{
  "best_global_step": 5724,
  "best_metric": 0.4691515564918518,
  "best_model_checkpoint": "saves_multiple/lora/llama-3-8b-instruct/train_codealpacapy_123_1762546585/checkpoint-5724",
  "epoch": 20.0,
  "eval_steps": 1908,
  "global_step": 38160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.002620545073375262,
      "grad_norm": 1.2127878665924072,
      "learning_rate": 5.241090146750525e-08,
      "loss": 1.2588,
      "num_input_tokens_seen": 2944,
      "step": 5
    },
    {
      "epoch": 0.005241090146750524,
      "grad_norm": 0.46947115659713745,
      "learning_rate": 1.1792452830188679e-07,
      "loss": 1.1171,
      "num_input_tokens_seen": 6816,
      "step": 10
    },
    {
      "epoch": 0.007861635220125786,
      "grad_norm": 0.7391409277915955,
      "learning_rate": 1.8343815513626837e-07,
      "loss": 1.2245,
      "num_input_tokens_seen": 9760,
      "step": 15
    },
    {
      "epoch": 0.010482180293501049,
      "grad_norm": 0.4756827652454376,
      "learning_rate": 2.489517819706499e-07,
      "loss": 1.1659,
      "num_input_tokens_seen": 13472,
      "step": 20
    },
    {
      "epoch": 0.01310272536687631,
      "grad_norm": 1.0817434787750244,
      "learning_rate": 3.144654088050315e-07,
      "loss": 1.2366,
      "num_input_tokens_seen": 16864,
      "step": 25
    },
    {
      "epoch": 0.015723270440251572,
      "grad_norm": 0.7966254949569702,
      "learning_rate": 3.79979035639413e-07,
      "loss": 1.3587,
      "num_input_tokens_seen": 20352,
      "step": 30
    },
    {
      "epoch": 0.018343815513626835,
      "grad_norm": 0.9355176687240601,
      "learning_rate": 4.454926624737945e-07,
      "loss": 1.3345,
      "num_input_tokens_seen": 23008,
      "step": 35
    },
    {
      "epoch": 0.020964360587002098,
      "grad_norm": 0.6985126733779907,
      "learning_rate": 5.110062893081761e-07,
      "loss": 1.0766,
      "num_input_tokens_seen": 25920,
      "step": 40
    },
    {
      "epoch": 0.02358490566037736,
      "grad_norm": 1.1450191736221313,
      "learning_rate": 5.765199161425577e-07,
      "loss": 1.2347,
      "num_input_tokens_seen": 28672,
      "step": 45
    },
    {
      "epoch": 0.02620545073375262,
      "grad_norm": 0.5917816162109375,
      "learning_rate": 6.420335429769392e-07,
      "loss": 1.0794,
      "num_input_tokens_seen": 31488,
      "step": 50
    },
    {
      "epoch": 0.028825995807127882,
      "grad_norm": 0.8176907896995544,
      "learning_rate": 7.075471698113208e-07,
      "loss": 1.2162,
      "num_input_tokens_seen": 34496,
      "step": 55
    },
    {
      "epoch": 0.031446540880503145,
      "grad_norm": 0.49047133326530457,
      "learning_rate": 7.730607966457024e-07,
      "loss": 1.2939,
      "num_input_tokens_seen": 38112,
      "step": 60
    },
    {
      "epoch": 0.034067085953878404,
      "grad_norm": 0.7897619605064392,
      "learning_rate": 8.38574423480084e-07,
      "loss": 1.3709,
      "num_input_tokens_seen": 40864,
      "step": 65
    },
    {
      "epoch": 0.03668763102725367,
      "grad_norm": 0.45301762223243713,
      "learning_rate": 9.040880503144654e-07,
      "loss": 1.2814,
      "num_input_tokens_seen": 44192,
      "step": 70
    },
    {
      "epoch": 0.03930817610062893,
      "grad_norm": 1.3057984113693237,
      "learning_rate": 9.69601677148847e-07,
      "loss": 1.407,
      "num_input_tokens_seen": 46784,
      "step": 75
    },
    {
      "epoch": 0.041928721174004195,
      "grad_norm": 2.336344003677368,
      "learning_rate": 1.0351153039832286e-06,
      "loss": 1.3949,
      "num_input_tokens_seen": 50176,
      "step": 80
    },
    {
      "epoch": 0.044549266247379454,
      "grad_norm": 0.4910266101360321,
      "learning_rate": 1.10062893081761e-06,
      "loss": 1.0891,
      "num_input_tokens_seen": 54400,
      "step": 85
    },
    {
      "epoch": 0.04716981132075472,
      "grad_norm": 1.2098689079284668,
      "learning_rate": 1.1661425576519918e-06,
      "loss": 1.4374,
      "num_input_tokens_seen": 56992,
      "step": 90
    },
    {
      "epoch": 0.04979035639412998,
      "grad_norm": 1.2719929218292236,
      "learning_rate": 1.2316561844863732e-06,
      "loss": 1.0148,
      "num_input_tokens_seen": 60896,
      "step": 95
    },
    {
      "epoch": 0.05241090146750524,
      "grad_norm": 0.5408589839935303,
      "learning_rate": 1.2971698113207548e-06,
      "loss": 1.1486,
      "num_input_tokens_seen": 64576,
      "step": 100
    },
    {
      "epoch": 0.055031446540880505,
      "grad_norm": 1.780749797821045,
      "learning_rate": 1.3626834381551364e-06,
      "loss": 1.3761,
      "num_input_tokens_seen": 67616,
      "step": 105
    },
    {
      "epoch": 0.057651991614255764,
      "grad_norm": 0.7092714309692383,
      "learning_rate": 1.4281970649895177e-06,
      "loss": 1.0211,
      "num_input_tokens_seen": 72544,
      "step": 110
    },
    {
      "epoch": 0.06027253668763103,
      "grad_norm": 0.7194042801856995,
      "learning_rate": 1.4937106918238995e-06,
      "loss": 1.2894,
      "num_input_tokens_seen": 75648,
      "step": 115
    },
    {
      "epoch": 0.06289308176100629,
      "grad_norm": 0.8944259881973267,
      "learning_rate": 1.559224318658281e-06,
      "loss": 1.3719,
      "num_input_tokens_seen": 78368,
      "step": 120
    },
    {
      "epoch": 0.06551362683438156,
      "grad_norm": 1.2700536251068115,
      "learning_rate": 1.6247379454926625e-06,
      "loss": 1.221,
      "num_input_tokens_seen": 81952,
      "step": 125
    },
    {
      "epoch": 0.06813417190775681,
      "grad_norm": 1.148658275604248,
      "learning_rate": 1.690251572327044e-06,
      "loss": 1.1627,
      "num_input_tokens_seen": 84960,
      "step": 130
    },
    {
      "epoch": 0.07075471698113207,
      "grad_norm": 1.3430958986282349,
      "learning_rate": 1.7557651991614257e-06,
      "loss": 1.2821,
      "num_input_tokens_seen": 88576,
      "step": 135
    },
    {
      "epoch": 0.07337526205450734,
      "grad_norm": 1.4252493381500244,
      "learning_rate": 1.821278825995807e-06,
      "loss": 0.9797,
      "num_input_tokens_seen": 92160,
      "step": 140
    },
    {
      "epoch": 0.0759958071278826,
      "grad_norm": 0.7527592778205872,
      "learning_rate": 1.8867924528301887e-06,
      "loss": 1.0781,
      "num_input_tokens_seen": 95840,
      "step": 145
    },
    {
      "epoch": 0.07861635220125786,
      "grad_norm": 1.3135706186294556,
      "learning_rate": 1.9523060796645703e-06,
      "loss": 1.1061,
      "num_input_tokens_seen": 98816,
      "step": 150
    },
    {
      "epoch": 0.08123689727463312,
      "grad_norm": 0.8705893158912659,
      "learning_rate": 2.017819706498952e-06,
      "loss": 0.9152,
      "num_input_tokens_seen": 102880,
      "step": 155
    },
    {
      "epoch": 0.08385744234800839,
      "grad_norm": 1.5332436561584473,
      "learning_rate": 2.0833333333333334e-06,
      "loss": 1.0812,
      "num_input_tokens_seen": 105920,
      "step": 160
    },
    {
      "epoch": 0.08647798742138364,
      "grad_norm": 0.8768753409385681,
      "learning_rate": 2.148846960167715e-06,
      "loss": 1.2431,
      "num_input_tokens_seen": 108800,
      "step": 165
    },
    {
      "epoch": 0.08909853249475891,
      "grad_norm": 1.3434324264526367,
      "learning_rate": 2.2143605870020966e-06,
      "loss": 0.9619,
      "num_input_tokens_seen": 111968,
      "step": 170
    },
    {
      "epoch": 0.09171907756813417,
      "grad_norm": 1.803437352180481,
      "learning_rate": 2.279874213836478e-06,
      "loss": 1.0127,
      "num_input_tokens_seen": 114976,
      "step": 175
    },
    {
      "epoch": 0.09433962264150944,
      "grad_norm": 4.882724761962891,
      "learning_rate": 2.34538784067086e-06,
      "loss": 1.16,
      "num_input_tokens_seen": 117568,
      "step": 180
    },
    {
      "epoch": 0.09696016771488469,
      "grad_norm": 1.5289909839630127,
      "learning_rate": 2.410901467505241e-06,
      "loss": 0.8258,
      "num_input_tokens_seen": 120384,
      "step": 185
    },
    {
      "epoch": 0.09958071278825996,
      "grad_norm": 1.1588318347930908,
      "learning_rate": 2.476415094339623e-06,
      "loss": 0.8277,
      "num_input_tokens_seen": 123680,
      "step": 190
    },
    {
      "epoch": 0.10220125786163523,
      "grad_norm": 1.8892508745193481,
      "learning_rate": 2.5419287211740044e-06,
      "loss": 0.9592,
      "num_input_tokens_seen": 127168,
      "step": 195
    },
    {
      "epoch": 0.10482180293501048,
      "grad_norm": 1.3397516012191772,
      "learning_rate": 2.6074423480083857e-06,
      "loss": 0.7008,
      "num_input_tokens_seen": 130176,
      "step": 200
    },
    {
      "epoch": 0.10744234800838574,
      "grad_norm": 1.6290249824523926,
      "learning_rate": 2.6729559748427675e-06,
      "loss": 0.8399,
      "num_input_tokens_seen": 133056,
      "step": 205
    },
    {
      "epoch": 0.11006289308176101,
      "grad_norm": 1.174363136291504,
      "learning_rate": 2.738469601677149e-06,
      "loss": 0.665,
      "num_input_tokens_seen": 136864,
      "step": 210
    },
    {
      "epoch": 0.11268343815513626,
      "grad_norm": 1.7376288175582886,
      "learning_rate": 2.8039832285115303e-06,
      "loss": 0.9658,
      "num_input_tokens_seen": 140064,
      "step": 215
    },
    {
      "epoch": 0.11530398322851153,
      "grad_norm": 1.4654614925384521,
      "learning_rate": 2.869496855345912e-06,
      "loss": 0.7364,
      "num_input_tokens_seen": 143712,
      "step": 220
    },
    {
      "epoch": 0.1179245283018868,
      "grad_norm": 1.238377332687378,
      "learning_rate": 2.935010482180294e-06,
      "loss": 0.7104,
      "num_input_tokens_seen": 146880,
      "step": 225
    },
    {
      "epoch": 0.12054507337526206,
      "grad_norm": 1.3718808889389038,
      "learning_rate": 3.0005241090146753e-06,
      "loss": 0.6654,
      "num_input_tokens_seen": 150560,
      "step": 230
    },
    {
      "epoch": 0.12316561844863731,
      "grad_norm": 1.0669552087783813,
      "learning_rate": 3.0660377358490567e-06,
      "loss": 0.8498,
      "num_input_tokens_seen": 153664,
      "step": 235
    },
    {
      "epoch": 0.12578616352201258,
      "grad_norm": 2.0938475131988525,
      "learning_rate": 3.131551362683438e-06,
      "loss": 0.5954,
      "num_input_tokens_seen": 157568,
      "step": 240
    },
    {
      "epoch": 0.12840670859538783,
      "grad_norm": 1.0705981254577637,
      "learning_rate": 3.1970649895178203e-06,
      "loss": 0.6299,
      "num_input_tokens_seen": 160512,
      "step": 245
    },
    {
      "epoch": 0.1310272536687631,
      "grad_norm": 0.4477166533470154,
      "learning_rate": 3.2625786163522012e-06,
      "loss": 0.5946,
      "num_input_tokens_seen": 164512,
      "step": 250
    },
    {
      "epoch": 0.13364779874213836,
      "grad_norm": 2.282764196395874,
      "learning_rate": 3.3280922431865826e-06,
      "loss": 0.562,
      "num_input_tokens_seen": 167200,
      "step": 255
    },
    {
      "epoch": 0.13626834381551362,
      "grad_norm": 1.4697599411010742,
      "learning_rate": 3.393605870020965e-06,
      "loss": 0.6459,
      "num_input_tokens_seen": 169984,
      "step": 260
    },
    {
      "epoch": 0.1388888888888889,
      "grad_norm": 0.7798205614089966,
      "learning_rate": 3.459119496855346e-06,
      "loss": 0.5688,
      "num_input_tokens_seen": 173856,
      "step": 265
    },
    {
      "epoch": 0.14150943396226415,
      "grad_norm": 1.2953965663909912,
      "learning_rate": 3.524633123689727e-06,
      "loss": 0.662,
      "num_input_tokens_seen": 176192,
      "step": 270
    },
    {
      "epoch": 0.1441299790356394,
      "grad_norm": 0.6025437116622925,
      "learning_rate": 3.5901467505241094e-06,
      "loss": 0.4491,
      "num_input_tokens_seen": 179840,
      "step": 275
    },
    {
      "epoch": 0.14675052410901468,
      "grad_norm": 0.8777902126312256,
      "learning_rate": 3.6556603773584908e-06,
      "loss": 0.5351,
      "num_input_tokens_seen": 182976,
      "step": 280
    },
    {
      "epoch": 0.14937106918238993,
      "grad_norm": 0.8129740953445435,
      "learning_rate": 3.7211740041928726e-06,
      "loss": 0.5822,
      "num_input_tokens_seen": 186400,
      "step": 285
    },
    {
      "epoch": 0.1519916142557652,
      "grad_norm": 1.0465656518936157,
      "learning_rate": 3.786687631027254e-06,
      "loss": 0.5798,
      "num_input_tokens_seen": 190144,
      "step": 290
    },
    {
      "epoch": 0.15461215932914046,
      "grad_norm": 1.4936944246292114,
      "learning_rate": 3.852201257861635e-06,
      "loss": 0.6191,
      "num_input_tokens_seen": 193344,
      "step": 295
    },
    {
      "epoch": 0.15723270440251572,
      "grad_norm": 1.450209140777588,
      "learning_rate": 3.917714884696017e-06,
      "loss": 0.4113,
      "num_input_tokens_seen": 195904,
      "step": 300
    },
    {
      "epoch": 0.159853249475891,
      "grad_norm": 1.8591293096542358,
      "learning_rate": 3.9832285115303985e-06,
      "loss": 0.5717,
      "num_input_tokens_seen": 198944,
      "step": 305
    },
    {
      "epoch": 0.16247379454926625,
      "grad_norm": 1.3647106885910034,
      "learning_rate": 4.04874213836478e-06,
      "loss": 0.5535,
      "num_input_tokens_seen": 201792,
      "step": 310
    },
    {
      "epoch": 0.1650943396226415,
      "grad_norm": 0.9549600481987,
      "learning_rate": 4.114255765199162e-06,
      "loss": 0.5157,
      "num_input_tokens_seen": 206080,
      "step": 315
    },
    {
      "epoch": 0.16771488469601678,
      "grad_norm": 0.818556547164917,
      "learning_rate": 4.1797693920335435e-06,
      "loss": 0.6303,
      "num_input_tokens_seen": 209376,
      "step": 320
    },
    {
      "epoch": 0.17033542976939203,
      "grad_norm": 1.1256197690963745,
      "learning_rate": 4.245283018867925e-06,
      "loss": 0.6296,
      "num_input_tokens_seen": 212160,
      "step": 325
    },
    {
      "epoch": 0.17295597484276728,
      "grad_norm": 0.7102301120758057,
      "learning_rate": 4.310796645702306e-06,
      "loss": 0.5717,
      "num_input_tokens_seen": 216128,
      "step": 330
    },
    {
      "epoch": 0.17557651991614256,
      "grad_norm": 0.7263284921646118,
      "learning_rate": 4.376310272536688e-06,
      "loss": 0.7549,
      "num_input_tokens_seen": 219232,
      "step": 335
    },
    {
      "epoch": 0.17819706498951782,
      "grad_norm": 0.6849648356437683,
      "learning_rate": 4.441823899371069e-06,
      "loss": 0.489,
      "num_input_tokens_seen": 222464,
      "step": 340
    },
    {
      "epoch": 0.18081761006289307,
      "grad_norm": 0.7484946846961975,
      "learning_rate": 4.507337526205451e-06,
      "loss": 0.7478,
      "num_input_tokens_seen": 225184,
      "step": 345
    },
    {
      "epoch": 0.18343815513626835,
      "grad_norm": 0.7505791187286377,
      "learning_rate": 4.572851153039833e-06,
      "loss": 0.5011,
      "num_input_tokens_seen": 228000,
      "step": 350
    },
    {
      "epoch": 0.1860587002096436,
      "grad_norm": 0.793000340461731,
      "learning_rate": 4.638364779874214e-06,
      "loss": 0.4632,
      "num_input_tokens_seen": 230944,
      "step": 355
    },
    {
      "epoch": 0.18867924528301888,
      "grad_norm": 1.2271275520324707,
      "learning_rate": 4.703878406708595e-06,
      "loss": 0.4807,
      "num_input_tokens_seen": 233728,
      "step": 360
    },
    {
      "epoch": 0.19129979035639413,
      "grad_norm": 0.9340527057647705,
      "learning_rate": 4.769392033542977e-06,
      "loss": 0.509,
      "num_input_tokens_seen": 236480,
      "step": 365
    },
    {
      "epoch": 0.19392033542976939,
      "grad_norm": 1.3452706336975098,
      "learning_rate": 4.834905660377359e-06,
      "loss": 0.5022,
      "num_input_tokens_seen": 239488,
      "step": 370
    },
    {
      "epoch": 0.19654088050314467,
      "grad_norm": 1.2965561151504517,
      "learning_rate": 4.90041928721174e-06,
      "loss": 0.6116,
      "num_input_tokens_seen": 242144,
      "step": 375
    },
    {
      "epoch": 0.19916142557651992,
      "grad_norm": 0.6673212051391602,
      "learning_rate": 4.965932914046122e-06,
      "loss": 0.5404,
      "num_input_tokens_seen": 245216,
      "step": 380
    },
    {
      "epoch": 0.20178197064989517,
      "grad_norm": 1.0752499103546143,
      "learning_rate": 5.031446540880504e-06,
      "loss": 0.6767,
      "num_input_tokens_seen": 249632,
      "step": 385
    },
    {
      "epoch": 0.20440251572327045,
      "grad_norm": 0.5316757559776306,
      "learning_rate": 5.096960167714885e-06,
      "loss": 0.5798,
      "num_input_tokens_seen": 253312,
      "step": 390
    },
    {
      "epoch": 0.2070230607966457,
      "grad_norm": 0.8235381841659546,
      "learning_rate": 5.162473794549266e-06,
      "loss": 0.5217,
      "num_input_tokens_seen": 256384,
      "step": 395
    },
    {
      "epoch": 0.20964360587002095,
      "grad_norm": 5.637340545654297,
      "learning_rate": 5.227987421383648e-06,
      "loss": 0.5806,
      "num_input_tokens_seen": 259744,
      "step": 400
    },
    {
      "epoch": 0.21226415094339623,
      "grad_norm": 0.9684422612190247,
      "learning_rate": 5.2935010482180295e-06,
      "loss": 0.7239,
      "num_input_tokens_seen": 263456,
      "step": 405
    },
    {
      "epoch": 0.2148846960167715,
      "grad_norm": 0.9401962757110596,
      "learning_rate": 5.359014675052411e-06,
      "loss": 0.5317,
      "num_input_tokens_seen": 266496,
      "step": 410
    },
    {
      "epoch": 0.21750524109014674,
      "grad_norm": 0.9476348757743835,
      "learning_rate": 5.424528301886793e-06,
      "loss": 0.6346,
      "num_input_tokens_seen": 269600,
      "step": 415
    },
    {
      "epoch": 0.22012578616352202,
      "grad_norm": 0.8082473278045654,
      "learning_rate": 5.4900419287211744e-06,
      "loss": 0.5527,
      "num_input_tokens_seen": 273152,
      "step": 420
    },
    {
      "epoch": 0.22274633123689727,
      "grad_norm": 0.7683749198913574,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.7132,
      "num_input_tokens_seen": 276992,
      "step": 425
    },
    {
      "epoch": 0.22536687631027252,
      "grad_norm": 1.2673919200897217,
      "learning_rate": 5.621069182389937e-06,
      "loss": 0.6571,
      "num_input_tokens_seen": 279296,
      "step": 430
    },
    {
      "epoch": 0.2279874213836478,
      "grad_norm": 0.5412548780441284,
      "learning_rate": 5.686582809224319e-06,
      "loss": 0.4393,
      "num_input_tokens_seen": 283296,
      "step": 435
    },
    {
      "epoch": 0.23060796645702306,
      "grad_norm": 1.7898181676864624,
      "learning_rate": 5.752096436058701e-06,
      "loss": 0.729,
      "num_input_tokens_seen": 286016,
      "step": 440
    },
    {
      "epoch": 0.23322851153039834,
      "grad_norm": 1.3507839441299438,
      "learning_rate": 5.817610062893082e-06,
      "loss": 0.5834,
      "num_input_tokens_seen": 289248,
      "step": 445
    },
    {
      "epoch": 0.2358490566037736,
      "grad_norm": 0.7601041197776794,
      "learning_rate": 5.8831236897274636e-06,
      "loss": 0.6264,
      "num_input_tokens_seen": 292160,
      "step": 450
    },
    {
      "epoch": 0.23846960167714884,
      "grad_norm": 0.968923807144165,
      "learning_rate": 5.948637316561846e-06,
      "loss": 0.5522,
      "num_input_tokens_seen": 295296,
      "step": 455
    },
    {
      "epoch": 0.24109014675052412,
      "grad_norm": 0.8205217123031616,
      "learning_rate": 6.014150943396227e-06,
      "loss": 0.6081,
      "num_input_tokens_seen": 299360,
      "step": 460
    },
    {
      "epoch": 0.24371069182389937,
      "grad_norm": 0.6574501395225525,
      "learning_rate": 6.079664570230608e-06,
      "loss": 0.5023,
      "num_input_tokens_seen": 303072,
      "step": 465
    },
    {
      "epoch": 0.24633123689727462,
      "grad_norm": 0.7788654565811157,
      "learning_rate": 6.14517819706499e-06,
      "loss": 0.4862,
      "num_input_tokens_seen": 307616,
      "step": 470
    },
    {
      "epoch": 0.2489517819706499,
      "grad_norm": 0.716489851474762,
      "learning_rate": 6.210691823899371e-06,
      "loss": 0.5074,
      "num_input_tokens_seen": 310112,
      "step": 475
    },
    {
      "epoch": 0.25157232704402516,
      "grad_norm": 0.7830471992492676,
      "learning_rate": 6.2762054507337535e-06,
      "loss": 0.4848,
      "num_input_tokens_seen": 313472,
      "step": 480
    },
    {
      "epoch": 0.25419287211740044,
      "grad_norm": 0.8694356083869934,
      "learning_rate": 6.341719077568134e-06,
      "loss": 0.5105,
      "num_input_tokens_seen": 316128,
      "step": 485
    },
    {
      "epoch": 0.25681341719077566,
      "grad_norm": 0.7905958890914917,
      "learning_rate": 6.407232704402516e-06,
      "loss": 0.475,
      "num_input_tokens_seen": 319584,
      "step": 490
    },
    {
      "epoch": 0.25943396226415094,
      "grad_norm": 0.6687126159667969,
      "learning_rate": 6.472746331236898e-06,
      "loss": 0.5692,
      "num_input_tokens_seen": 322496,
      "step": 495
    },
    {
      "epoch": 0.2620545073375262,
      "grad_norm": 0.8297496438026428,
      "learning_rate": 6.538259958071278e-06,
      "loss": 0.8498,
      "num_input_tokens_seen": 325120,
      "step": 500
    },
    {
      "epoch": 0.26467505241090145,
      "grad_norm": 1.034053921699524,
      "learning_rate": 6.60377358490566e-06,
      "loss": 0.6094,
      "num_input_tokens_seen": 327840,
      "step": 505
    },
    {
      "epoch": 0.2672955974842767,
      "grad_norm": 0.67457115650177,
      "learning_rate": 6.669287211740043e-06,
      "loss": 0.4201,
      "num_input_tokens_seen": 331776,
      "step": 510
    },
    {
      "epoch": 0.269916142557652,
      "grad_norm": 1.9690123796463013,
      "learning_rate": 6.734800838574423e-06,
      "loss": 0.5139,
      "num_input_tokens_seen": 336544,
      "step": 515
    },
    {
      "epoch": 0.27253668763102723,
      "grad_norm": 0.6818830370903015,
      "learning_rate": 6.800314465408805e-06,
      "loss": 0.6796,
      "num_input_tokens_seen": 340768,
      "step": 520
    },
    {
      "epoch": 0.2751572327044025,
      "grad_norm": 1.1146661043167114,
      "learning_rate": 6.865828092243188e-06,
      "loss": 0.6693,
      "num_input_tokens_seen": 343296,
      "step": 525
    },
    {
      "epoch": 0.2777777777777778,
      "grad_norm": 0.8182678818702698,
      "learning_rate": 6.931341719077568e-06,
      "loss": 0.3967,
      "num_input_tokens_seen": 346848,
      "step": 530
    },
    {
      "epoch": 0.280398322851153,
      "grad_norm": 0.744745671749115,
      "learning_rate": 6.9968553459119495e-06,
      "loss": 0.5686,
      "num_input_tokens_seen": 349632,
      "step": 535
    },
    {
      "epoch": 0.2830188679245283,
      "grad_norm": 0.8877318501472473,
      "learning_rate": 7.062368972746332e-06,
      "loss": 0.5291,
      "num_input_tokens_seen": 353024,
      "step": 540
    },
    {
      "epoch": 0.2856394129979036,
      "grad_norm": 0.7972932457923889,
      "learning_rate": 7.127882599580712e-06,
      "loss": 0.5171,
      "num_input_tokens_seen": 355552,
      "step": 545
    },
    {
      "epoch": 0.2882599580712788,
      "grad_norm": 1.0433536767959595,
      "learning_rate": 7.1933962264150945e-06,
      "loss": 0.4898,
      "num_input_tokens_seen": 360128,
      "step": 550
    },
    {
      "epoch": 0.2908805031446541,
      "grad_norm": 0.5782132148742676,
      "learning_rate": 7.258909853249477e-06,
      "loss": 0.5362,
      "num_input_tokens_seen": 363360,
      "step": 555
    },
    {
      "epoch": 0.29350104821802936,
      "grad_norm": 0.9726479053497314,
      "learning_rate": 7.324423480083858e-06,
      "loss": 0.5338,
      "num_input_tokens_seen": 366144,
      "step": 560
    },
    {
      "epoch": 0.29612159329140464,
      "grad_norm": 0.7549358010292053,
      "learning_rate": 7.3899371069182395e-06,
      "loss": 0.5507,
      "num_input_tokens_seen": 369216,
      "step": 565
    },
    {
      "epoch": 0.29874213836477986,
      "grad_norm": 1.3198691606521606,
      "learning_rate": 7.455450733752621e-06,
      "loss": 0.6116,
      "num_input_tokens_seen": 372512,
      "step": 570
    },
    {
      "epoch": 0.30136268343815514,
      "grad_norm": 1.2303029298782349,
      "learning_rate": 7.520964360587003e-06,
      "loss": 0.5496,
      "num_input_tokens_seen": 374976,
      "step": 575
    },
    {
      "epoch": 0.3039832285115304,
      "grad_norm": 1.6215330362319946,
      "learning_rate": 7.586477987421384e-06,
      "loss": 0.4374,
      "num_input_tokens_seen": 377632,
      "step": 580
    },
    {
      "epoch": 0.30660377358490565,
      "grad_norm": 0.6961995363235474,
      "learning_rate": 7.651991614255765e-06,
      "loss": 0.6749,
      "num_input_tokens_seen": 380768,
      "step": 585
    },
    {
      "epoch": 0.30922431865828093,
      "grad_norm": 0.8278395533561707,
      "learning_rate": 7.717505241090148e-06,
      "loss": 0.5122,
      "num_input_tokens_seen": 383520,
      "step": 590
    },
    {
      "epoch": 0.3118448637316562,
      "grad_norm": 1.8187737464904785,
      "learning_rate": 7.783018867924528e-06,
      "loss": 0.7816,
      "num_input_tokens_seen": 386496,
      "step": 595
    },
    {
      "epoch": 0.31446540880503143,
      "grad_norm": 0.9431346654891968,
      "learning_rate": 7.848532494758911e-06,
      "loss": 0.5949,
      "num_input_tokens_seen": 390464,
      "step": 600
    },
    {
      "epoch": 0.3170859538784067,
      "grad_norm": 1.0517154932022095,
      "learning_rate": 7.914046121593292e-06,
      "loss": 0.4512,
      "num_input_tokens_seen": 393440,
      "step": 605
    },
    {
      "epoch": 0.319706498951782,
      "grad_norm": 0.9073853492736816,
      "learning_rate": 7.979559748427674e-06,
      "loss": 0.461,
      "num_input_tokens_seen": 398336,
      "step": 610
    },
    {
      "epoch": 0.3223270440251572,
      "grad_norm": 1.4231126308441162,
      "learning_rate": 8.045073375262055e-06,
      "loss": 0.4246,
      "num_input_tokens_seen": 401504,
      "step": 615
    },
    {
      "epoch": 0.3249475890985325,
      "grad_norm": 1.3209675550460815,
      "learning_rate": 8.110587002096436e-06,
      "loss": 0.5048,
      "num_input_tokens_seen": 404864,
      "step": 620
    },
    {
      "epoch": 0.3275681341719078,
      "grad_norm": 0.5791663527488708,
      "learning_rate": 8.176100628930818e-06,
      "loss": 0.5725,
      "num_input_tokens_seen": 409056,
      "step": 625
    },
    {
      "epoch": 0.330188679245283,
      "grad_norm": 0.8607786893844604,
      "learning_rate": 8.241614255765199e-06,
      "loss": 0.5641,
      "num_input_tokens_seen": 412544,
      "step": 630
    },
    {
      "epoch": 0.3328092243186583,
      "grad_norm": 1.5001403093338013,
      "learning_rate": 8.307127882599582e-06,
      "loss": 0.5394,
      "num_input_tokens_seen": 415520,
      "step": 635
    },
    {
      "epoch": 0.33542976939203356,
      "grad_norm": 1.2909969091415405,
      "learning_rate": 8.372641509433962e-06,
      "loss": 0.9309,
      "num_input_tokens_seen": 417760,
      "step": 640
    },
    {
      "epoch": 0.3380503144654088,
      "grad_norm": 0.6109410524368286,
      "learning_rate": 8.438155136268345e-06,
      "loss": 0.5685,
      "num_input_tokens_seen": 421728,
      "step": 645
    },
    {
      "epoch": 0.34067085953878407,
      "grad_norm": 4.333765029907227,
      "learning_rate": 8.503668763102726e-06,
      "loss": 0.5014,
      "num_input_tokens_seen": 425664,
      "step": 650
    },
    {
      "epoch": 0.34329140461215935,
      "grad_norm": 1.0831236839294434,
      "learning_rate": 8.569182389937106e-06,
      "loss": 0.4649,
      "num_input_tokens_seen": 428896,
      "step": 655
    },
    {
      "epoch": 0.34591194968553457,
      "grad_norm": 1.4196419715881348,
      "learning_rate": 8.634696016771489e-06,
      "loss": 0.5261,
      "num_input_tokens_seen": 431744,
      "step": 660
    },
    {
      "epoch": 0.34853249475890985,
      "grad_norm": 0.8383955359458923,
      "learning_rate": 8.70020964360587e-06,
      "loss": 0.5443,
      "num_input_tokens_seen": 435520,
      "step": 665
    },
    {
      "epoch": 0.35115303983228513,
      "grad_norm": 0.66987144947052,
      "learning_rate": 8.765723270440252e-06,
      "loss": 0.4883,
      "num_input_tokens_seen": 438592,
      "step": 670
    },
    {
      "epoch": 0.35377358490566035,
      "grad_norm": 0.7857100963592529,
      "learning_rate": 8.831236897274633e-06,
      "loss": 0.5312,
      "num_input_tokens_seen": 441664,
      "step": 675
    },
    {
      "epoch": 0.35639412997903563,
      "grad_norm": 1.537290334701538,
      "learning_rate": 8.896750524109015e-06,
      "loss": 0.59,
      "num_input_tokens_seen": 444832,
      "step": 680
    },
    {
      "epoch": 0.3590146750524109,
      "grad_norm": 1.0513705015182495,
      "learning_rate": 8.962264150943396e-06,
      "loss": 0.6046,
      "num_input_tokens_seen": 447392,
      "step": 685
    },
    {
      "epoch": 0.36163522012578614,
      "grad_norm": 1.2972558736801147,
      "learning_rate": 9.027777777777777e-06,
      "loss": 0.5575,
      "num_input_tokens_seen": 450528,
      "step": 690
    },
    {
      "epoch": 0.3642557651991614,
      "grad_norm": 1.0609298944473267,
      "learning_rate": 9.09329140461216e-06,
      "loss": 0.4183,
      "num_input_tokens_seen": 453504,
      "step": 695
    },
    {
      "epoch": 0.3668763102725367,
      "grad_norm": 1.2141389846801758,
      "learning_rate": 9.15880503144654e-06,
      "loss": 0.6326,
      "num_input_tokens_seen": 456288,
      "step": 700
    },
    {
      "epoch": 0.3694968553459119,
      "grad_norm": 1.4611376523971558,
      "learning_rate": 9.224318658280923e-06,
      "loss": 0.4041,
      "num_input_tokens_seen": 459840,
      "step": 705
    },
    {
      "epoch": 0.3721174004192872,
      "grad_norm": 0.6853867769241333,
      "learning_rate": 9.289832285115305e-06,
      "loss": 0.4243,
      "num_input_tokens_seen": 463168,
      "step": 710
    },
    {
      "epoch": 0.3747379454926625,
      "grad_norm": 1.072186827659607,
      "learning_rate": 9.355345911949686e-06,
      "loss": 0.4882,
      "num_input_tokens_seen": 466112,
      "step": 715
    },
    {
      "epoch": 0.37735849056603776,
      "grad_norm": 1.300741195678711,
      "learning_rate": 9.420859538784067e-06,
      "loss": 0.6441,
      "num_input_tokens_seen": 469568,
      "step": 720
    },
    {
      "epoch": 0.379979035639413,
      "grad_norm": 0.8578652143478394,
      "learning_rate": 9.486373165618449e-06,
      "loss": 0.4169,
      "num_input_tokens_seen": 472896,
      "step": 725
    },
    {
      "epoch": 0.38259958071278827,
      "grad_norm": 1.1069036722183228,
      "learning_rate": 9.551886792452832e-06,
      "loss": 0.3374,
      "num_input_tokens_seen": 475744,
      "step": 730
    },
    {
      "epoch": 0.38522012578616355,
      "grad_norm": 0.3411802649497986,
      "learning_rate": 9.617400419287211e-06,
      "loss": 0.5299,
      "num_input_tokens_seen": 481216,
      "step": 735
    },
    {
      "epoch": 0.38784067085953877,
      "grad_norm": 1.147443413734436,
      "learning_rate": 9.682914046121595e-06,
      "loss": 0.5329,
      "num_input_tokens_seen": 483488,
      "step": 740
    },
    {
      "epoch": 0.39046121593291405,
      "grad_norm": 0.8487539887428284,
      "learning_rate": 9.748427672955976e-06,
      "loss": 0.5875,
      "num_input_tokens_seen": 487072,
      "step": 745
    },
    {
      "epoch": 0.39308176100628933,
      "grad_norm": 0.8702980875968933,
      "learning_rate": 9.813941299790357e-06,
      "loss": 0.5249,
      "num_input_tokens_seen": 490496,
      "step": 750
    },
    {
      "epoch": 0.39570230607966456,
      "grad_norm": 0.7943355441093445,
      "learning_rate": 9.879454926624739e-06,
      "loss": 0.4644,
      "num_input_tokens_seen": 493824,
      "step": 755
    },
    {
      "epoch": 0.39832285115303984,
      "grad_norm": 1.2097253799438477,
      "learning_rate": 9.94496855345912e-06,
      "loss": 0.6809,
      "num_input_tokens_seen": 496192,
      "step": 760
    },
    {
      "epoch": 0.4009433962264151,
      "grad_norm": 0.7809175252914429,
      "learning_rate": 1.0010482180293501e-05,
      "loss": 0.5845,
      "num_input_tokens_seen": 498880,
      "step": 765
    },
    {
      "epoch": 0.40356394129979034,
      "grad_norm": 0.8147527575492859,
      "learning_rate": 1.0075995807127883e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 502016,
      "step": 770
    },
    {
      "epoch": 0.4061844863731656,
      "grad_norm": 0.5472318530082703,
      "learning_rate": 1.0141509433962266e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 505120,
      "step": 775
    },
    {
      "epoch": 0.4088050314465409,
      "grad_norm": 0.9467899799346924,
      "learning_rate": 1.0207023060796646e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 508256,
      "step": 780
    },
    {
      "epoch": 0.4114255765199161,
      "grad_norm": 0.9587962627410889,
      "learning_rate": 1.0272536687631027e-05,
      "loss": 0.5066,
      "num_input_tokens_seen": 511776,
      "step": 785
    },
    {
      "epoch": 0.4140461215932914,
      "grad_norm": 1.6650651693344116,
      "learning_rate": 1.033805031446541e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 514464,
      "step": 790
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 0.9744753241539001,
      "learning_rate": 1.040356394129979e-05,
      "loss": 0.5012,
      "num_input_tokens_seen": 518144,
      "step": 795
    },
    {
      "epoch": 0.4192872117400419,
      "grad_norm": 1.2468008995056152,
      "learning_rate": 1.0469077568134173e-05,
      "loss": 0.5248,
      "num_input_tokens_seen": 520768,
      "step": 800
    },
    {
      "epoch": 0.4219077568134172,
      "grad_norm": 1.388175368309021,
      "learning_rate": 1.0534591194968554e-05,
      "loss": 0.6468,
      "num_input_tokens_seen": 523808,
      "step": 805
    },
    {
      "epoch": 0.42452830188679247,
      "grad_norm": 1.035539150238037,
      "learning_rate": 1.0600104821802936e-05,
      "loss": 0.3761,
      "num_input_tokens_seen": 527136,
      "step": 810
    },
    {
      "epoch": 0.4271488469601677,
      "grad_norm": 2.3244001865386963,
      "learning_rate": 1.0665618448637317e-05,
      "loss": 0.6178,
      "num_input_tokens_seen": 530240,
      "step": 815
    },
    {
      "epoch": 0.429769392033543,
      "grad_norm": 1.188620686531067,
      "learning_rate": 1.0731132075471698e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 534496,
      "step": 820
    },
    {
      "epoch": 0.43238993710691825,
      "grad_norm": 1.5427753925323486,
      "learning_rate": 1.079664570230608e-05,
      "loss": 0.539,
      "num_input_tokens_seen": 538624,
      "step": 825
    },
    {
      "epoch": 0.4350104821802935,
      "grad_norm": 1.5885844230651855,
      "learning_rate": 1.0862159329140461e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 542112,
      "step": 830
    },
    {
      "epoch": 0.43763102725366876,
      "grad_norm": 1.0200605392456055,
      "learning_rate": 1.0927672955974844e-05,
      "loss": 0.6292,
      "num_input_tokens_seen": 545408,
      "step": 835
    },
    {
      "epoch": 0.44025157232704404,
      "grad_norm": 0.6758912205696106,
      "learning_rate": 1.0993186582809224e-05,
      "loss": 0.5048,
      "num_input_tokens_seen": 548416,
      "step": 840
    },
    {
      "epoch": 0.44287211740041926,
      "grad_norm": 1.0210635662078857,
      "learning_rate": 1.1058700209643607e-05,
      "loss": 0.5326,
      "num_input_tokens_seen": 551264,
      "step": 845
    },
    {
      "epoch": 0.44549266247379454,
      "grad_norm": 0.7288973927497864,
      "learning_rate": 1.1124213836477988e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 554592,
      "step": 850
    },
    {
      "epoch": 0.4481132075471698,
      "grad_norm": 1.3487876653671265,
      "learning_rate": 1.118972746331237e-05,
      "loss": 0.382,
      "num_input_tokens_seen": 558304,
      "step": 855
    },
    {
      "epoch": 0.45073375262054505,
      "grad_norm": 1.0377235412597656,
      "learning_rate": 1.1255241090146751e-05,
      "loss": 0.5466,
      "num_input_tokens_seen": 561248,
      "step": 860
    },
    {
      "epoch": 0.4533542976939203,
      "grad_norm": 1.1714564561843872,
      "learning_rate": 1.1320754716981132e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 564352,
      "step": 865
    },
    {
      "epoch": 0.4559748427672956,
      "grad_norm": 2.348745346069336,
      "learning_rate": 1.1386268343815514e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 567712,
      "step": 870
    },
    {
      "epoch": 0.4585953878406709,
      "grad_norm": 1.525247573852539,
      "learning_rate": 1.1451781970649895e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 570240,
      "step": 875
    },
    {
      "epoch": 0.4612159329140461,
      "grad_norm": 1.7257378101348877,
      "learning_rate": 1.1517295597484278e-05,
      "loss": 0.5573,
      "num_input_tokens_seen": 572768,
      "step": 880
    },
    {
      "epoch": 0.4638364779874214,
      "grad_norm": 1.4012850522994995,
      "learning_rate": 1.158280922431866e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 576512,
      "step": 885
    },
    {
      "epoch": 0.46645702306079667,
      "grad_norm": 1.379889965057373,
      "learning_rate": 1.164832285115304e-05,
      "loss": 0.5497,
      "num_input_tokens_seen": 579264,
      "step": 890
    },
    {
      "epoch": 0.4690775681341719,
      "grad_norm": 0.8231465220451355,
      "learning_rate": 1.1713836477987422e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 581440,
      "step": 895
    },
    {
      "epoch": 0.4716981132075472,
      "grad_norm": 0.7701451778411865,
      "learning_rate": 1.1779350104821804e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 584544,
      "step": 900
    },
    {
      "epoch": 0.47431865828092246,
      "grad_norm": 1.3979400396347046,
      "learning_rate": 1.1844863731656185e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 588160,
      "step": 905
    },
    {
      "epoch": 0.4769392033542977,
      "grad_norm": 2.0374956130981445,
      "learning_rate": 1.1910377358490566e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 590400,
      "step": 910
    },
    {
      "epoch": 0.47955974842767296,
      "grad_norm": 0.9611548781394958,
      "learning_rate": 1.197589098532495e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 593440,
      "step": 915
    },
    {
      "epoch": 0.48218029350104824,
      "grad_norm": 1.410500168800354,
      "learning_rate": 1.204140461215933e-05,
      "loss": 0.4679,
      "num_input_tokens_seen": 596480,
      "step": 920
    },
    {
      "epoch": 0.48480083857442346,
      "grad_norm": 0.896182119846344,
      "learning_rate": 1.210691823899371e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 599392,
      "step": 925
    },
    {
      "epoch": 0.48742138364779874,
      "grad_norm": 1.2971687316894531,
      "learning_rate": 1.2172431865828094e-05,
      "loss": 0.4062,
      "num_input_tokens_seen": 602432,
      "step": 930
    },
    {
      "epoch": 0.490041928721174,
      "grad_norm": 0.9469761848449707,
      "learning_rate": 1.2237945492662473e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 604512,
      "step": 935
    },
    {
      "epoch": 0.49266247379454925,
      "grad_norm": 0.594935417175293,
      "learning_rate": 1.2303459119496856e-05,
      "loss": 0.4398,
      "num_input_tokens_seen": 608864,
      "step": 940
    },
    {
      "epoch": 0.49528301886792453,
      "grad_norm": 1.445914387702942,
      "learning_rate": 1.2368972746331238e-05,
      "loss": 0.504,
      "num_input_tokens_seen": 612448,
      "step": 945
    },
    {
      "epoch": 0.4979035639412998,
      "grad_norm": 3.4456372261047363,
      "learning_rate": 1.243448637316562e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 614944,
      "step": 950
    },
    {
      "epoch": 0.500524109014675,
      "grad_norm": 1.4669665098190308,
      "learning_rate": 1.25e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 617472,
      "step": 955
    },
    {
      "epoch": 0.5031446540880503,
      "grad_norm": 1.7000058889389038,
      "learning_rate": 1.2565513626834382e-05,
      "loss": 0.431,
      "num_input_tokens_seen": 620192,
      "step": 960
    },
    {
      "epoch": 0.5057651991614256,
      "grad_norm": 0.8384459018707275,
      "learning_rate": 1.2631027253668765e-05,
      "loss": 0.5388,
      "num_input_tokens_seen": 623232,
      "step": 965
    },
    {
      "epoch": 0.5083857442348009,
      "grad_norm": 1.3518370389938354,
      "learning_rate": 1.2696540880503146e-05,
      "loss": 0.642,
      "num_input_tokens_seen": 626016,
      "step": 970
    },
    {
      "epoch": 0.5110062893081762,
      "grad_norm": 1.1484125852584839,
      "learning_rate": 1.2762054507337526e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 629472,
      "step": 975
    },
    {
      "epoch": 0.5136268343815513,
      "grad_norm": 1.2724742889404297,
      "learning_rate": 1.2827568134171907e-05,
      "loss": 0.5402,
      "num_input_tokens_seen": 632608,
      "step": 980
    },
    {
      "epoch": 0.5162473794549266,
      "grad_norm": 1.5012109279632568,
      "learning_rate": 1.289308176100629e-05,
      "loss": 0.393,
      "num_input_tokens_seen": 636192,
      "step": 985
    },
    {
      "epoch": 0.5188679245283019,
      "grad_norm": 0.9749370217323303,
      "learning_rate": 1.2958595387840672e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 640544,
      "step": 990
    },
    {
      "epoch": 0.5214884696016772,
      "grad_norm": 1.2814892530441284,
      "learning_rate": 1.3024109014675053e-05,
      "loss": 0.459,
      "num_input_tokens_seen": 644448,
      "step": 995
    },
    {
      "epoch": 0.5241090146750524,
      "grad_norm": 0.7873020172119141,
      "learning_rate": 1.3089622641509436e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 648288,
      "step": 1000
    },
    {
      "epoch": 0.5267295597484277,
      "grad_norm": 2.0086476802825928,
      "learning_rate": 1.3155136268343814e-05,
      "loss": 0.6344,
      "num_input_tokens_seen": 651296,
      "step": 1005
    },
    {
      "epoch": 0.5293501048218029,
      "grad_norm": 0.8954798579216003,
      "learning_rate": 1.3220649895178197e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 654176,
      "step": 1010
    },
    {
      "epoch": 0.5319706498951782,
      "grad_norm": 1.047168493270874,
      "learning_rate": 1.3286163522012579e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 657664,
      "step": 1015
    },
    {
      "epoch": 0.5345911949685535,
      "grad_norm": 1.256142020225525,
      "learning_rate": 1.3351677148846962e-05,
      "loss": 0.5941,
      "num_input_tokens_seen": 660064,
      "step": 1020
    },
    {
      "epoch": 0.5372117400419287,
      "grad_norm": 1.553135871887207,
      "learning_rate": 1.3417190775681343e-05,
      "loss": 0.5872,
      "num_input_tokens_seen": 662880,
      "step": 1025
    },
    {
      "epoch": 0.539832285115304,
      "grad_norm": 1.2017161846160889,
      "learning_rate": 1.3482704402515725e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 665920,
      "step": 1030
    },
    {
      "epoch": 0.5424528301886793,
      "grad_norm": 1.5468571186065674,
      "learning_rate": 1.3548218029350104e-05,
      "loss": 0.4367,
      "num_input_tokens_seen": 668864,
      "step": 1035
    },
    {
      "epoch": 0.5450733752620545,
      "grad_norm": 1.4029545783996582,
      "learning_rate": 1.3613731656184486e-05,
      "loss": 0.4442,
      "num_input_tokens_seen": 671776,
      "step": 1040
    },
    {
      "epoch": 0.5476939203354297,
      "grad_norm": 1.2014282941818237,
      "learning_rate": 1.3679245283018869e-05,
      "loss": 0.5206,
      "num_input_tokens_seen": 674912,
      "step": 1045
    },
    {
      "epoch": 0.550314465408805,
      "grad_norm": 1.1566749811172485,
      "learning_rate": 1.374475890985325e-05,
      "loss": 0.3837,
      "num_input_tokens_seen": 678368,
      "step": 1050
    },
    {
      "epoch": 0.5529350104821803,
      "grad_norm": 0.6789112091064453,
      "learning_rate": 1.3810272536687632e-05,
      "loss": 0.5459,
      "num_input_tokens_seen": 682336,
      "step": 1055
    },
    {
      "epoch": 0.5555555555555556,
      "grad_norm": 2.199918270111084,
      "learning_rate": 1.3875786163522015e-05,
      "loss": 0.4788,
      "num_input_tokens_seen": 685760,
      "step": 1060
    },
    {
      "epoch": 0.5581761006289309,
      "grad_norm": 1.847589373588562,
      "learning_rate": 1.3941299790356394e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 688480,
      "step": 1065
    },
    {
      "epoch": 0.560796645702306,
      "grad_norm": 1.0683168172836304,
      "learning_rate": 1.4006813417190776e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 692064,
      "step": 1070
    },
    {
      "epoch": 0.5634171907756813,
      "grad_norm": 2.041823625564575,
      "learning_rate": 1.4072327044025157e-05,
      "loss": 0.5657,
      "num_input_tokens_seen": 694464,
      "step": 1075
    },
    {
      "epoch": 0.5660377358490566,
      "grad_norm": 1.083713173866272,
      "learning_rate": 1.413784067085954e-05,
      "loss": 0.4907,
      "num_input_tokens_seen": 697344,
      "step": 1080
    },
    {
      "epoch": 0.5686582809224319,
      "grad_norm": 1.5443570613861084,
      "learning_rate": 1.4203354297693922e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 700256,
      "step": 1085
    },
    {
      "epoch": 0.5712788259958071,
      "grad_norm": 0.9995582103729248,
      "learning_rate": 1.4268867924528303e-05,
      "loss": 0.4493,
      "num_input_tokens_seen": 703968,
      "step": 1090
    },
    {
      "epoch": 0.5738993710691824,
      "grad_norm": 0.834894597530365,
      "learning_rate": 1.4334381551362683e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 708160,
      "step": 1095
    },
    {
      "epoch": 0.5765199161425576,
      "grad_norm": 1.2520649433135986,
      "learning_rate": 1.4399895178197064e-05,
      "loss": 0.6817,
      "num_input_tokens_seen": 710976,
      "step": 1100
    },
    {
      "epoch": 0.5791404612159329,
      "grad_norm": 3.037137746810913,
      "learning_rate": 1.4465408805031447e-05,
      "loss": 0.5854,
      "num_input_tokens_seen": 713696,
      "step": 1105
    },
    {
      "epoch": 0.5817610062893082,
      "grad_norm": 1.80252206325531,
      "learning_rate": 1.4530922431865828e-05,
      "loss": 0.4378,
      "num_input_tokens_seen": 716608,
      "step": 1110
    },
    {
      "epoch": 0.5843815513626834,
      "grad_norm": 1.4816187620162964,
      "learning_rate": 1.4596436058700211e-05,
      "loss": 0.4786,
      "num_input_tokens_seen": 719488,
      "step": 1115
    },
    {
      "epoch": 0.5870020964360587,
      "grad_norm": 0.9957678914070129,
      "learning_rate": 1.4661949685534593e-05,
      "loss": 0.384,
      "num_input_tokens_seen": 722304,
      "step": 1120
    },
    {
      "epoch": 0.589622641509434,
      "grad_norm": 1.307634711265564,
      "learning_rate": 1.4727463312368974e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 725472,
      "step": 1125
    },
    {
      "epoch": 0.5922431865828093,
      "grad_norm": 1.1849116086959839,
      "learning_rate": 1.4792976939203354e-05,
      "loss": 0.3555,
      "num_input_tokens_seen": 728896,
      "step": 1130
    },
    {
      "epoch": 0.5948637316561844,
      "grad_norm": 0.9431798458099365,
      "learning_rate": 1.4858490566037735e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 732000,
      "step": 1135
    },
    {
      "epoch": 0.5974842767295597,
      "grad_norm": 1.2277644872665405,
      "learning_rate": 1.4924004192872118e-05,
      "loss": 0.5477,
      "num_input_tokens_seen": 735328,
      "step": 1140
    },
    {
      "epoch": 0.600104821802935,
      "grad_norm": 0.9730405807495117,
      "learning_rate": 1.49895178197065e-05,
      "loss": 0.3414,
      "num_input_tokens_seen": 738528,
      "step": 1145
    },
    {
      "epoch": 0.6027253668763103,
      "grad_norm": 1.2166637182235718,
      "learning_rate": 1.5055031446540883e-05,
      "loss": 0.5687,
      "num_input_tokens_seen": 741376,
      "step": 1150
    },
    {
      "epoch": 0.6053459119496856,
      "grad_norm": 1.1247636079788208,
      "learning_rate": 1.5120545073375264e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 743872,
      "step": 1155
    },
    {
      "epoch": 0.6079664570230608,
      "grad_norm": 1.1070125102996826,
      "learning_rate": 1.5186058700209644e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 746720,
      "step": 1160
    },
    {
      "epoch": 0.610587002096436,
      "grad_norm": 1.1410273313522339,
      "learning_rate": 1.5251572327044025e-05,
      "loss": 0.5449,
      "num_input_tokens_seen": 750240,
      "step": 1165
    },
    {
      "epoch": 0.6132075471698113,
      "grad_norm": 1.3340420722961426,
      "learning_rate": 1.5317085953878407e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 753344,
      "step": 1170
    },
    {
      "epoch": 0.6158280922431866,
      "grad_norm": 0.881894588470459,
      "learning_rate": 1.538259958071279e-05,
      "loss": 0.4529,
      "num_input_tokens_seen": 755968,
      "step": 1175
    },
    {
      "epoch": 0.6184486373165619,
      "grad_norm": 0.851213812828064,
      "learning_rate": 1.5448113207547173e-05,
      "loss": 0.5571,
      "num_input_tokens_seen": 759520,
      "step": 1180
    },
    {
      "epoch": 0.6210691823899371,
      "grad_norm": 1.2282451391220093,
      "learning_rate": 1.5513626834381552e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 762048,
      "step": 1185
    },
    {
      "epoch": 0.6236897274633124,
      "grad_norm": 1.4346284866333008,
      "learning_rate": 1.5579140461215932e-05,
      "loss": 0.5803,
      "num_input_tokens_seen": 765504,
      "step": 1190
    },
    {
      "epoch": 0.6263102725366876,
      "grad_norm": 1.0165141820907593,
      "learning_rate": 1.5644654088050315e-05,
      "loss": 0.4512,
      "num_input_tokens_seen": 768192,
      "step": 1195
    },
    {
      "epoch": 0.6289308176100629,
      "grad_norm": 1.0896462202072144,
      "learning_rate": 1.5710167714884695e-05,
      "loss": 0.4972,
      "num_input_tokens_seen": 771168,
      "step": 1200
    },
    {
      "epoch": 0.6315513626834381,
      "grad_norm": 0.9294119477272034,
      "learning_rate": 1.5775681341719078e-05,
      "loss": 0.2978,
      "num_input_tokens_seen": 777504,
      "step": 1205
    },
    {
      "epoch": 0.6341719077568134,
      "grad_norm": 1.5067579746246338,
      "learning_rate": 1.584119496855346e-05,
      "loss": 0.4441,
      "num_input_tokens_seen": 780512,
      "step": 1210
    },
    {
      "epoch": 0.6367924528301887,
      "grad_norm": 1.5305171012878418,
      "learning_rate": 1.5906708595387844e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 784192,
      "step": 1215
    },
    {
      "epoch": 0.639412997903564,
      "grad_norm": 1.3792356252670288,
      "learning_rate": 1.597222222222222e-05,
      "loss": 0.5606,
      "num_input_tokens_seen": 787552,
      "step": 1220
    },
    {
      "epoch": 0.6420335429769392,
      "grad_norm": 0.8774745464324951,
      "learning_rate": 1.6037735849056604e-05,
      "loss": 0.6115,
      "num_input_tokens_seen": 790720,
      "step": 1225
    },
    {
      "epoch": 0.6446540880503144,
      "grad_norm": 1.6989121437072754,
      "learning_rate": 1.6103249475890987e-05,
      "loss": 0.5549,
      "num_input_tokens_seen": 793696,
      "step": 1230
    },
    {
      "epoch": 0.6472746331236897,
      "grad_norm": 1.8355845212936401,
      "learning_rate": 1.6168763102725366e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 796640,
      "step": 1235
    },
    {
      "epoch": 0.649895178197065,
      "grad_norm": 2.067753553390503,
      "learning_rate": 1.623427672955975e-05,
      "loss": 0.5081,
      "num_input_tokens_seen": 799648,
      "step": 1240
    },
    {
      "epoch": 0.6525157232704403,
      "grad_norm": 1.9289723634719849,
      "learning_rate": 1.6299790356394132e-05,
      "loss": 0.3392,
      "num_input_tokens_seen": 802496,
      "step": 1245
    },
    {
      "epoch": 0.6551362683438156,
      "grad_norm": 1.200361728668213,
      "learning_rate": 1.6365303983228512e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 805760,
      "step": 1250
    },
    {
      "epoch": 0.6577568134171907,
      "grad_norm": 2.6751766204833984,
      "learning_rate": 1.6430817610062892e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 808832,
      "step": 1255
    },
    {
      "epoch": 0.660377358490566,
      "grad_norm": 1.3863588571548462,
      "learning_rate": 1.6496331236897275e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 812256,
      "step": 1260
    },
    {
      "epoch": 0.6629979035639413,
      "grad_norm": 1.679952621459961,
      "learning_rate": 1.6561844863731658e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 815680,
      "step": 1265
    },
    {
      "epoch": 0.6656184486373166,
      "grad_norm": 0.8282990455627441,
      "learning_rate": 1.6627358490566038e-05,
      "loss": 0.3576,
      "num_input_tokens_seen": 818400,
      "step": 1270
    },
    {
      "epoch": 0.6682389937106918,
      "grad_norm": 1.242930293083191,
      "learning_rate": 1.669287211740042e-05,
      "loss": 0.5443,
      "num_input_tokens_seen": 821536,
      "step": 1275
    },
    {
      "epoch": 0.6708595387840671,
      "grad_norm": 1.6194554567337036,
      "learning_rate": 1.6758385744234804e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 825024,
      "step": 1280
    },
    {
      "epoch": 0.6734800838574424,
      "grad_norm": 0.900963544845581,
      "learning_rate": 1.6823899371069183e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 829248,
      "step": 1285
    },
    {
      "epoch": 0.6761006289308176,
      "grad_norm": 1.1463847160339355,
      "learning_rate": 1.6889412997903563e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 831872,
      "step": 1290
    },
    {
      "epoch": 0.6787211740041929,
      "grad_norm": 1.4006637334823608,
      "learning_rate": 1.6954926624737946e-05,
      "loss": 0.5771,
      "num_input_tokens_seen": 834880,
      "step": 1295
    },
    {
      "epoch": 0.6813417190775681,
      "grad_norm": 0.9309133291244507,
      "learning_rate": 1.702044025157233e-05,
      "loss": 0.5613,
      "num_input_tokens_seen": 837824,
      "step": 1300
    },
    {
      "epoch": 0.6839622641509434,
      "grad_norm": 1.1003601551055908,
      "learning_rate": 1.708595387840671e-05,
      "loss": 0.498,
      "num_input_tokens_seen": 840896,
      "step": 1305
    },
    {
      "epoch": 0.6865828092243187,
      "grad_norm": 1.1695489883422852,
      "learning_rate": 1.7151467505241092e-05,
      "loss": 0.6692,
      "num_input_tokens_seen": 844000,
      "step": 1310
    },
    {
      "epoch": 0.689203354297694,
      "grad_norm": 0.6874983906745911,
      "learning_rate": 1.7216981132075472e-05,
      "loss": 0.4513,
      "num_input_tokens_seen": 847840,
      "step": 1315
    },
    {
      "epoch": 0.6918238993710691,
      "grad_norm": 1.0842262506484985,
      "learning_rate": 1.7282494758909855e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 850976,
      "step": 1320
    },
    {
      "epoch": 0.6944444444444444,
      "grad_norm": 0.8498819470405579,
      "learning_rate": 1.7348008385744234e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 855616,
      "step": 1325
    },
    {
      "epoch": 0.6970649895178197,
      "grad_norm": 0.992221474647522,
      "learning_rate": 1.7413522012578618e-05,
      "loss": 0.3071,
      "num_input_tokens_seen": 858528,
      "step": 1330
    },
    {
      "epoch": 0.699685534591195,
      "grad_norm": 1.449661374092102,
      "learning_rate": 1.7479035639413e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 862144,
      "step": 1335
    },
    {
      "epoch": 0.7023060796645703,
      "grad_norm": 1.1315070390701294,
      "learning_rate": 1.754454926624738e-05,
      "loss": 0.5737,
      "num_input_tokens_seen": 865440,
      "step": 1340
    },
    {
      "epoch": 0.7049266247379455,
      "grad_norm": 1.068007230758667,
      "learning_rate": 1.761006289308176e-05,
      "loss": 0.5072,
      "num_input_tokens_seen": 868992,
      "step": 1345
    },
    {
      "epoch": 0.7075471698113207,
      "grad_norm": 1.609686255455017,
      "learning_rate": 1.7675576519916143e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 871968,
      "step": 1350
    },
    {
      "epoch": 0.710167714884696,
      "grad_norm": 1.4275932312011719,
      "learning_rate": 1.7741090146750526e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 874912,
      "step": 1355
    },
    {
      "epoch": 0.7127882599580713,
      "grad_norm": 1.2501240968704224,
      "learning_rate": 1.7806603773584906e-05,
      "loss": 0.5088,
      "num_input_tokens_seen": 878112,
      "step": 1360
    },
    {
      "epoch": 0.7154088050314465,
      "grad_norm": 1.7398126125335693,
      "learning_rate": 1.787211740041929e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 880928,
      "step": 1365
    },
    {
      "epoch": 0.7180293501048218,
      "grad_norm": 0.9613440632820129,
      "learning_rate": 1.7937631027253672e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 884832,
      "step": 1370
    },
    {
      "epoch": 0.7206498951781971,
      "grad_norm": 0.7573508024215698,
      "learning_rate": 1.8003144654088048e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 888800,
      "step": 1375
    },
    {
      "epoch": 0.7232704402515723,
      "grad_norm": 0.957529604434967,
      "learning_rate": 1.806865828092243e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 892416,
      "step": 1380
    },
    {
      "epoch": 0.7258909853249476,
      "grad_norm": 1.317569375038147,
      "learning_rate": 1.8134171907756814e-05,
      "loss": 0.5645,
      "num_input_tokens_seen": 895840,
      "step": 1385
    },
    {
      "epoch": 0.7285115303983228,
      "grad_norm": 1.1882200241088867,
      "learning_rate": 1.8199685534591197e-05,
      "loss": 0.5095,
      "num_input_tokens_seen": 899392,
      "step": 1390
    },
    {
      "epoch": 0.7311320754716981,
      "grad_norm": 1.183388113975525,
      "learning_rate": 1.8265199161425577e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 903264,
      "step": 1395
    },
    {
      "epoch": 0.7337526205450734,
      "grad_norm": 1.2069904804229736,
      "learning_rate": 1.833071278825996e-05,
      "loss": 0.4499,
      "num_input_tokens_seen": 906016,
      "step": 1400
    },
    {
      "epoch": 0.7363731656184487,
      "grad_norm": 1.5491236448287964,
      "learning_rate": 1.839622641509434e-05,
      "loss": 0.5893,
      "num_input_tokens_seen": 909152,
      "step": 1405
    },
    {
      "epoch": 0.7389937106918238,
      "grad_norm": 1.5980843305587769,
      "learning_rate": 1.846174004192872e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 912416,
      "step": 1410
    },
    {
      "epoch": 0.7416142557651991,
      "grad_norm": 1.4913450479507446,
      "learning_rate": 1.8527253668763103e-05,
      "loss": 0.6131,
      "num_input_tokens_seen": 915616,
      "step": 1415
    },
    {
      "epoch": 0.7442348008385744,
      "grad_norm": 2.3381500244140625,
      "learning_rate": 1.8592767295597486e-05,
      "loss": 0.5128,
      "num_input_tokens_seen": 921568,
      "step": 1420
    },
    {
      "epoch": 0.7468553459119497,
      "grad_norm": 0.9894816279411316,
      "learning_rate": 1.865828092243187e-05,
      "loss": 0.4479,
      "num_input_tokens_seen": 924864,
      "step": 1425
    },
    {
      "epoch": 0.749475890985325,
      "grad_norm": 1.538128137588501,
      "learning_rate": 1.872379454926625e-05,
      "loss": 0.5181,
      "num_input_tokens_seen": 927968,
      "step": 1430
    },
    {
      "epoch": 0.7520964360587002,
      "grad_norm": 1.3000891208648682,
      "learning_rate": 1.878930817610063e-05,
      "loss": 0.5011,
      "num_input_tokens_seen": 931584,
      "step": 1435
    },
    {
      "epoch": 0.7547169811320755,
      "grad_norm": 1.8249589204788208,
      "learning_rate": 1.885482180293501e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 934816,
      "step": 1440
    },
    {
      "epoch": 0.7573375262054507,
      "grad_norm": 1.508600115776062,
      "learning_rate": 1.892033542976939e-05,
      "loss": 0.5268,
      "num_input_tokens_seen": 937856,
      "step": 1445
    },
    {
      "epoch": 0.759958071278826,
      "grad_norm": 1.3927509784698486,
      "learning_rate": 1.8985849056603774e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 945632,
      "step": 1450
    },
    {
      "epoch": 0.7625786163522013,
      "grad_norm": 1.277258276939392,
      "learning_rate": 1.9051362683438157e-05,
      "loss": 0.5428,
      "num_input_tokens_seen": 948608,
      "step": 1455
    },
    {
      "epoch": 0.7651991614255765,
      "grad_norm": 1.4495875835418701,
      "learning_rate": 1.9116876310272537e-05,
      "loss": 0.4075,
      "num_input_tokens_seen": 951488,
      "step": 1460
    },
    {
      "epoch": 0.7678197064989518,
      "grad_norm": 1.3176020383834839,
      "learning_rate": 1.918238993710692e-05,
      "loss": 0.4967,
      "num_input_tokens_seen": 955072,
      "step": 1465
    },
    {
      "epoch": 0.7704402515723271,
      "grad_norm": 1.492506980895996,
      "learning_rate": 1.92479035639413e-05,
      "loss": 0.5655,
      "num_input_tokens_seen": 958272,
      "step": 1470
    },
    {
      "epoch": 0.7730607966457023,
      "grad_norm": 0.7839031219482422,
      "learning_rate": 1.9313417190775683e-05,
      "loss": 0.6418,
      "num_input_tokens_seen": 961888,
      "step": 1475
    },
    {
      "epoch": 0.7756813417190775,
      "grad_norm": 2.726471424102783,
      "learning_rate": 1.9378930817610062e-05,
      "loss": 0.6758,
      "num_input_tokens_seen": 964256,
      "step": 1480
    },
    {
      "epoch": 0.7783018867924528,
      "grad_norm": 1.6470816135406494,
      "learning_rate": 1.9444444444444445e-05,
      "loss": 0.5108,
      "num_input_tokens_seen": 968224,
      "step": 1485
    },
    {
      "epoch": 0.7809224318658281,
      "grad_norm": 0.8348334431648254,
      "learning_rate": 1.950995807127883e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 971136,
      "step": 1490
    },
    {
      "epoch": 0.7835429769392034,
      "grad_norm": 1.4424186944961548,
      "learning_rate": 1.9575471698113208e-05,
      "loss": 0.4917,
      "num_input_tokens_seen": 975936,
      "step": 1495
    },
    {
      "epoch": 0.7861635220125787,
      "grad_norm": 1.6124675273895264,
      "learning_rate": 1.9640985324947588e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 979360,
      "step": 1500
    },
    {
      "epoch": 0.7887840670859538,
      "grad_norm": 0.6098265051841736,
      "learning_rate": 1.970649895178197e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 983616,
      "step": 1505
    },
    {
      "epoch": 0.7914046121593291,
      "grad_norm": 1.408664345741272,
      "learning_rate": 1.9772012578616354e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 986944,
      "step": 1510
    },
    {
      "epoch": 0.7940251572327044,
      "grad_norm": 1.2067890167236328,
      "learning_rate": 1.9837526205450734e-05,
      "loss": 0.4822,
      "num_input_tokens_seen": 989888,
      "step": 1515
    },
    {
      "epoch": 0.7966457023060797,
      "grad_norm": 1.420892596244812,
      "learning_rate": 1.9903039832285117e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 992768,
      "step": 1520
    },
    {
      "epoch": 0.799266247379455,
      "grad_norm": 1.150388240814209,
      "learning_rate": 1.99685534591195e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 996896,
      "step": 1525
    },
    {
      "epoch": 0.8018867924528302,
      "grad_norm": 1.0258524417877197,
      "learning_rate": 2.003406708595388e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 999840,
      "step": 1530
    },
    {
      "epoch": 0.8045073375262054,
      "grad_norm": 0.9107629656791687,
      "learning_rate": 2.009958071278826e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 1003680,
      "step": 1535
    },
    {
      "epoch": 0.8071278825995807,
      "grad_norm": 1.3751723766326904,
      "learning_rate": 2.0165094339622642e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 1006656,
      "step": 1540
    },
    {
      "epoch": 0.809748427672956,
      "grad_norm": 2.5770654678344727,
      "learning_rate": 2.0230607966457025e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 1008768,
      "step": 1545
    },
    {
      "epoch": 0.8123689727463312,
      "grad_norm": 0.9366227984428406,
      "learning_rate": 2.0296121593291405e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 1011968,
      "step": 1550
    },
    {
      "epoch": 0.8149895178197065,
      "grad_norm": 1.1994231939315796,
      "learning_rate": 2.0361635220125788e-05,
      "loss": 0.4702,
      "num_input_tokens_seen": 1018784,
      "step": 1555
    },
    {
      "epoch": 0.8176100628930818,
      "grad_norm": 0.9687456488609314,
      "learning_rate": 2.0427148846960168e-05,
      "loss": 0.3471,
      "num_input_tokens_seen": 1023520,
      "step": 1560
    },
    {
      "epoch": 0.820230607966457,
      "grad_norm": 1.1904103755950928,
      "learning_rate": 2.049266247379455e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 1027200,
      "step": 1565
    },
    {
      "epoch": 0.8228511530398323,
      "grad_norm": 0.6912742853164673,
      "learning_rate": 2.055817610062893e-05,
      "loss": 0.7732,
      "num_input_tokens_seen": 1030848,
      "step": 1570
    },
    {
      "epoch": 0.8254716981132075,
      "grad_norm": 1.1211975812911987,
      "learning_rate": 2.0623689727463314e-05,
      "loss": 0.5356,
      "num_input_tokens_seen": 1035232,
      "step": 1575
    },
    {
      "epoch": 0.8280922431865828,
      "grad_norm": 1.6517441272735596,
      "learning_rate": 2.0689203354297697e-05,
      "loss": 0.4918,
      "num_input_tokens_seen": 1038176,
      "step": 1580
    },
    {
      "epoch": 0.8307127882599581,
      "grad_norm": 2.6102170944213867,
      "learning_rate": 2.0754716981132076e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 1041216,
      "step": 1585
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 1.2312190532684326,
      "learning_rate": 2.0820230607966456e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 1043840,
      "step": 1590
    },
    {
      "epoch": 0.8359538784067087,
      "grad_norm": 1.0607041120529175,
      "learning_rate": 2.088574423480084e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 1046656,
      "step": 1595
    },
    {
      "epoch": 0.8385744234800838,
      "grad_norm": 1.1346946954727173,
      "learning_rate": 2.0951257861635222e-05,
      "loss": 0.4215,
      "num_input_tokens_seen": 1051616,
      "step": 1600
    },
    {
      "epoch": 0.8411949685534591,
      "grad_norm": 1.5963236093521118,
      "learning_rate": 2.1016771488469602e-05,
      "loss": 0.5972,
      "num_input_tokens_seen": 1053824,
      "step": 1605
    },
    {
      "epoch": 0.8438155136268344,
      "grad_norm": 1.3434704542160034,
      "learning_rate": 2.1082285115303985e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 1056512,
      "step": 1610
    },
    {
      "epoch": 0.8464360587002097,
      "grad_norm": 0.9758290648460388,
      "learning_rate": 2.1147798742138368e-05,
      "loss": 0.4772,
      "num_input_tokens_seen": 1060160,
      "step": 1615
    },
    {
      "epoch": 0.8490566037735849,
      "grad_norm": 1.564477801322937,
      "learning_rate": 2.1213312368972748e-05,
      "loss": 0.6068,
      "num_input_tokens_seen": 1062880,
      "step": 1620
    },
    {
      "epoch": 0.8516771488469602,
      "grad_norm": 1.647552728652954,
      "learning_rate": 2.1278825995807127e-05,
      "loss": 0.5661,
      "num_input_tokens_seen": 1066528,
      "step": 1625
    },
    {
      "epoch": 0.8542976939203354,
      "grad_norm": 1.1753463745117188,
      "learning_rate": 2.134433962264151e-05,
      "loss": 0.3973,
      "num_input_tokens_seen": 1069472,
      "step": 1630
    },
    {
      "epoch": 0.8569182389937107,
      "grad_norm": 1.3032065629959106,
      "learning_rate": 2.1409853249475894e-05,
      "loss": 0.4902,
      "num_input_tokens_seen": 1074304,
      "step": 1635
    },
    {
      "epoch": 0.859538784067086,
      "grad_norm": 1.8093535900115967,
      "learning_rate": 2.1475366876310273e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 1077376,
      "step": 1640
    },
    {
      "epoch": 0.8621593291404612,
      "grad_norm": 0.9956007599830627,
      "learning_rate": 2.1540880503144656e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 1080672,
      "step": 1645
    },
    {
      "epoch": 0.8647798742138365,
      "grad_norm": 1.1790646314620972,
      "learning_rate": 2.160639412997904e-05,
      "loss": 0.5266,
      "num_input_tokens_seen": 1084224,
      "step": 1650
    },
    {
      "epoch": 0.8674004192872118,
      "grad_norm": 1.407495379447937,
      "learning_rate": 2.1671907756813416e-05,
      "loss": 0.6544,
      "num_input_tokens_seen": 1086912,
      "step": 1655
    },
    {
      "epoch": 0.870020964360587,
      "grad_norm": 1.578322172164917,
      "learning_rate": 2.17374213836478e-05,
      "loss": 0.581,
      "num_input_tokens_seen": 1089696,
      "step": 1660
    },
    {
      "epoch": 0.8726415094339622,
      "grad_norm": 1.055503249168396,
      "learning_rate": 2.1802935010482182e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 1092608,
      "step": 1665
    },
    {
      "epoch": 0.8752620545073375,
      "grad_norm": 1.1203181743621826,
      "learning_rate": 2.186844863731656e-05,
      "loss": 0.6558,
      "num_input_tokens_seen": 1095808,
      "step": 1670
    },
    {
      "epoch": 0.8778825995807128,
      "grad_norm": 0.9507994055747986,
      "learning_rate": 2.1933962264150945e-05,
      "loss": 0.5447,
      "num_input_tokens_seen": 1098816,
      "step": 1675
    },
    {
      "epoch": 0.8805031446540881,
      "grad_norm": 0.8525803685188293,
      "learning_rate": 2.1999475890985328e-05,
      "loss": 0.5563,
      "num_input_tokens_seen": 1103136,
      "step": 1680
    },
    {
      "epoch": 0.8831236897274634,
      "grad_norm": 0.992513120174408,
      "learning_rate": 2.2064989517819707e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 1107872,
      "step": 1685
    },
    {
      "epoch": 0.8857442348008385,
      "grad_norm": 1.2677968740463257,
      "learning_rate": 2.2130503144654087e-05,
      "loss": 0.6202,
      "num_input_tokens_seen": 1111072,
      "step": 1690
    },
    {
      "epoch": 0.8883647798742138,
      "grad_norm": 1.6781073808670044,
      "learning_rate": 2.219601677148847e-05,
      "loss": 0.5386,
      "num_input_tokens_seen": 1113792,
      "step": 1695
    },
    {
      "epoch": 0.8909853249475891,
      "grad_norm": 1.7389599084854126,
      "learning_rate": 2.2261530398322853e-05,
      "loss": 0.4305,
      "num_input_tokens_seen": 1116384,
      "step": 1700
    },
    {
      "epoch": 0.8936058700209644,
      "grad_norm": 0.9067497253417969,
      "learning_rate": 2.2327044025157233e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 1119296,
      "step": 1705
    },
    {
      "epoch": 0.8962264150943396,
      "grad_norm": 0.9933650493621826,
      "learning_rate": 2.2392557651991616e-05,
      "loss": 0.5935,
      "num_input_tokens_seen": 1122720,
      "step": 1710
    },
    {
      "epoch": 0.8988469601677149,
      "grad_norm": 3.7508480548858643,
      "learning_rate": 2.2458071278825996e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 1125088,
      "step": 1715
    },
    {
      "epoch": 0.9014675052410901,
      "grad_norm": 1.3584188222885132,
      "learning_rate": 2.252358490566038e-05,
      "loss": 0.6149,
      "num_input_tokens_seen": 1128288,
      "step": 1720
    },
    {
      "epoch": 0.9040880503144654,
      "grad_norm": 2.1770894527435303,
      "learning_rate": 2.258909853249476e-05,
      "loss": 0.5371,
      "num_input_tokens_seen": 1131040,
      "step": 1725
    },
    {
      "epoch": 0.9067085953878407,
      "grad_norm": 0.8626812100410461,
      "learning_rate": 2.265461215932914e-05,
      "loss": 0.5138,
      "num_input_tokens_seen": 1135136,
      "step": 1730
    },
    {
      "epoch": 0.9093291404612159,
      "grad_norm": 1.0651603937149048,
      "learning_rate": 2.2720125786163524e-05,
      "loss": 0.6284,
      "num_input_tokens_seen": 1139008,
      "step": 1735
    },
    {
      "epoch": 0.9119496855345912,
      "grad_norm": 1.0495299100875854,
      "learning_rate": 2.2785639412997904e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 1141920,
      "step": 1740
    },
    {
      "epoch": 0.9145702306079665,
      "grad_norm": 1.347430944442749,
      "learning_rate": 2.2851153039832284e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 1144960,
      "step": 1745
    },
    {
      "epoch": 0.9171907756813418,
      "grad_norm": 1.3524119853973389,
      "learning_rate": 2.2916666666666667e-05,
      "loss": 0.5337,
      "num_input_tokens_seen": 1147712,
      "step": 1750
    },
    {
      "epoch": 0.9198113207547169,
      "grad_norm": 1.2409439086914062,
      "learning_rate": 2.298218029350105e-05,
      "loss": 0.5192,
      "num_input_tokens_seen": 1150528,
      "step": 1755
    },
    {
      "epoch": 0.9224318658280922,
      "grad_norm": 1.6672214269638062,
      "learning_rate": 2.304769392033543e-05,
      "loss": 0.5521,
      "num_input_tokens_seen": 1154272,
      "step": 1760
    },
    {
      "epoch": 0.9250524109014675,
      "grad_norm": 1.9797773361206055,
      "learning_rate": 2.3113207547169813e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 1157088,
      "step": 1765
    },
    {
      "epoch": 0.9276729559748428,
      "grad_norm": 1.1141971349716187,
      "learning_rate": 2.3178721174004196e-05,
      "loss": 0.6102,
      "num_input_tokens_seen": 1160192,
      "step": 1770
    },
    {
      "epoch": 0.9302935010482181,
      "grad_norm": 1.2566362619400024,
      "learning_rate": 2.3244234800838576e-05,
      "loss": 0.5305,
      "num_input_tokens_seen": 1163424,
      "step": 1775
    },
    {
      "epoch": 0.9329140461215933,
      "grad_norm": 0.9857488870620728,
      "learning_rate": 2.3309748427672955e-05,
      "loss": 0.5788,
      "num_input_tokens_seen": 1167104,
      "step": 1780
    },
    {
      "epoch": 0.9355345911949685,
      "grad_norm": 1.5482431650161743,
      "learning_rate": 2.3375262054507338e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 1170368,
      "step": 1785
    },
    {
      "epoch": 0.9381551362683438,
      "grad_norm": 0.9865766167640686,
      "learning_rate": 2.344077568134172e-05,
      "loss": 0.5968,
      "num_input_tokens_seen": 1173888,
      "step": 1790
    },
    {
      "epoch": 0.9407756813417191,
      "grad_norm": 0.8442274928092957,
      "learning_rate": 2.35062893081761e-05,
      "loss": 0.4947,
      "num_input_tokens_seen": 1177408,
      "step": 1795
    },
    {
      "epoch": 0.9433962264150944,
      "grad_norm": 1.5175985097885132,
      "learning_rate": 2.3571802935010484e-05,
      "loss": 0.4773,
      "num_input_tokens_seen": 1181056,
      "step": 1800
    },
    {
      "epoch": 0.9460167714884696,
      "grad_norm": 1.3901777267456055,
      "learning_rate": 2.3637316561844867e-05,
      "loss": 0.6613,
      "num_input_tokens_seen": 1184256,
      "step": 1805
    },
    {
      "epoch": 0.9486373165618449,
      "grad_norm": 1.1978044509887695,
      "learning_rate": 2.3702830188679247e-05,
      "loss": 0.42,
      "num_input_tokens_seen": 1187168,
      "step": 1810
    },
    {
      "epoch": 0.9512578616352201,
      "grad_norm": 0.8089858889579773,
      "learning_rate": 2.3768343815513627e-05,
      "loss": 0.4466,
      "num_input_tokens_seen": 1190784,
      "step": 1815
    },
    {
      "epoch": 0.9538784067085954,
      "grad_norm": 1.3247634172439575,
      "learning_rate": 2.383385744234801e-05,
      "loss": 0.5457,
      "num_input_tokens_seen": 1193312,
      "step": 1820
    },
    {
      "epoch": 0.9564989517819706,
      "grad_norm": 1.0153871774673462,
      "learning_rate": 2.3899371069182393e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 1196608,
      "step": 1825
    },
    {
      "epoch": 0.9591194968553459,
      "grad_norm": 2.0007128715515137,
      "learning_rate": 2.3964884696016772e-05,
      "loss": 0.5531,
      "num_input_tokens_seen": 1200160,
      "step": 1830
    },
    {
      "epoch": 0.9617400419287212,
      "grad_norm": 2.3584542274475098,
      "learning_rate": 2.4030398322851155e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 1203040,
      "step": 1835
    },
    {
      "epoch": 0.9643605870020965,
      "grad_norm": 1.00979483127594,
      "learning_rate": 2.4095911949685535e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 1205408,
      "step": 1840
    },
    {
      "epoch": 0.9669811320754716,
      "grad_norm": 4.916698455810547,
      "learning_rate": 2.4161425576519918e-05,
      "loss": 0.3536,
      "num_input_tokens_seen": 1208896,
      "step": 1845
    },
    {
      "epoch": 0.9696016771488469,
      "grad_norm": 1.1193126440048218,
      "learning_rate": 2.4226939203354298e-05,
      "loss": 0.4694,
      "num_input_tokens_seen": 1212576,
      "step": 1850
    },
    {
      "epoch": 0.9722222222222222,
      "grad_norm": 0.8890129923820496,
      "learning_rate": 2.429245283018868e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 1215872,
      "step": 1855
    },
    {
      "epoch": 0.9748427672955975,
      "grad_norm": 1.8729305267333984,
      "learning_rate": 2.4357966457023064e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 1219968,
      "step": 1860
    },
    {
      "epoch": 0.9774633123689728,
      "grad_norm": 0.997357964515686,
      "learning_rate": 2.4423480083857444e-05,
      "loss": 0.4885,
      "num_input_tokens_seen": 1222048,
      "step": 1865
    },
    {
      "epoch": 0.980083857442348,
      "grad_norm": 1.5450555086135864,
      "learning_rate": 2.4488993710691823e-05,
      "loss": 0.5155,
      "num_input_tokens_seen": 1225120,
      "step": 1870
    },
    {
      "epoch": 0.9827044025157232,
      "grad_norm": 1.0446343421936035,
      "learning_rate": 2.4554507337526206e-05,
      "loss": 0.5146,
      "num_input_tokens_seen": 1229056,
      "step": 1875
    },
    {
      "epoch": 0.9853249475890985,
      "grad_norm": 1.1415332555770874,
      "learning_rate": 2.4620020964360586e-05,
      "loss": 0.3727,
      "num_input_tokens_seen": 1231776,
      "step": 1880
    },
    {
      "epoch": 0.9879454926624738,
      "grad_norm": 0.8646486401557922,
      "learning_rate": 2.468553459119497e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 1235040,
      "step": 1885
    },
    {
      "epoch": 0.9905660377358491,
      "grad_norm": 1.6135566234588623,
      "learning_rate": 2.4751048218029352e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 1238048,
      "step": 1890
    },
    {
      "epoch": 0.9931865828092243,
      "grad_norm": 1.089516043663025,
      "learning_rate": 2.4816561844863735e-05,
      "loss": 0.5253,
      "num_input_tokens_seen": 1241344,
      "step": 1895
    },
    {
      "epoch": 0.9958071278825996,
      "grad_norm": 1.4433825016021729,
      "learning_rate": 2.488207547169811e-05,
      "loss": 0.3521,
      "num_input_tokens_seen": 1243840,
      "step": 1900
    },
    {
      "epoch": 0.9984276729559748,
      "grad_norm": 1.029323935508728,
      "learning_rate": 2.4947589098532495e-05,
      "loss": 0.5856,
      "num_input_tokens_seen": 1246816,
      "step": 1905
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.4838300049304962,
      "eval_runtime": 11.741,
      "eval_samples_per_second": 72.226,
      "eval_steps_per_second": 18.056,
      "num_input_tokens_seen": 1248304,
      "step": 1908
    },
    {
      "epoch": 1.00104821802935,
      "grad_norm": 1.5639374256134033,
      "learning_rate": 2.5013102725366878e-05,
      "loss": 0.4641,
      "num_input_tokens_seen": 1249200,
      "step": 1910
    },
    {
      "epoch": 1.0036687631027255,
      "grad_norm": 1.5963870286941528,
      "learning_rate": 2.507861635220126e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 1251664,
      "step": 1915
    },
    {
      "epoch": 1.0062893081761006,
      "grad_norm": 0.6603704690933228,
      "learning_rate": 2.514412997903564e-05,
      "loss": 0.4514,
      "num_input_tokens_seen": 1254928,
      "step": 1920
    },
    {
      "epoch": 1.0089098532494758,
      "grad_norm": 2.112800359725952,
      "learning_rate": 2.520964360587002e-05,
      "loss": 0.3823,
      "num_input_tokens_seen": 1257744,
      "step": 1925
    },
    {
      "epoch": 1.0115303983228512,
      "grad_norm": 1.5294440984725952,
      "learning_rate": 2.5275157232704407e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 1261904,
      "step": 1930
    },
    {
      "epoch": 1.0141509433962264,
      "grad_norm": 0.8666194081306458,
      "learning_rate": 2.5340670859538783e-05,
      "loss": 0.3906,
      "num_input_tokens_seen": 1264432,
      "step": 1935
    },
    {
      "epoch": 1.0167714884696017,
      "grad_norm": 1.249788761138916,
      "learning_rate": 2.540618448637317e-05,
      "loss": 0.5223,
      "num_input_tokens_seen": 1268400,
      "step": 1940
    },
    {
      "epoch": 1.019392033542977,
      "grad_norm": 1.1010940074920654,
      "learning_rate": 2.547169811320755e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 1272752,
      "step": 1945
    },
    {
      "epoch": 1.0220125786163523,
      "grad_norm": 1.297809362411499,
      "learning_rate": 2.553721174004193e-05,
      "loss": 0.5506,
      "num_input_tokens_seen": 1275568,
      "step": 1950
    },
    {
      "epoch": 1.0246331236897275,
      "grad_norm": 0.9485344886779785,
      "learning_rate": 2.5602725366876312e-05,
      "loss": 0.5376,
      "num_input_tokens_seen": 1278160,
      "step": 1955
    },
    {
      "epoch": 1.0272536687631026,
      "grad_norm": 1.4968823194503784,
      "learning_rate": 2.566823899371069e-05,
      "loss": 0.4488,
      "num_input_tokens_seen": 1281424,
      "step": 1960
    },
    {
      "epoch": 1.029874213836478,
      "grad_norm": 1.0812995433807373,
      "learning_rate": 2.5733752620545075e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 1283920,
      "step": 1965
    },
    {
      "epoch": 1.0324947589098532,
      "grad_norm": 1.8017246723175049,
      "learning_rate": 2.5799266247379454e-05,
      "loss": 0.4141,
      "num_input_tokens_seen": 1287920,
      "step": 1970
    },
    {
      "epoch": 1.0351153039832286,
      "grad_norm": 2.14921498298645,
      "learning_rate": 2.586477987421384e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 1290960,
      "step": 1975
    },
    {
      "epoch": 1.0377358490566038,
      "grad_norm": 1.0997629165649414,
      "learning_rate": 2.593029350104822e-05,
      "loss": 0.3531,
      "num_input_tokens_seen": 1294544,
      "step": 1980
    },
    {
      "epoch": 1.040356394129979,
      "grad_norm": 0.8998910188674927,
      "learning_rate": 2.59958071278826e-05,
      "loss": 0.6244,
      "num_input_tokens_seen": 1298160,
      "step": 1985
    },
    {
      "epoch": 1.0429769392033543,
      "grad_norm": 1.0903269052505493,
      "learning_rate": 2.6061320754716983e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 1300368,
      "step": 1990
    },
    {
      "epoch": 1.0455974842767295,
      "grad_norm": 0.7534974217414856,
      "learning_rate": 2.6126834381551363e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 1307184,
      "step": 1995
    },
    {
      "epoch": 1.0482180293501049,
      "grad_norm": 1.661173701286316,
      "learning_rate": 2.6192348008385746e-05,
      "loss": 0.7226,
      "num_input_tokens_seen": 1309840,
      "step": 2000
    },
    {
      "epoch": 1.05083857442348,
      "grad_norm": 1.3201872110366821,
      "learning_rate": 2.6257861635220126e-05,
      "loss": 0.4379,
      "num_input_tokens_seen": 1313648,
      "step": 2005
    },
    {
      "epoch": 1.0534591194968554,
      "grad_norm": 1.2389806509017944,
      "learning_rate": 2.6323375262054505e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 1316496,
      "step": 2010
    },
    {
      "epoch": 1.0560796645702306,
      "grad_norm": 1.9144022464752197,
      "learning_rate": 2.6388888888888892e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 1319952,
      "step": 2015
    },
    {
      "epoch": 1.0587002096436058,
      "grad_norm": 1.194970965385437,
      "learning_rate": 2.645440251572327e-05,
      "loss": 0.4267,
      "num_input_tokens_seen": 1323344,
      "step": 2020
    },
    {
      "epoch": 1.0613207547169812,
      "grad_norm": 1.4783294200897217,
      "learning_rate": 2.6519916142557655e-05,
      "loss": 0.5222,
      "num_input_tokens_seen": 1326704,
      "step": 2025
    },
    {
      "epoch": 1.0639412997903563,
      "grad_norm": 0.9192572832107544,
      "learning_rate": 2.6585429769392034e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 1330128,
      "step": 2030
    },
    {
      "epoch": 1.0665618448637317,
      "grad_norm": 1.17734694480896,
      "learning_rate": 2.6650943396226417e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 1333040,
      "step": 2035
    },
    {
      "epoch": 1.069182389937107,
      "grad_norm": 1.2605372667312622,
      "learning_rate": 2.6716457023060797e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 1335504,
      "step": 2040
    },
    {
      "epoch": 1.0718029350104823,
      "grad_norm": 1.7234807014465332,
      "learning_rate": 2.6781970649895177e-05,
      "loss": 0.5059,
      "num_input_tokens_seen": 1339280,
      "step": 2045
    },
    {
      "epoch": 1.0744234800838575,
      "grad_norm": 1.7322921752929688,
      "learning_rate": 2.6847484276729563e-05,
      "loss": 0.5321,
      "num_input_tokens_seen": 1342576,
      "step": 2050
    },
    {
      "epoch": 1.0770440251572326,
      "grad_norm": 1.2220631837844849,
      "learning_rate": 2.6912997903563943e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 1346256,
      "step": 2055
    },
    {
      "epoch": 1.079664570230608,
      "grad_norm": 1.4665502309799194,
      "learning_rate": 2.6978511530398326e-05,
      "loss": 0.5555,
      "num_input_tokens_seen": 1349008,
      "step": 2060
    },
    {
      "epoch": 1.0822851153039832,
      "grad_norm": 1.181148648262024,
      "learning_rate": 2.7044025157232706e-05,
      "loss": 0.5261,
      "num_input_tokens_seen": 1353552,
      "step": 2065
    },
    {
      "epoch": 1.0849056603773586,
      "grad_norm": 1.4291411638259888,
      "learning_rate": 2.710953878406709e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 1356400,
      "step": 2070
    },
    {
      "epoch": 1.0875262054507338,
      "grad_norm": 1.5800479650497437,
      "learning_rate": 2.717505241090147e-05,
      "loss": 0.4783,
      "num_input_tokens_seen": 1361008,
      "step": 2075
    },
    {
      "epoch": 1.090146750524109,
      "grad_norm": 2.8716955184936523,
      "learning_rate": 2.7240566037735848e-05,
      "loss": 0.5299,
      "num_input_tokens_seen": 1363184,
      "step": 2080
    },
    {
      "epoch": 1.0927672955974843,
      "grad_norm": 1.9465808868408203,
      "learning_rate": 2.7306079664570235e-05,
      "loss": 0.4511,
      "num_input_tokens_seen": 1365872,
      "step": 2085
    },
    {
      "epoch": 1.0953878406708595,
      "grad_norm": 1.0104401111602783,
      "learning_rate": 2.737159329140461e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 1369168,
      "step": 2090
    },
    {
      "epoch": 1.0980083857442349,
      "grad_norm": 1.1121481657028198,
      "learning_rate": 2.7437106918238997e-05,
      "loss": 0.4588,
      "num_input_tokens_seen": 1372208,
      "step": 2095
    },
    {
      "epoch": 1.10062893081761,
      "grad_norm": 1.1901297569274902,
      "learning_rate": 2.7502620545073377e-05,
      "loss": 0.3748,
      "num_input_tokens_seen": 1377360,
      "step": 2100
    },
    {
      "epoch": 1.1032494758909852,
      "grad_norm": 2.1477787494659424,
      "learning_rate": 2.7568134171907757e-05,
      "loss": 0.402,
      "num_input_tokens_seen": 1380400,
      "step": 2105
    },
    {
      "epoch": 1.1058700209643606,
      "grad_norm": 1.869063377380371,
      "learning_rate": 2.763364779874214e-05,
      "loss": 0.4416,
      "num_input_tokens_seen": 1383056,
      "step": 2110
    },
    {
      "epoch": 1.1084905660377358,
      "grad_norm": 1.6130743026733398,
      "learning_rate": 2.769916142557652e-05,
      "loss": 0.5296,
      "num_input_tokens_seen": 1386160,
      "step": 2115
    },
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 1.365220308303833,
      "learning_rate": 2.7764675052410906e-05,
      "loss": 0.4361,
      "num_input_tokens_seen": 1389232,
      "step": 2120
    },
    {
      "epoch": 1.1137316561844863,
      "grad_norm": 1.1250954866409302,
      "learning_rate": 2.7830188679245282e-05,
      "loss": 0.5491,
      "num_input_tokens_seen": 1393328,
      "step": 2125
    },
    {
      "epoch": 1.1163522012578617,
      "grad_norm": 1.1218178272247314,
      "learning_rate": 2.789570230607967e-05,
      "loss": 0.4165,
      "num_input_tokens_seen": 1396816,
      "step": 2130
    },
    {
      "epoch": 1.118972746331237,
      "grad_norm": 1.4751542806625366,
      "learning_rate": 2.796121593291405e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 1400176,
      "step": 2135
    },
    {
      "epoch": 1.121593291404612,
      "grad_norm": 1.2682143449783325,
      "learning_rate": 2.8026729559748428e-05,
      "loss": 0.5527,
      "num_input_tokens_seen": 1403408,
      "step": 2140
    },
    {
      "epoch": 1.1242138364779874,
      "grad_norm": 1.1258819103240967,
      "learning_rate": 2.809224318658281e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 1406448,
      "step": 2145
    },
    {
      "epoch": 1.1268343815513626,
      "grad_norm": 3.434342384338379,
      "learning_rate": 2.815775681341719e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 1409584,
      "step": 2150
    },
    {
      "epoch": 1.129454926624738,
      "grad_norm": 1.5856891870498657,
      "learning_rate": 2.8223270440251577e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 1413872,
      "step": 2155
    },
    {
      "epoch": 1.1320754716981132,
      "grad_norm": 1.3720319271087646,
      "learning_rate": 2.8288784067085954e-05,
      "loss": 0.6167,
      "num_input_tokens_seen": 1416528,
      "step": 2160
    },
    {
      "epoch": 1.1346960167714886,
      "grad_norm": 1.5753345489501953,
      "learning_rate": 2.8354297693920333e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 1418896,
      "step": 2165
    },
    {
      "epoch": 1.1373165618448637,
      "grad_norm": 1.6418986320495605,
      "learning_rate": 2.841981132075472e-05,
      "loss": 0.516,
      "num_input_tokens_seen": 1421712,
      "step": 2170
    },
    {
      "epoch": 1.139937106918239,
      "grad_norm": 1.1151642799377441,
      "learning_rate": 2.84853249475891e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 1425072,
      "step": 2175
    },
    {
      "epoch": 1.1425576519916143,
      "grad_norm": 1.064461588859558,
      "learning_rate": 2.8550838574423482e-05,
      "loss": 0.6512,
      "num_input_tokens_seen": 1428848,
      "step": 2180
    },
    {
      "epoch": 1.1451781970649895,
      "grad_norm": 1.4366748332977295,
      "learning_rate": 2.8616352201257862e-05,
      "loss": 0.4974,
      "num_input_tokens_seen": 1432432,
      "step": 2185
    },
    {
      "epoch": 1.1477987421383649,
      "grad_norm": 1.2078688144683838,
      "learning_rate": 2.868186582809225e-05,
      "loss": 0.6417,
      "num_input_tokens_seen": 1436048,
      "step": 2190
    },
    {
      "epoch": 1.15041928721174,
      "grad_norm": 1.9002025127410889,
      "learning_rate": 2.8747379454926625e-05,
      "loss": 0.6179,
      "num_input_tokens_seen": 1440144,
      "step": 2195
    },
    {
      "epoch": 1.1530398322851152,
      "grad_norm": 1.2152127027511597,
      "learning_rate": 2.8812893081761005e-05,
      "loss": 0.5435,
      "num_input_tokens_seen": 1443760,
      "step": 2200
    },
    {
      "epoch": 1.1556603773584906,
      "grad_norm": 1.2940795421600342,
      "learning_rate": 2.887840670859539e-05,
      "loss": 0.7429,
      "num_input_tokens_seen": 1446864,
      "step": 2205
    },
    {
      "epoch": 1.1582809224318658,
      "grad_norm": 1.1875152587890625,
      "learning_rate": 2.894392033542977e-05,
      "loss": 0.3158,
      "num_input_tokens_seen": 1449456,
      "step": 2210
    },
    {
      "epoch": 1.1609014675052411,
      "grad_norm": 1.2192507982254028,
      "learning_rate": 2.9009433962264154e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 1452944,
      "step": 2215
    },
    {
      "epoch": 1.1635220125786163,
      "grad_norm": 0.9514321684837341,
      "learning_rate": 2.9074947589098533e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 1455952,
      "step": 2220
    },
    {
      "epoch": 1.1661425576519917,
      "grad_norm": 1.2864208221435547,
      "learning_rate": 2.9140461215932913e-05,
      "loss": 0.5714,
      "num_input_tokens_seen": 1459376,
      "step": 2225
    },
    {
      "epoch": 1.1687631027253669,
      "grad_norm": 1.1172130107879639,
      "learning_rate": 2.9205974842767296e-05,
      "loss": 0.4568,
      "num_input_tokens_seen": 1462352,
      "step": 2230
    },
    {
      "epoch": 1.171383647798742,
      "grad_norm": 1.678428053855896,
      "learning_rate": 2.9271488469601676e-05,
      "loss": 0.55,
      "num_input_tokens_seen": 1465104,
      "step": 2235
    },
    {
      "epoch": 1.1740041928721174,
      "grad_norm": 0.9615904688835144,
      "learning_rate": 2.9337002096436062e-05,
      "loss": 0.5201,
      "num_input_tokens_seen": 1468624,
      "step": 2240
    },
    {
      "epoch": 1.1766247379454926,
      "grad_norm": 1.3666311502456665,
      "learning_rate": 2.9402515723270442e-05,
      "loss": 0.4042,
      "num_input_tokens_seen": 1475184,
      "step": 2245
    },
    {
      "epoch": 1.179245283018868,
      "grad_norm": 0.6408969759941101,
      "learning_rate": 2.9468029350104825e-05,
      "loss": 0.3685,
      "num_input_tokens_seen": 1478640,
      "step": 2250
    },
    {
      "epoch": 1.1818658280922432,
      "grad_norm": 1.6460199356079102,
      "learning_rate": 2.9533542976939205e-05,
      "loss": 0.5231,
      "num_input_tokens_seen": 1481552,
      "step": 2255
    },
    {
      "epoch": 1.1844863731656186,
      "grad_norm": 1.2504994869232178,
      "learning_rate": 2.9599056603773584e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 1485104,
      "step": 2260
    },
    {
      "epoch": 1.1871069182389937,
      "grad_norm": 1.2435284852981567,
      "learning_rate": 2.9664570230607968e-05,
      "loss": 0.498,
      "num_input_tokens_seen": 1487728,
      "step": 2265
    },
    {
      "epoch": 1.189727463312369,
      "grad_norm": 1.1403037309646606,
      "learning_rate": 2.9730083857442347e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 1491312,
      "step": 2270
    },
    {
      "epoch": 1.1923480083857443,
      "grad_norm": 2.0713672637939453,
      "learning_rate": 2.9795597484276734e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 1494192,
      "step": 2275
    },
    {
      "epoch": 1.1949685534591195,
      "grad_norm": 1.5903410911560059,
      "learning_rate": 2.9861111111111113e-05,
      "loss": 0.6157,
      "num_input_tokens_seen": 1497168,
      "step": 2280
    },
    {
      "epoch": 1.1975890985324948,
      "grad_norm": 1.4199756383895874,
      "learning_rate": 2.9926624737945496e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 1501936,
      "step": 2285
    },
    {
      "epoch": 1.20020964360587,
      "grad_norm": 0.9725185036659241,
      "learning_rate": 2.9992138364779876e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 1505040,
      "step": 2290
    },
    {
      "epoch": 1.2028301886792452,
      "grad_norm": 1.7497847080230713,
      "learning_rate": 3.0057651991614256e-05,
      "loss": 0.6102,
      "num_input_tokens_seen": 1507856,
      "step": 2295
    },
    {
      "epoch": 1.2054507337526206,
      "grad_norm": 1.6324553489685059,
      "learning_rate": 3.012316561844864e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 1510768,
      "step": 2300
    },
    {
      "epoch": 1.2080712788259957,
      "grad_norm": 1.417231559753418,
      "learning_rate": 3.018867924528302e-05,
      "loss": 0.5981,
      "num_input_tokens_seen": 1513808,
      "step": 2305
    },
    {
      "epoch": 1.2106918238993711,
      "grad_norm": 0.8619780540466309,
      "learning_rate": 3.0254192872117405e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 1517392,
      "step": 2310
    },
    {
      "epoch": 1.2133123689727463,
      "grad_norm": 0.7549493908882141,
      "learning_rate": 3.0319706498951785e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 1521648,
      "step": 2315
    },
    {
      "epoch": 1.2159329140461215,
      "grad_norm": 1.3048080205917358,
      "learning_rate": 3.038522012578616e-05,
      "loss": 0.4232,
      "num_input_tokens_seen": 1524496,
      "step": 2320
    },
    {
      "epoch": 1.2185534591194969,
      "grad_norm": 1.2213858366012573,
      "learning_rate": 3.0450733752620547e-05,
      "loss": 0.4965,
      "num_input_tokens_seen": 1528144,
      "step": 2325
    },
    {
      "epoch": 1.221174004192872,
      "grad_norm": 1.4192748069763184,
      "learning_rate": 3.0516247379454927e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 1531664,
      "step": 2330
    },
    {
      "epoch": 1.2237945492662474,
      "grad_norm": 1.670950174331665,
      "learning_rate": 3.0581761006289314e-05,
      "loss": 0.5805,
      "num_input_tokens_seen": 1535408,
      "step": 2335
    },
    {
      "epoch": 1.2264150943396226,
      "grad_norm": 1.0210933685302734,
      "learning_rate": 3.064727463312369e-05,
      "loss": 0.4774,
      "num_input_tokens_seen": 1538448,
      "step": 2340
    },
    {
      "epoch": 1.229035639412998,
      "grad_norm": 1.2445347309112549,
      "learning_rate": 3.071278825995807e-05,
      "loss": 0.5075,
      "num_input_tokens_seen": 1541968,
      "step": 2345
    },
    {
      "epoch": 1.2316561844863732,
      "grad_norm": 1.1533499956130981,
      "learning_rate": 3.0778301886792456e-05,
      "loss": 0.5195,
      "num_input_tokens_seen": 1544560,
      "step": 2350
    },
    {
      "epoch": 1.2342767295597485,
      "grad_norm": 1.6192859411239624,
      "learning_rate": 3.084381551362683e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 1548080,
      "step": 2355
    },
    {
      "epoch": 1.2368972746331237,
      "grad_norm": 1.7127760648727417,
      "learning_rate": 3.0909329140461215e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 1551472,
      "step": 2360
    },
    {
      "epoch": 1.2395178197064989,
      "grad_norm": 1.0494384765625,
      "learning_rate": 3.09748427672956e-05,
      "loss": 0.5382,
      "num_input_tokens_seen": 1553904,
      "step": 2365
    },
    {
      "epoch": 1.2421383647798743,
      "grad_norm": 1.1294807195663452,
      "learning_rate": 3.104035639412998e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 1558352,
      "step": 2370
    },
    {
      "epoch": 1.2447589098532494,
      "grad_norm": 1.8631935119628906,
      "learning_rate": 3.1105870020964365e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 1561616,
      "step": 2375
    },
    {
      "epoch": 1.2473794549266248,
      "grad_norm": 1.037449836730957,
      "learning_rate": 3.117138364779874e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 1566000,
      "step": 2380
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.9012345671653748,
      "learning_rate": 3.1236897274633124e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 1569776,
      "step": 2385
    },
    {
      "epoch": 1.2526205450733752,
      "grad_norm": 1.6752454042434692,
      "learning_rate": 3.130241090146751e-05,
      "loss": 0.5438,
      "num_input_tokens_seen": 1572112,
      "step": 2390
    },
    {
      "epoch": 1.2552410901467506,
      "grad_norm": 0.8405994772911072,
      "learning_rate": 3.136792452830189e-05,
      "loss": 0.4919,
      "num_input_tokens_seen": 1575312,
      "step": 2395
    },
    {
      "epoch": 1.2578616352201257,
      "grad_norm": 0.9886555075645447,
      "learning_rate": 3.1433438155136266e-05,
      "loss": 0.5564,
      "num_input_tokens_seen": 1578256,
      "step": 2400
    },
    {
      "epoch": 1.2604821802935011,
      "grad_norm": 1.1776065826416016,
      "learning_rate": 3.1498951781970656e-05,
      "loss": 0.6539,
      "num_input_tokens_seen": 1581328,
      "step": 2405
    },
    {
      "epoch": 1.2631027253668763,
      "grad_norm": 1.1553181409835815,
      "learning_rate": 3.156446540880503e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 1583952,
      "step": 2410
    },
    {
      "epoch": 1.2657232704402515,
      "grad_norm": 2.2757766246795654,
      "learning_rate": 3.162997903563941e-05,
      "loss": 0.4849,
      "num_input_tokens_seen": 1587472,
      "step": 2415
    },
    {
      "epoch": 1.2683438155136268,
      "grad_norm": 1.4335726499557495,
      "learning_rate": 3.16954926624738e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 1590384,
      "step": 2420
    },
    {
      "epoch": 1.270964360587002,
      "grad_norm": 1.3732612133026123,
      "learning_rate": 3.1761006289308175e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 1593264,
      "step": 2425
    },
    {
      "epoch": 1.2735849056603774,
      "grad_norm": 0.853746235370636,
      "learning_rate": 3.182651991614256e-05,
      "loss": 0.4205,
      "num_input_tokens_seen": 1597680,
      "step": 2430
    },
    {
      "epoch": 1.2762054507337526,
      "grad_norm": 2.043006181716919,
      "learning_rate": 3.189203354297694e-05,
      "loss": 0.5132,
      "num_input_tokens_seen": 1600912,
      "step": 2435
    },
    {
      "epoch": 1.2788259958071277,
      "grad_norm": 1.5203347206115723,
      "learning_rate": 3.1957547169811324e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 1603472,
      "step": 2440
    },
    {
      "epoch": 1.2814465408805031,
      "grad_norm": 2.713966131210327,
      "learning_rate": 3.202306079664571e-05,
      "loss": 0.4372,
      "num_input_tokens_seen": 1606384,
      "step": 2445
    },
    {
      "epoch": 1.2840670859538785,
      "grad_norm": 1.038334846496582,
      "learning_rate": 3.2088574423480084e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 1609360,
      "step": 2450
    },
    {
      "epoch": 1.2866876310272537,
      "grad_norm": 1.7261629104614258,
      "learning_rate": 3.215408805031447e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 1612336,
      "step": 2455
    },
    {
      "epoch": 1.2893081761006289,
      "grad_norm": 1.0960794687271118,
      "learning_rate": 3.221960167714885e-05,
      "loss": 0.3954,
      "num_input_tokens_seen": 1616048,
      "step": 2460
    },
    {
      "epoch": 1.2919287211740043,
      "grad_norm": 2.2848775386810303,
      "learning_rate": 3.228511530398323e-05,
      "loss": 0.3752,
      "num_input_tokens_seen": 1618960,
      "step": 2465
    },
    {
      "epoch": 1.2945492662473794,
      "grad_norm": 0.7887179851531982,
      "learning_rate": 3.235062893081761e-05,
      "loss": 0.336,
      "num_input_tokens_seen": 1622544,
      "step": 2470
    },
    {
      "epoch": 1.2971698113207548,
      "grad_norm": 1.0174744129180908,
      "learning_rate": 3.241614255765199e-05,
      "loss": 0.4439,
      "num_input_tokens_seen": 1625872,
      "step": 2475
    },
    {
      "epoch": 1.29979035639413,
      "grad_norm": 0.8798580765724182,
      "learning_rate": 3.2481656184486375e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 1629456,
      "step": 2480
    },
    {
      "epoch": 1.3024109014675052,
      "grad_norm": 1.0128560066223145,
      "learning_rate": 3.254716981132075e-05,
      "loss": 0.5714,
      "num_input_tokens_seen": 1632432,
      "step": 2485
    },
    {
      "epoch": 1.3050314465408805,
      "grad_norm": 0.8342604637145996,
      "learning_rate": 3.261268343815514e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 1636080,
      "step": 2490
    },
    {
      "epoch": 1.3076519916142557,
      "grad_norm": 0.9787384867668152,
      "learning_rate": 3.267819706498952e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 1638960,
      "step": 2495
    },
    {
      "epoch": 1.310272536687631,
      "grad_norm": 0.52146315574646,
      "learning_rate": 3.27437106918239e-05,
      "loss": 0.4534,
      "num_input_tokens_seen": 1643120,
      "step": 2500
    },
    {
      "epoch": 1.3128930817610063,
      "grad_norm": 1.4440808296203613,
      "learning_rate": 3.2809224318658284e-05,
      "loss": 0.385,
      "num_input_tokens_seen": 1646768,
      "step": 2505
    },
    {
      "epoch": 1.3155136268343814,
      "grad_norm": 0.9708005785942078,
      "learning_rate": 3.287473794549266e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 1650256,
      "step": 2510
    },
    {
      "epoch": 1.3181341719077568,
      "grad_norm": 1.0552177429199219,
      "learning_rate": 3.294025157232704e-05,
      "loss": 0.4953,
      "num_input_tokens_seen": 1652912,
      "step": 2515
    },
    {
      "epoch": 1.320754716981132,
      "grad_norm": 0.6639091968536377,
      "learning_rate": 3.3005765199161426e-05,
      "loss": 0.4725,
      "num_input_tokens_seen": 1656432,
      "step": 2520
    },
    {
      "epoch": 1.3233752620545074,
      "grad_norm": 1.0614778995513916,
      "learning_rate": 3.307127882599581e-05,
      "loss": 0.3954,
      "num_input_tokens_seen": 1659472,
      "step": 2525
    },
    {
      "epoch": 1.3259958071278826,
      "grad_norm": 2.045283555984497,
      "learning_rate": 3.313679245283019e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 1662480,
      "step": 2530
    },
    {
      "epoch": 1.3286163522012577,
      "grad_norm": 0.9547843337059021,
      "learning_rate": 3.320230607966457e-05,
      "loss": 0.5204,
      "num_input_tokens_seen": 1665424,
      "step": 2535
    },
    {
      "epoch": 1.3312368972746331,
      "grad_norm": 1.353561520576477,
      "learning_rate": 3.326781970649895e-05,
      "loss": 0.53,
      "num_input_tokens_seen": 1667760,
      "step": 2540
    },
    {
      "epoch": 1.3338574423480085,
      "grad_norm": 1.1859780550003052,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.3832,
      "num_input_tokens_seen": 1670800,
      "step": 2545
    },
    {
      "epoch": 1.3364779874213837,
      "grad_norm": 1.2120261192321777,
      "learning_rate": 3.339884696016772e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 1675760,
      "step": 2550
    },
    {
      "epoch": 1.3390985324947589,
      "grad_norm": 2.5782012939453125,
      "learning_rate": 3.3464360587002094e-05,
      "loss": 0.5548,
      "num_input_tokens_seen": 1677936,
      "step": 2555
    },
    {
      "epoch": 1.3417190775681342,
      "grad_norm": 0.9149587154388428,
      "learning_rate": 3.3529874213836484e-05,
      "loss": 0.5665,
      "num_input_tokens_seen": 1685008,
      "step": 2560
    },
    {
      "epoch": 1.3443396226415094,
      "grad_norm": 1.3638646602630615,
      "learning_rate": 3.359538784067086e-05,
      "loss": 0.5916,
      "num_input_tokens_seen": 1688240,
      "step": 2565
    },
    {
      "epoch": 1.3469601677148848,
      "grad_norm": 1.5252044200897217,
      "learning_rate": 3.3660901467505244e-05,
      "loss": 0.6286,
      "num_input_tokens_seen": 1690832,
      "step": 2570
    },
    {
      "epoch": 1.34958071278826,
      "grad_norm": 1.0303153991699219,
      "learning_rate": 3.3726415094339627e-05,
      "loss": 0.6035,
      "num_input_tokens_seen": 1695248,
      "step": 2575
    },
    {
      "epoch": 1.3522012578616351,
      "grad_norm": 1.1225000619888306,
      "learning_rate": 3.3791928721174e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 1698896,
      "step": 2580
    },
    {
      "epoch": 1.3548218029350105,
      "grad_norm": 1.6490352153778076,
      "learning_rate": 3.3857442348008386e-05,
      "loss": 0.5938,
      "num_input_tokens_seen": 1701296,
      "step": 2585
    },
    {
      "epoch": 1.3574423480083857,
      "grad_norm": 1.7535251379013062,
      "learning_rate": 3.392295597484277e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 1705136,
      "step": 2590
    },
    {
      "epoch": 1.360062893081761,
      "grad_norm": 1.2281982898712158,
      "learning_rate": 3.398846960167715e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 1708016,
      "step": 2595
    },
    {
      "epoch": 1.3626834381551363,
      "grad_norm": 3.9152426719665527,
      "learning_rate": 3.4053983228511535e-05,
      "loss": 0.604,
      "num_input_tokens_seen": 1710864,
      "step": 2600
    },
    {
      "epoch": 1.3653039832285114,
      "grad_norm": 1.0465699434280396,
      "learning_rate": 3.411949685534591e-05,
      "loss": 0.3936,
      "num_input_tokens_seen": 1713552,
      "step": 2605
    },
    {
      "epoch": 1.3679245283018868,
      "grad_norm": 1.0872142314910889,
      "learning_rate": 3.4185010482180295e-05,
      "loss": 0.4097,
      "num_input_tokens_seen": 1716528,
      "step": 2610
    },
    {
      "epoch": 1.370545073375262,
      "grad_norm": 3.542109966278076,
      "learning_rate": 3.425052410901468e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 1719216,
      "step": 2615
    },
    {
      "epoch": 1.3731656184486374,
      "grad_norm": 0.9788249731063843,
      "learning_rate": 3.431603773584906e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 1722384,
      "step": 2620
    },
    {
      "epoch": 1.3757861635220126,
      "grad_norm": 0.9060404896736145,
      "learning_rate": 3.438155136268344e-05,
      "loss": 0.4267,
      "num_input_tokens_seen": 1725104,
      "step": 2625
    },
    {
      "epoch": 1.3784067085953877,
      "grad_norm": 4.117213726043701,
      "learning_rate": 3.444706498951782e-05,
      "loss": 0.4437,
      "num_input_tokens_seen": 1728112,
      "step": 2630
    },
    {
      "epoch": 1.381027253668763,
      "grad_norm": 1.3817458152770996,
      "learning_rate": 3.45125786163522e-05,
      "loss": 0.4852,
      "num_input_tokens_seen": 1731536,
      "step": 2635
    },
    {
      "epoch": 1.3836477987421385,
      "grad_norm": 0.8503803014755249,
      "learning_rate": 3.457809224318658e-05,
      "loss": 0.5252,
      "num_input_tokens_seen": 1734896,
      "step": 2640
    },
    {
      "epoch": 1.3862683438155137,
      "grad_norm": 1.917037844657898,
      "learning_rate": 3.464360587002097e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 1737584,
      "step": 2645
    },
    {
      "epoch": 1.3888888888888888,
      "grad_norm": 1.2518603801727295,
      "learning_rate": 3.4709119496855346e-05,
      "loss": 0.5482,
      "num_input_tokens_seen": 1740720,
      "step": 2650
    },
    {
      "epoch": 1.3915094339622642,
      "grad_norm": 0.9410755038261414,
      "learning_rate": 3.477463312368973e-05,
      "loss": 0.3392,
      "num_input_tokens_seen": 1743216,
      "step": 2655
    },
    {
      "epoch": 1.3941299790356394,
      "grad_norm": 1.401098608970642,
      "learning_rate": 3.484014675052411e-05,
      "loss": 0.5271,
      "num_input_tokens_seen": 1746000,
      "step": 2660
    },
    {
      "epoch": 1.3967505241090148,
      "grad_norm": 2.123605251312256,
      "learning_rate": 3.490566037735849e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 1748560,
      "step": 2665
    },
    {
      "epoch": 1.39937106918239,
      "grad_norm": 1.2824671268463135,
      "learning_rate": 3.497117400419288e-05,
      "loss": 0.542,
      "num_input_tokens_seen": 1751504,
      "step": 2670
    },
    {
      "epoch": 1.4019916142557651,
      "grad_norm": 1.5118260383605957,
      "learning_rate": 3.5036687631027254e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 1754864,
      "step": 2675
    },
    {
      "epoch": 1.4046121593291405,
      "grad_norm": 2.348820209503174,
      "learning_rate": 3.510220125786164e-05,
      "loss": 0.4715,
      "num_input_tokens_seen": 1757648,
      "step": 2680
    },
    {
      "epoch": 1.4072327044025157,
      "grad_norm": 1.04496431350708,
      "learning_rate": 3.516771488469602e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 1760432,
      "step": 2685
    },
    {
      "epoch": 1.409853249475891,
      "grad_norm": 0.7505095601081848,
      "learning_rate": 3.5233228511530397e-05,
      "loss": 0.4834,
      "num_input_tokens_seen": 1763664,
      "step": 2690
    },
    {
      "epoch": 1.4124737945492662,
      "grad_norm": 1.0793790817260742,
      "learning_rate": 3.529874213836478e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 1766640,
      "step": 2695
    },
    {
      "epoch": 1.4150943396226414,
      "grad_norm": 1.1729893684387207,
      "learning_rate": 3.536425576519916e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 1769776,
      "step": 2700
    },
    {
      "epoch": 1.4177148846960168,
      "grad_norm": 1.827548623085022,
      "learning_rate": 3.5429769392033546e-05,
      "loss": 0.513,
      "num_input_tokens_seen": 1772112,
      "step": 2705
    },
    {
      "epoch": 1.420335429769392,
      "grad_norm": 1.0716309547424316,
      "learning_rate": 3.549528301886792e-05,
      "loss": 0.4658,
      "num_input_tokens_seen": 1775312,
      "step": 2710
    },
    {
      "epoch": 1.4229559748427674,
      "grad_norm": 0.8407943248748779,
      "learning_rate": 3.556079664570231e-05,
      "loss": 0.5959,
      "num_input_tokens_seen": 1779504,
      "step": 2715
    },
    {
      "epoch": 1.4255765199161425,
      "grad_norm": 0.8635473847389221,
      "learning_rate": 3.562631027253669e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 1783248,
      "step": 2720
    },
    {
      "epoch": 1.4281970649895177,
      "grad_norm": 2.047332286834717,
      "learning_rate": 3.569182389937107e-05,
      "loss": 0.3472,
      "num_input_tokens_seen": 1789168,
      "step": 2725
    },
    {
      "epoch": 1.430817610062893,
      "grad_norm": 1.2397516965866089,
      "learning_rate": 3.5757337526205454e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 1792592,
      "step": 2730
    },
    {
      "epoch": 1.4334381551362683,
      "grad_norm": 1.6285920143127441,
      "learning_rate": 3.582285115303983e-05,
      "loss": 0.4904,
      "num_input_tokens_seen": 1795536,
      "step": 2735
    },
    {
      "epoch": 1.4360587002096437,
      "grad_norm": 0.9286166429519653,
      "learning_rate": 3.588836477987422e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 1798064,
      "step": 2740
    },
    {
      "epoch": 1.4386792452830188,
      "grad_norm": 1.123016357421875,
      "learning_rate": 3.59538784067086e-05,
      "loss": 0.4659,
      "num_input_tokens_seen": 1801392,
      "step": 2745
    },
    {
      "epoch": 1.441299790356394,
      "grad_norm": 1.5374678373336792,
      "learning_rate": 3.601939203354298e-05,
      "loss": 0.5285,
      "num_input_tokens_seen": 1804368,
      "step": 2750
    },
    {
      "epoch": 1.4439203354297694,
      "grad_norm": 1.7887508869171143,
      "learning_rate": 3.608490566037736e-05,
      "loss": 0.5262,
      "num_input_tokens_seen": 1807504,
      "step": 2755
    },
    {
      "epoch": 1.4465408805031448,
      "grad_norm": 1.3802886009216309,
      "learning_rate": 3.615041928721174e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 1810160,
      "step": 2760
    },
    {
      "epoch": 1.44916142557652,
      "grad_norm": 1.3658630847930908,
      "learning_rate": 3.621593291404612e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 1813616,
      "step": 2765
    },
    {
      "epoch": 1.4517819706498951,
      "grad_norm": 1.0603950023651123,
      "learning_rate": 3.6281446540880505e-05,
      "loss": 0.4432,
      "num_input_tokens_seen": 1816752,
      "step": 2770
    },
    {
      "epoch": 1.4544025157232705,
      "grad_norm": 1.154689908027649,
      "learning_rate": 3.634696016771489e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 1820784,
      "step": 2775
    },
    {
      "epoch": 1.4570230607966457,
      "grad_norm": 1.4365513324737549,
      "learning_rate": 3.6412473794549265e-05,
      "loss": 0.3938,
      "num_input_tokens_seen": 1823024,
      "step": 2780
    },
    {
      "epoch": 1.459643605870021,
      "grad_norm": 1.12371027469635,
      "learning_rate": 3.647798742138365e-05,
      "loss": 0.5382,
      "num_input_tokens_seen": 1828016,
      "step": 2785
    },
    {
      "epoch": 1.4622641509433962,
      "grad_norm": 1.557904839515686,
      "learning_rate": 3.654350104821803e-05,
      "loss": 0.5113,
      "num_input_tokens_seen": 1831280,
      "step": 2790
    },
    {
      "epoch": 1.4648846960167714,
      "grad_norm": 0.730713963508606,
      "learning_rate": 3.6609014675052414e-05,
      "loss": 0.4139,
      "num_input_tokens_seen": 1834896,
      "step": 2795
    },
    {
      "epoch": 1.4675052410901468,
      "grad_norm": 1.0424472093582153,
      "learning_rate": 3.66745283018868e-05,
      "loss": 0.4414,
      "num_input_tokens_seen": 1837488,
      "step": 2800
    },
    {
      "epoch": 1.470125786163522,
      "grad_norm": 0.9047855138778687,
      "learning_rate": 3.6740041928721173e-05,
      "loss": 0.5171,
      "num_input_tokens_seen": 1841584,
      "step": 2805
    },
    {
      "epoch": 1.4727463312368974,
      "grad_norm": 1.1977747678756714,
      "learning_rate": 3.6805555555555556e-05,
      "loss": 0.5345,
      "num_input_tokens_seen": 1844688,
      "step": 2810
    },
    {
      "epoch": 1.4753668763102725,
      "grad_norm": 1.2651900053024292,
      "learning_rate": 3.687106918238994e-05,
      "loss": 0.4745,
      "num_input_tokens_seen": 1847472,
      "step": 2815
    },
    {
      "epoch": 1.4779874213836477,
      "grad_norm": 1.4831770658493042,
      "learning_rate": 3.6936582809224316e-05,
      "loss": 0.682,
      "num_input_tokens_seen": 1851088,
      "step": 2820
    },
    {
      "epoch": 1.480607966457023,
      "grad_norm": 0.8768284916877747,
      "learning_rate": 3.7002096436058706e-05,
      "loss": 0.4793,
      "num_input_tokens_seen": 1855216,
      "step": 2825
    },
    {
      "epoch": 1.4832285115303983,
      "grad_norm": 0.856367826461792,
      "learning_rate": 3.706761006289308e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 1857744,
      "step": 2830
    },
    {
      "epoch": 1.4858490566037736,
      "grad_norm": 1.181350588798523,
      "learning_rate": 3.7133123689727465e-05,
      "loss": 0.4102,
      "num_input_tokens_seen": 1860400,
      "step": 2835
    },
    {
      "epoch": 1.4884696016771488,
      "grad_norm": 1.355531930923462,
      "learning_rate": 3.719863731656185e-05,
      "loss": 0.521,
      "num_input_tokens_seen": 1863248,
      "step": 2840
    },
    {
      "epoch": 1.491090146750524,
      "grad_norm": 1.3130433559417725,
      "learning_rate": 3.7264150943396224e-05,
      "loss": 0.4603,
      "num_input_tokens_seen": 1866576,
      "step": 2845
    },
    {
      "epoch": 1.4937106918238994,
      "grad_norm": 1.2162578105926514,
      "learning_rate": 3.732966457023061e-05,
      "loss": 0.4304,
      "num_input_tokens_seen": 1869840,
      "step": 2850
    },
    {
      "epoch": 1.4963312368972748,
      "grad_norm": 1.1129281520843506,
      "learning_rate": 3.739517819706499e-05,
      "loss": 0.5746,
      "num_input_tokens_seen": 1872624,
      "step": 2855
    },
    {
      "epoch": 1.49895178197065,
      "grad_norm": 2.7817912101745605,
      "learning_rate": 3.7460691823899374e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 1876016,
      "step": 2860
    },
    {
      "epoch": 1.501572327044025,
      "grad_norm": 1.2089736461639404,
      "learning_rate": 3.752620545073376e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 1880976,
      "step": 2865
    },
    {
      "epoch": 1.5041928721174003,
      "grad_norm": 2.396070718765259,
      "learning_rate": 3.759171907756814e-05,
      "loss": 0.5604,
      "num_input_tokens_seen": 1883568,
      "step": 2870
    },
    {
      "epoch": 1.5068134171907757,
      "grad_norm": 1.4910128116607666,
      "learning_rate": 3.7657232704402516e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 1886800,
      "step": 2875
    },
    {
      "epoch": 1.509433962264151,
      "grad_norm": 1.349703073501587,
      "learning_rate": 3.77227463312369e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 1890672,
      "step": 2880
    },
    {
      "epoch": 1.5120545073375262,
      "grad_norm": 1.1058118343353271,
      "learning_rate": 3.778825995807128e-05,
      "loss": 0.3561,
      "num_input_tokens_seen": 1893456,
      "step": 2885
    },
    {
      "epoch": 1.5146750524109014,
      "grad_norm": 1.8564906120300293,
      "learning_rate": 3.785377358490566e-05,
      "loss": 0.4757,
      "num_input_tokens_seen": 1896752,
      "step": 2890
    },
    {
      "epoch": 1.5172955974842768,
      "grad_norm": 0.7322782278060913,
      "learning_rate": 3.791928721174005e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 1899664,
      "step": 2895
    },
    {
      "epoch": 1.519916142557652,
      "grad_norm": 1.0574018955230713,
      "learning_rate": 3.7984800838574425e-05,
      "loss": 0.3692,
      "num_input_tokens_seen": 1902320,
      "step": 2900
    },
    {
      "epoch": 1.5225366876310273,
      "grad_norm": 1.1070363521575928,
      "learning_rate": 3.805031446540881e-05,
      "loss": 0.5233,
      "num_input_tokens_seen": 1905488,
      "step": 2905
    },
    {
      "epoch": 1.5251572327044025,
      "grad_norm": 1.1629977226257324,
      "learning_rate": 3.811582809224319e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 1908144,
      "step": 2910
    },
    {
      "epoch": 1.5277777777777777,
      "grad_norm": 0.724809467792511,
      "learning_rate": 3.818134171907757e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 1911536,
      "step": 2915
    },
    {
      "epoch": 1.530398322851153,
      "grad_norm": 0.6803159117698669,
      "learning_rate": 3.824685534591195e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 1916176,
      "step": 2920
    },
    {
      "epoch": 1.5330188679245285,
      "grad_norm": 1.195420265197754,
      "learning_rate": 3.831236897274633e-05,
      "loss": 0.4567,
      "num_input_tokens_seen": 1920592,
      "step": 2925
    },
    {
      "epoch": 1.5356394129979036,
      "grad_norm": 1.187882900238037,
      "learning_rate": 3.8377882599580716e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 1923152,
      "step": 2930
    },
    {
      "epoch": 1.5382599580712788,
      "grad_norm": 1.329315423965454,
      "learning_rate": 3.844339622641509e-05,
      "loss": 0.4956,
      "num_input_tokens_seen": 1926000,
      "step": 2935
    },
    {
      "epoch": 1.540880503144654,
      "grad_norm": 1.0629950761795044,
      "learning_rate": 3.8508909853249476e-05,
      "loss": 0.5237,
      "num_input_tokens_seen": 1929072,
      "step": 2940
    },
    {
      "epoch": 1.5435010482180294,
      "grad_norm": 0.6240044832229614,
      "learning_rate": 3.857442348008386e-05,
      "loss": 0.3307,
      "num_input_tokens_seen": 1932208,
      "step": 2945
    },
    {
      "epoch": 1.5461215932914047,
      "grad_norm": 1.277529239654541,
      "learning_rate": 3.863993710691824e-05,
      "loss": 0.6308,
      "num_input_tokens_seen": 1935152,
      "step": 2950
    },
    {
      "epoch": 1.54874213836478,
      "grad_norm": 1.0281760692596436,
      "learning_rate": 3.8705450733752625e-05,
      "loss": 0.5213,
      "num_input_tokens_seen": 1938704,
      "step": 2955
    },
    {
      "epoch": 1.551362683438155,
      "grad_norm": 0.9927992224693298,
      "learning_rate": 3.8770964360587e-05,
      "loss": 0.5464,
      "num_input_tokens_seen": 1942192,
      "step": 2960
    },
    {
      "epoch": 1.5539832285115303,
      "grad_norm": 1.8382102251052856,
      "learning_rate": 3.883647798742139e-05,
      "loss": 0.4301,
      "num_input_tokens_seen": 1945744,
      "step": 2965
    },
    {
      "epoch": 1.5566037735849056,
      "grad_norm": 0.988781750202179,
      "learning_rate": 3.890199161425577e-05,
      "loss": 0.456,
      "num_input_tokens_seen": 1948944,
      "step": 2970
    },
    {
      "epoch": 1.559224318658281,
      "grad_norm": 1.1608911752700806,
      "learning_rate": 3.8967505241090144e-05,
      "loss": 0.4307,
      "num_input_tokens_seen": 1951792,
      "step": 2975
    },
    {
      "epoch": 1.5618448637316562,
      "grad_norm": 0.9789219498634338,
      "learning_rate": 3.9033018867924534e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 1955376,
      "step": 2980
    },
    {
      "epoch": 1.5644654088050314,
      "grad_norm": 0.8732677102088928,
      "learning_rate": 3.909853249475891e-05,
      "loss": 0.4855,
      "num_input_tokens_seen": 1959120,
      "step": 2985
    },
    {
      "epoch": 1.5670859538784065,
      "grad_norm": 0.711575984954834,
      "learning_rate": 3.916404612159329e-05,
      "loss": 0.3743,
      "num_input_tokens_seen": 1964304,
      "step": 2990
    },
    {
      "epoch": 1.569706498951782,
      "grad_norm": 1.3834688663482666,
      "learning_rate": 3.9229559748427676e-05,
      "loss": 0.5761,
      "num_input_tokens_seen": 1966832,
      "step": 2995
    },
    {
      "epoch": 1.5723270440251573,
      "grad_norm": 0.8239392042160034,
      "learning_rate": 3.929507337526205e-05,
      "loss": 0.5471,
      "num_input_tokens_seen": 1970224,
      "step": 3000
    },
    {
      "epoch": 1.5749475890985325,
      "grad_norm": 0.6849537491798401,
      "learning_rate": 3.9360587002096435e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 1974288,
      "step": 3005
    },
    {
      "epoch": 1.5775681341719077,
      "grad_norm": 0.9583351612091064,
      "learning_rate": 3.942610062893082e-05,
      "loss": 0.4063,
      "num_input_tokens_seen": 1978064,
      "step": 3010
    },
    {
      "epoch": 1.580188679245283,
      "grad_norm": 1.5907776355743408,
      "learning_rate": 3.94916142557652e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 1981744,
      "step": 3015
    },
    {
      "epoch": 1.5828092243186582,
      "grad_norm": 1.7945626974105835,
      "learning_rate": 3.9557127882599585e-05,
      "loss": 0.4914,
      "num_input_tokens_seen": 1984016,
      "step": 3020
    },
    {
      "epoch": 1.5854297693920336,
      "grad_norm": 1.2637096643447876,
      "learning_rate": 3.962264150943397e-05,
      "loss": 0.3403,
      "num_input_tokens_seen": 1987056,
      "step": 3025
    },
    {
      "epoch": 1.5880503144654088,
      "grad_norm": 0.9007876515388489,
      "learning_rate": 3.9688155136268344e-05,
      "loss": 0.5119,
      "num_input_tokens_seen": 1992048,
      "step": 3030
    },
    {
      "epoch": 1.590670859538784,
      "grad_norm": 0.8497835397720337,
      "learning_rate": 3.975366876310273e-05,
      "loss": 0.6167,
      "num_input_tokens_seen": 1995472,
      "step": 3035
    },
    {
      "epoch": 1.5932914046121593,
      "grad_norm": 0.8558198809623718,
      "learning_rate": 3.981918238993711e-05,
      "loss": 0.3579,
      "num_input_tokens_seen": 1999152,
      "step": 3040
    },
    {
      "epoch": 1.5959119496855347,
      "grad_norm": 1.1094955205917358,
      "learning_rate": 3.9884696016771486e-05,
      "loss": 0.4982,
      "num_input_tokens_seen": 2001744,
      "step": 3045
    },
    {
      "epoch": 1.59853249475891,
      "grad_norm": 1.2778831720352173,
      "learning_rate": 3.9950209643605876e-05,
      "loss": 0.5746,
      "num_input_tokens_seen": 2004848,
      "step": 3050
    },
    {
      "epoch": 1.601153039832285,
      "grad_norm": 1.5115489959716797,
      "learning_rate": 4.001572327044025e-05,
      "loss": 0.5147,
      "num_input_tokens_seen": 2007920,
      "step": 3055
    },
    {
      "epoch": 1.6037735849056602,
      "grad_norm": 1.2122528553009033,
      "learning_rate": 4.0081236897274636e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 2011152,
      "step": 3060
    },
    {
      "epoch": 1.6063941299790356,
      "grad_norm": 1.4435789585113525,
      "learning_rate": 4.014675052410902e-05,
      "loss": 0.5602,
      "num_input_tokens_seen": 2013968,
      "step": 3065
    },
    {
      "epoch": 1.609014675052411,
      "grad_norm": 2.3309624195098877,
      "learning_rate": 4.0212264150943395e-05,
      "loss": 0.5073,
      "num_input_tokens_seen": 2017936,
      "step": 3070
    },
    {
      "epoch": 1.6116352201257862,
      "grad_norm": 0.6182041764259338,
      "learning_rate": 4.027777777777778e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 2020528,
      "step": 3075
    },
    {
      "epoch": 1.6142557651991614,
      "grad_norm": 1.307565689086914,
      "learning_rate": 4.034329140461216e-05,
      "loss": 0.6204,
      "num_input_tokens_seen": 2023792,
      "step": 3080
    },
    {
      "epoch": 1.6168763102725365,
      "grad_norm": 0.8717908263206482,
      "learning_rate": 4.0408805031446544e-05,
      "loss": 0.5078,
      "num_input_tokens_seen": 2027152,
      "step": 3085
    },
    {
      "epoch": 1.619496855345912,
      "grad_norm": 0.8633650541305542,
      "learning_rate": 4.047431865828093e-05,
      "loss": 0.3421,
      "num_input_tokens_seen": 2029968,
      "step": 3090
    },
    {
      "epoch": 1.6221174004192873,
      "grad_norm": 1.4575767517089844,
      "learning_rate": 4.0539832285115304e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 2033488,
      "step": 3095
    },
    {
      "epoch": 1.6247379454926625,
      "grad_norm": 1.6705766916275024,
      "learning_rate": 4.0605345911949687e-05,
      "loss": 0.5312,
      "num_input_tokens_seen": 2037136,
      "step": 3100
    },
    {
      "epoch": 1.6273584905660377,
      "grad_norm": 1.0354257822036743,
      "learning_rate": 4.067085953878407e-05,
      "loss": 0.4617,
      "num_input_tokens_seen": 2039888,
      "step": 3105
    },
    {
      "epoch": 1.629979035639413,
      "grad_norm": 1.1805715560913086,
      "learning_rate": 4.073637316561845e-05,
      "loss": 0.6456,
      "num_input_tokens_seen": 2043024,
      "step": 3110
    },
    {
      "epoch": 1.6325995807127882,
      "grad_norm": 1.1457258462905884,
      "learning_rate": 4.080188679245283e-05,
      "loss": 0.3562,
      "num_input_tokens_seen": 2047056,
      "step": 3115
    },
    {
      "epoch": 1.6352201257861636,
      "grad_norm": 1.6641687154769897,
      "learning_rate": 4.086740041928722e-05,
      "loss": 0.5273,
      "num_input_tokens_seen": 2050000,
      "step": 3120
    },
    {
      "epoch": 1.6378406708595388,
      "grad_norm": 0.9415584206581116,
      "learning_rate": 4.0932914046121595e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 2054032,
      "step": 3125
    },
    {
      "epoch": 1.640461215932914,
      "grad_norm": 1.4511919021606445,
      "learning_rate": 4.099842767295597e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 2057072,
      "step": 3130
    },
    {
      "epoch": 1.6430817610062893,
      "grad_norm": 0.8433658480644226,
      "learning_rate": 4.106394129979036e-05,
      "loss": 0.4452,
      "num_input_tokens_seen": 2060656,
      "step": 3135
    },
    {
      "epoch": 1.6457023060796647,
      "grad_norm": 2.4026052951812744,
      "learning_rate": 4.112945492662474e-05,
      "loss": 0.5115,
      "num_input_tokens_seen": 2063568,
      "step": 3140
    },
    {
      "epoch": 1.64832285115304,
      "grad_norm": 1.3429535627365112,
      "learning_rate": 4.119496855345912e-05,
      "loss": 0.3865,
      "num_input_tokens_seen": 2066960,
      "step": 3145
    },
    {
      "epoch": 1.650943396226415,
      "grad_norm": 0.8712383508682251,
      "learning_rate": 4.1260482180293504e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 2070096,
      "step": 3150
    },
    {
      "epoch": 1.6535639412997902,
      "grad_norm": 0.8716954588890076,
      "learning_rate": 4.132599580712788e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 2073072,
      "step": 3155
    },
    {
      "epoch": 1.6561844863731656,
      "grad_norm": 0.8913594484329224,
      "learning_rate": 4.139150943396227e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 2075632,
      "step": 3160
    },
    {
      "epoch": 1.658805031446541,
      "grad_norm": 1.4996721744537354,
      "learning_rate": 4.1457023060796646e-05,
      "loss": 0.3975,
      "num_input_tokens_seen": 2079024,
      "step": 3165
    },
    {
      "epoch": 1.6614255765199162,
      "grad_norm": 1.0098106861114502,
      "learning_rate": 4.152253668763103e-05,
      "loss": 0.6084,
      "num_input_tokens_seen": 2081744,
      "step": 3170
    },
    {
      "epoch": 1.6640461215932913,
      "grad_norm": 0.7177658677101135,
      "learning_rate": 4.158805031446541e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 2084784,
      "step": 3175
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.9701419472694397,
      "learning_rate": 4.1653563941299795e-05,
      "loss": 0.4022,
      "num_input_tokens_seen": 2087984,
      "step": 3180
    },
    {
      "epoch": 1.669287211740042,
      "grad_norm": 0.8129101991653442,
      "learning_rate": 4.171907756813417e-05,
      "loss": 0.4352,
      "num_input_tokens_seen": 2091344,
      "step": 3185
    },
    {
      "epoch": 1.6719077568134173,
      "grad_norm": 0.6051977872848511,
      "learning_rate": 4.1784591194968555e-05,
      "loss": 0.4085,
      "num_input_tokens_seen": 2095312,
      "step": 3190
    },
    {
      "epoch": 1.6745283018867925,
      "grad_norm": 0.8479169011116028,
      "learning_rate": 4.185010482180294e-05,
      "loss": 0.3947,
      "num_input_tokens_seen": 2098832,
      "step": 3195
    },
    {
      "epoch": 1.6771488469601676,
      "grad_norm": 1.5010048151016235,
      "learning_rate": 4.1915618448637314e-05,
      "loss": 0.5665,
      "num_input_tokens_seen": 2101168,
      "step": 3200
    },
    {
      "epoch": 1.679769392033543,
      "grad_norm": 1.5457565784454346,
      "learning_rate": 4.1981132075471704e-05,
      "loss": 0.6716,
      "num_input_tokens_seen": 2104816,
      "step": 3205
    },
    {
      "epoch": 1.6823899371069182,
      "grad_norm": 1.4708168506622314,
      "learning_rate": 4.204664570230608e-05,
      "loss": 0.4636,
      "num_input_tokens_seen": 2107728,
      "step": 3210
    },
    {
      "epoch": 1.6850104821802936,
      "grad_norm": 2.5884079933166504,
      "learning_rate": 4.2112159329140463e-05,
      "loss": 0.6681,
      "num_input_tokens_seen": 2110992,
      "step": 3215
    },
    {
      "epoch": 1.6876310272536688,
      "grad_norm": 1.0902680158615112,
      "learning_rate": 4.2177672955974846e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 2113840,
      "step": 3220
    },
    {
      "epoch": 1.690251572327044,
      "grad_norm": 1.0863803625106812,
      "learning_rate": 4.224318658280922e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 2116784,
      "step": 3225
    },
    {
      "epoch": 1.6928721174004193,
      "grad_norm": 0.8332738876342773,
      "learning_rate": 4.2308700209643606e-05,
      "loss": 0.6238,
      "num_input_tokens_seen": 2119696,
      "step": 3230
    },
    {
      "epoch": 1.6954926624737947,
      "grad_norm": 1.2924506664276123,
      "learning_rate": 4.237421383647799e-05,
      "loss": 0.531,
      "num_input_tokens_seen": 2122608,
      "step": 3235
    },
    {
      "epoch": 1.6981132075471699,
      "grad_norm": 1.3109737634658813,
      "learning_rate": 4.243972746331237e-05,
      "loss": 0.4474,
      "num_input_tokens_seen": 2125808,
      "step": 3240
    },
    {
      "epoch": 1.700733752620545,
      "grad_norm": 1.0331168174743652,
      "learning_rate": 4.2505241090146755e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 2129584,
      "step": 3245
    },
    {
      "epoch": 1.7033542976939202,
      "grad_norm": 1.198137879371643,
      "learning_rate": 4.257075471698113e-05,
      "loss": 0.4304,
      "num_input_tokens_seen": 2132208,
      "step": 3250
    },
    {
      "epoch": 1.7059748427672956,
      "grad_norm": 0.8305608630180359,
      "learning_rate": 4.2636268343815514e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 2135632,
      "step": 3255
    },
    {
      "epoch": 1.708595387840671,
      "grad_norm": 1.303464651107788,
      "learning_rate": 4.27017819706499e-05,
      "loss": 0.6695,
      "num_input_tokens_seen": 2139760,
      "step": 3260
    },
    {
      "epoch": 1.7112159329140462,
      "grad_norm": 1.2149865627288818,
      "learning_rate": 4.276729559748428e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 2143280,
      "step": 3265
    },
    {
      "epoch": 1.7138364779874213,
      "grad_norm": 0.8523284196853638,
      "learning_rate": 4.283280922431866e-05,
      "loss": 0.4259,
      "num_input_tokens_seen": 2145648,
      "step": 3270
    },
    {
      "epoch": 1.7164570230607965,
      "grad_norm": 1.2453815937042236,
      "learning_rate": 4.289832285115305e-05,
      "loss": 0.4551,
      "num_input_tokens_seen": 2149040,
      "step": 3275
    },
    {
      "epoch": 1.719077568134172,
      "grad_norm": 0.7604999542236328,
      "learning_rate": 4.296383647798742e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 2151792,
      "step": 3280
    },
    {
      "epoch": 1.7216981132075473,
      "grad_norm": 1.0474059581756592,
      "learning_rate": 4.3029350104821806e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 2155600,
      "step": 3285
    },
    {
      "epoch": 1.7243186582809225,
      "grad_norm": 0.7748103141784668,
      "learning_rate": 4.309486373165619e-05,
      "loss": 0.3938,
      "num_input_tokens_seen": 2159088,
      "step": 3290
    },
    {
      "epoch": 1.7269392033542976,
      "grad_norm": 1.2993237972259521,
      "learning_rate": 4.3160377358490565e-05,
      "loss": 0.3509,
      "num_input_tokens_seen": 2162800,
      "step": 3295
    },
    {
      "epoch": 1.7295597484276728,
      "grad_norm": 1.5319031476974487,
      "learning_rate": 4.322589098532495e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 2165872,
      "step": 3300
    },
    {
      "epoch": 1.7321802935010482,
      "grad_norm": 1.055769920349121,
      "learning_rate": 4.329140461215933e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 2169456,
      "step": 3305
    },
    {
      "epoch": 1.7348008385744236,
      "grad_norm": 0.9534486532211304,
      "learning_rate": 4.335691823899371e-05,
      "loss": 0.4633,
      "num_input_tokens_seen": 2172368,
      "step": 3310
    },
    {
      "epoch": 1.7374213836477987,
      "grad_norm": 0.7563013434410095,
      "learning_rate": 4.34224318658281e-05,
      "loss": 0.5658,
      "num_input_tokens_seen": 2175504,
      "step": 3315
    },
    {
      "epoch": 1.740041928721174,
      "grad_norm": 1.0940120220184326,
      "learning_rate": 4.3487945492662474e-05,
      "loss": 0.476,
      "num_input_tokens_seen": 2179440,
      "step": 3320
    },
    {
      "epoch": 1.7426624737945493,
      "grad_norm": 1.4431226253509521,
      "learning_rate": 4.355345911949686e-05,
      "loss": 0.5656,
      "num_input_tokens_seen": 2181904,
      "step": 3325
    },
    {
      "epoch": 1.7452830188679245,
      "grad_norm": 1.1267253160476685,
      "learning_rate": 4.361897274633124e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 2184816,
      "step": 3330
    },
    {
      "epoch": 1.7479035639412999,
      "grad_norm": 1.325600504875183,
      "learning_rate": 4.368448637316562e-05,
      "loss": 0.5502,
      "num_input_tokens_seen": 2187920,
      "step": 3335
    },
    {
      "epoch": 1.750524109014675,
      "grad_norm": 0.6263946294784546,
      "learning_rate": 4.375e-05,
      "loss": 0.4816,
      "num_input_tokens_seen": 2191216,
      "step": 3340
    },
    {
      "epoch": 1.7531446540880502,
      "grad_norm": 1.3308621644973755,
      "learning_rate": 4.381551362683438e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 2193744,
      "step": 3345
    },
    {
      "epoch": 1.7557651991614256,
      "grad_norm": 0.8700528740882874,
      "learning_rate": 4.3881027253668766e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 2196368,
      "step": 3350
    },
    {
      "epoch": 1.758385744234801,
      "grad_norm": 0.7918739318847656,
      "learning_rate": 4.394654088050314e-05,
      "loss": 0.3615,
      "num_input_tokens_seen": 2199632,
      "step": 3355
    },
    {
      "epoch": 1.7610062893081762,
      "grad_norm": 0.7617687582969666,
      "learning_rate": 4.401205450733753e-05,
      "loss": 0.4848,
      "num_input_tokens_seen": 2202704,
      "step": 3360
    },
    {
      "epoch": 1.7636268343815513,
      "grad_norm": 0.9885025024414062,
      "learning_rate": 4.407756813417191e-05,
      "loss": 0.6286,
      "num_input_tokens_seen": 2205616,
      "step": 3365
    },
    {
      "epoch": 1.7662473794549265,
      "grad_norm": 0.6348594427108765,
      "learning_rate": 4.414308176100629e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 2209008,
      "step": 3370
    },
    {
      "epoch": 1.7688679245283019,
      "grad_norm": 0.948408305644989,
      "learning_rate": 4.4208595387840674e-05,
      "loss": 0.4054,
      "num_input_tokens_seen": 2211824,
      "step": 3375
    },
    {
      "epoch": 1.7714884696016773,
      "grad_norm": 0.620255708694458,
      "learning_rate": 4.427410901467505e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 2215632,
      "step": 3380
    },
    {
      "epoch": 1.7741090146750524,
      "grad_norm": 1.1833486557006836,
      "learning_rate": 4.433962264150944e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 2218512,
      "step": 3385
    },
    {
      "epoch": 1.7767295597484276,
      "grad_norm": 1.0561902523040771,
      "learning_rate": 4.440513626834382e-05,
      "loss": 0.3741,
      "num_input_tokens_seen": 2221776,
      "step": 3390
    },
    {
      "epoch": 1.7793501048218028,
      "grad_norm": 0.5751020908355713,
      "learning_rate": 4.44706498951782e-05,
      "loss": 0.3794,
      "num_input_tokens_seen": 2224784,
      "step": 3395
    },
    {
      "epoch": 1.7819706498951782,
      "grad_norm": 1.6191028356552124,
      "learning_rate": 4.453616352201258e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 2227408,
      "step": 3400
    },
    {
      "epoch": 1.7845911949685536,
      "grad_norm": 1.2426648139953613,
      "learning_rate": 4.460167714884696e-05,
      "loss": 0.4317,
      "num_input_tokens_seen": 2230064,
      "step": 3405
    },
    {
      "epoch": 1.7872117400419287,
      "grad_norm": 0.9830101132392883,
      "learning_rate": 4.466719077568134e-05,
      "loss": 0.3442,
      "num_input_tokens_seen": 2234288,
      "step": 3410
    },
    {
      "epoch": 1.789832285115304,
      "grad_norm": 0.5475638508796692,
      "learning_rate": 4.4732704402515725e-05,
      "loss": 0.3477,
      "num_input_tokens_seen": 2237072,
      "step": 3415
    },
    {
      "epoch": 1.7924528301886793,
      "grad_norm": 0.6749818325042725,
      "learning_rate": 4.479821802935011e-05,
      "loss": 0.3449,
      "num_input_tokens_seen": 2240016,
      "step": 3420
    },
    {
      "epoch": 1.7950733752620545,
      "grad_norm": 0.7662815451622009,
      "learning_rate": 4.4863731656184485e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 2242800,
      "step": 3425
    },
    {
      "epoch": 1.7976939203354299,
      "grad_norm": 1.1825385093688965,
      "learning_rate": 4.4929245283018875e-05,
      "loss": 0.4742,
      "num_input_tokens_seen": 2245680,
      "step": 3430
    },
    {
      "epoch": 1.800314465408805,
      "grad_norm": 1.1253209114074707,
      "learning_rate": 4.499475890985325e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 2248272,
      "step": 3435
    },
    {
      "epoch": 1.8029350104821802,
      "grad_norm": 1.406171202659607,
      "learning_rate": 4.5060272536687634e-05,
      "loss": 0.3095,
      "num_input_tokens_seen": 2251376,
      "step": 3440
    },
    {
      "epoch": 1.8055555555555556,
      "grad_norm": 2.030001401901245,
      "learning_rate": 4.512578616352202e-05,
      "loss": 0.5099,
      "num_input_tokens_seen": 2254256,
      "step": 3445
    },
    {
      "epoch": 1.808176100628931,
      "grad_norm": 1.2542152404785156,
      "learning_rate": 4.519129979035639e-05,
      "loss": 0.5393,
      "num_input_tokens_seen": 2257104,
      "step": 3450
    },
    {
      "epoch": 1.8107966457023061,
      "grad_norm": 1.2001535892486572,
      "learning_rate": 4.525681341719078e-05,
      "loss": 0.4933,
      "num_input_tokens_seen": 2259920,
      "step": 3455
    },
    {
      "epoch": 1.8134171907756813,
      "grad_norm": 0.7913087010383606,
      "learning_rate": 4.532232704402516e-05,
      "loss": 0.3662,
      "num_input_tokens_seen": 2262736,
      "step": 3460
    },
    {
      "epoch": 1.8160377358490565,
      "grad_norm": 0.46319377422332764,
      "learning_rate": 4.5387840670859536e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 2266864,
      "step": 3465
    },
    {
      "epoch": 1.8186582809224319,
      "grad_norm": 1.9051378965377808,
      "learning_rate": 4.5453354297693926e-05,
      "loss": 0.6449,
      "num_input_tokens_seen": 2269584,
      "step": 3470
    },
    {
      "epoch": 1.8212788259958073,
      "grad_norm": 0.6460825204849243,
      "learning_rate": 4.55188679245283e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 2272720,
      "step": 3475
    },
    {
      "epoch": 1.8238993710691824,
      "grad_norm": 0.7731202244758606,
      "learning_rate": 4.5584381551362685e-05,
      "loss": 0.5554,
      "num_input_tokens_seen": 2276880,
      "step": 3480
    },
    {
      "epoch": 1.8265199161425576,
      "grad_norm": 3.3764760494232178,
      "learning_rate": 4.564989517819707e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 2280016,
      "step": 3485
    },
    {
      "epoch": 1.8291404612159328,
      "grad_norm": 0.9064461588859558,
      "learning_rate": 4.571540880503145e-05,
      "loss": 0.4464,
      "num_input_tokens_seen": 2282640,
      "step": 3490
    },
    {
      "epoch": 1.8317610062893082,
      "grad_norm": 1.4080784320831299,
      "learning_rate": 4.578092243186583e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 2286032,
      "step": 3495
    },
    {
      "epoch": 1.8343815513626835,
      "grad_norm": 0.8067936897277832,
      "learning_rate": 4.584643605870021e-05,
      "loss": 0.5222,
      "num_input_tokens_seen": 2289136,
      "step": 3500
    },
    {
      "epoch": 1.8370020964360587,
      "grad_norm": 0.9298325777053833,
      "learning_rate": 4.5911949685534594e-05,
      "loss": 0.3612,
      "num_input_tokens_seen": 2292944,
      "step": 3505
    },
    {
      "epoch": 1.8396226415094339,
      "grad_norm": 0.9098219871520996,
      "learning_rate": 4.5977463312368977e-05,
      "loss": 0.4369,
      "num_input_tokens_seen": 2296240,
      "step": 3510
    },
    {
      "epoch": 1.8422431865828093,
      "grad_norm": 1.4328676462173462,
      "learning_rate": 4.604297693920336e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 2299088,
      "step": 3515
    },
    {
      "epoch": 1.8448637316561844,
      "grad_norm": 1.3096126317977905,
      "learning_rate": 4.6108490566037736e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 2302736,
      "step": 3520
    },
    {
      "epoch": 1.8474842767295598,
      "grad_norm": 0.8984668850898743,
      "learning_rate": 4.617400419287212e-05,
      "loss": 0.5232,
      "num_input_tokens_seen": 2306640,
      "step": 3525
    },
    {
      "epoch": 1.850104821802935,
      "grad_norm": 0.9730688333511353,
      "learning_rate": 4.62395178197065e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 2309840,
      "step": 3530
    },
    {
      "epoch": 1.8527253668763102,
      "grad_norm": 1.6916528940200806,
      "learning_rate": 4.630503144654088e-05,
      "loss": 0.4911,
      "num_input_tokens_seen": 2312208,
      "step": 3535
    },
    {
      "epoch": 1.8553459119496856,
      "grad_norm": 2.4578135013580322,
      "learning_rate": 4.637054507337527e-05,
      "loss": 0.574,
      "num_input_tokens_seen": 2315088,
      "step": 3540
    },
    {
      "epoch": 1.857966457023061,
      "grad_norm": 1.4763623476028442,
      "learning_rate": 4.6436058700209645e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 2318224,
      "step": 3545
    },
    {
      "epoch": 1.8605870020964361,
      "grad_norm": 0.524951696395874,
      "learning_rate": 4.650157232704403e-05,
      "loss": 0.3959,
      "num_input_tokens_seen": 2322800,
      "step": 3550
    },
    {
      "epoch": 1.8632075471698113,
      "grad_norm": 1.3985940217971802,
      "learning_rate": 4.656708595387841e-05,
      "loss": 0.3209,
      "num_input_tokens_seen": 2329008,
      "step": 3555
    },
    {
      "epoch": 1.8658280922431865,
      "grad_norm": 1.7476978302001953,
      "learning_rate": 4.663259958071279e-05,
      "loss": 0.6168,
      "num_input_tokens_seen": 2331344,
      "step": 3560
    },
    {
      "epoch": 1.8684486373165619,
      "grad_norm": 0.9636356830596924,
      "learning_rate": 4.669811320754717e-05,
      "loss": 0.4587,
      "num_input_tokens_seen": 2334352,
      "step": 3565
    },
    {
      "epoch": 1.8710691823899372,
      "grad_norm": 0.6869973540306091,
      "learning_rate": 4.676362683438155e-05,
      "loss": 0.5249,
      "num_input_tokens_seen": 2338512,
      "step": 3570
    },
    {
      "epoch": 1.8736897274633124,
      "grad_norm": 0.6853863596916199,
      "learning_rate": 4.6829140461215936e-05,
      "loss": 0.5512,
      "num_input_tokens_seen": 2342224,
      "step": 3575
    },
    {
      "epoch": 1.8763102725366876,
      "grad_norm": 2.167863368988037,
      "learning_rate": 4.689465408805032e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 2345808,
      "step": 3580
    },
    {
      "epoch": 1.8789308176100628,
      "grad_norm": 0.609732449054718,
      "learning_rate": 4.69601677148847e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 2348560,
      "step": 3585
    },
    {
      "epoch": 1.8815513626834381,
      "grad_norm": 0.8691074848175049,
      "learning_rate": 4.702568134171908e-05,
      "loss": 0.3608,
      "num_input_tokens_seen": 2351824,
      "step": 3590
    },
    {
      "epoch": 1.8841719077568135,
      "grad_norm": 1.184690237045288,
      "learning_rate": 4.709119496855346e-05,
      "loss": 0.3704,
      "num_input_tokens_seen": 2355696,
      "step": 3595
    },
    {
      "epoch": 1.8867924528301887,
      "grad_norm": 0.8486743569374084,
      "learning_rate": 4.7156708595387845e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 2358352,
      "step": 3600
    },
    {
      "epoch": 1.8894129979035639,
      "grad_norm": 0.6490517258644104,
      "learning_rate": 4.722222222222222e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 2361776,
      "step": 3605
    },
    {
      "epoch": 1.892033542976939,
      "grad_norm": 1.053784966468811,
      "learning_rate": 4.728773584905661e-05,
      "loss": 0.5221,
      "num_input_tokens_seen": 2365264,
      "step": 3610
    },
    {
      "epoch": 1.8946540880503144,
      "grad_norm": 1.3080251216888428,
      "learning_rate": 4.735324947589099e-05,
      "loss": 0.6779,
      "num_input_tokens_seen": 2368080,
      "step": 3615
    },
    {
      "epoch": 1.8972746331236898,
      "grad_norm": 0.7440404295921326,
      "learning_rate": 4.7418763102725364e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 2371696,
      "step": 3620
    },
    {
      "epoch": 1.899895178197065,
      "grad_norm": 1.0299056768417358,
      "learning_rate": 4.7484276729559753e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 2374160,
      "step": 3625
    },
    {
      "epoch": 1.9025157232704402,
      "grad_norm": 1.1182126998901367,
      "learning_rate": 4.754979035639413e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 2376656,
      "step": 3630
    },
    {
      "epoch": 1.9051362683438156,
      "grad_norm": 0.867205798625946,
      "learning_rate": 4.761530398322851e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 2379856,
      "step": 3635
    },
    {
      "epoch": 1.9077568134171907,
      "grad_norm": 0.8277135491371155,
      "learning_rate": 4.7680817610062896e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 2382736,
      "step": 3640
    },
    {
      "epoch": 1.9103773584905661,
      "grad_norm": 1.153232216835022,
      "learning_rate": 4.774633123689728e-05,
      "loss": 0.4241,
      "num_input_tokens_seen": 2386672,
      "step": 3645
    },
    {
      "epoch": 1.9129979035639413,
      "grad_norm": 0.8728160262107849,
      "learning_rate": 4.7811844863731655e-05,
      "loss": 0.3796,
      "num_input_tokens_seen": 2389296,
      "step": 3650
    },
    {
      "epoch": 1.9156184486373165,
      "grad_norm": 0.962847113609314,
      "learning_rate": 4.787735849056604e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 2392080,
      "step": 3655
    },
    {
      "epoch": 1.9182389937106918,
      "grad_norm": 0.5763393640518188,
      "learning_rate": 4.794287211740042e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 2395952,
      "step": 3660
    },
    {
      "epoch": 1.9208595387840672,
      "grad_norm": 0.5958591103553772,
      "learning_rate": 4.8008385744234804e-05,
      "loss": 0.3813,
      "num_input_tokens_seen": 2398512,
      "step": 3665
    },
    {
      "epoch": 1.9234800838574424,
      "grad_norm": 0.8564649224281311,
      "learning_rate": 4.807389937106919e-05,
      "loss": 0.524,
      "num_input_tokens_seen": 2401904,
      "step": 3670
    },
    {
      "epoch": 1.9261006289308176,
      "grad_norm": 0.671838641166687,
      "learning_rate": 4.8139412997903564e-05,
      "loss": 0.4899,
      "num_input_tokens_seen": 2405584,
      "step": 3675
    },
    {
      "epoch": 1.9287211740041927,
      "grad_norm": 0.5613347291946411,
      "learning_rate": 4.8204926624737954e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 2409712,
      "step": 3680
    },
    {
      "epoch": 1.9313417190775681,
      "grad_norm": 0.9844335317611694,
      "learning_rate": 4.827044025157233e-05,
      "loss": 0.6036,
      "num_input_tokens_seen": 2412240,
      "step": 3685
    },
    {
      "epoch": 1.9339622641509435,
      "grad_norm": 0.7533885836601257,
      "learning_rate": 4.8335953878406706e-05,
      "loss": 0.3927,
      "num_input_tokens_seen": 2415056,
      "step": 3690
    },
    {
      "epoch": 1.9365828092243187,
      "grad_norm": 1.234908938407898,
      "learning_rate": 4.8401467505241096e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 2419184,
      "step": 3695
    },
    {
      "epoch": 1.9392033542976939,
      "grad_norm": 1.487202763557434,
      "learning_rate": 4.846698113207547e-05,
      "loss": 0.503,
      "num_input_tokens_seen": 2422512,
      "step": 3700
    },
    {
      "epoch": 1.941823899371069,
      "grad_norm": 1.6397887468338013,
      "learning_rate": 4.8532494758909855e-05,
      "loss": 0.4436,
      "num_input_tokens_seen": 2425936,
      "step": 3705
    },
    {
      "epoch": 1.9444444444444444,
      "grad_norm": 1.264580249786377,
      "learning_rate": 4.859800838574424e-05,
      "loss": 0.5833,
      "num_input_tokens_seen": 2428944,
      "step": 3710
    },
    {
      "epoch": 1.9470649895178198,
      "grad_norm": 0.8766135573387146,
      "learning_rate": 4.8663522012578615e-05,
      "loss": 0.365,
      "num_input_tokens_seen": 2432080,
      "step": 3715
    },
    {
      "epoch": 1.949685534591195,
      "grad_norm": 0.9716737866401672,
      "learning_rate": 4.8729035639413e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 2434864,
      "step": 3720
    },
    {
      "epoch": 1.9523060796645701,
      "grad_norm": 1.0023905038833618,
      "learning_rate": 4.879454926624738e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 2437104,
      "step": 3725
    },
    {
      "epoch": 1.9549266247379455,
      "grad_norm": 1.3583959341049194,
      "learning_rate": 4.8860062893081764e-05,
      "loss": 0.6277,
      "num_input_tokens_seen": 2439792,
      "step": 3730
    },
    {
      "epoch": 1.9575471698113207,
      "grad_norm": 0.5970690846443176,
      "learning_rate": 4.892557651991615e-05,
      "loss": 0.6071,
      "num_input_tokens_seen": 2442320,
      "step": 3735
    },
    {
      "epoch": 1.960167714884696,
      "grad_norm": 0.7311194539070129,
      "learning_rate": 4.899109014675053e-05,
      "loss": 0.3889,
      "num_input_tokens_seen": 2445136,
      "step": 3740
    },
    {
      "epoch": 1.9627882599580713,
      "grad_norm": 0.7701296210289001,
      "learning_rate": 4.9056603773584906e-05,
      "loss": 0.44,
      "num_input_tokens_seen": 2448080,
      "step": 3745
    },
    {
      "epoch": 1.9654088050314464,
      "grad_norm": 0.7176717519760132,
      "learning_rate": 4.912211740041929e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 2451376,
      "step": 3750
    },
    {
      "epoch": 1.9680293501048218,
      "grad_norm": 0.8987816572189331,
      "learning_rate": 4.918763102725367e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 2454672,
      "step": 3755
    },
    {
      "epoch": 1.9706498951781972,
      "grad_norm": 0.9972012639045715,
      "learning_rate": 4.925314465408805e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 2457712,
      "step": 3760
    },
    {
      "epoch": 1.9732704402515724,
      "grad_norm": 1.0456798076629639,
      "learning_rate": 4.931865828092244e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 2461936,
      "step": 3765
    },
    {
      "epoch": 1.9758909853249476,
      "grad_norm": 0.9483128190040588,
      "learning_rate": 4.9384171907756815e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 2465296,
      "step": 3770
    },
    {
      "epoch": 1.9785115303983227,
      "grad_norm": 1.0111498832702637,
      "learning_rate": 4.944968553459119e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 2469872,
      "step": 3775
    },
    {
      "epoch": 1.9811320754716981,
      "grad_norm": 0.9154354333877563,
      "learning_rate": 4.951519916142558e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 2473840,
      "step": 3780
    },
    {
      "epoch": 1.9837526205450735,
      "grad_norm": 0.971025288105011,
      "learning_rate": 4.958071278825996e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 2476944,
      "step": 3785
    },
    {
      "epoch": 1.9863731656184487,
      "grad_norm": 0.9575963616371155,
      "learning_rate": 4.964622641509434e-05,
      "loss": 0.4406,
      "num_input_tokens_seen": 2479792,
      "step": 3790
    },
    {
      "epoch": 1.9889937106918238,
      "grad_norm": 1.1203691959381104,
      "learning_rate": 4.9711740041928724e-05,
      "loss": 0.5654,
      "num_input_tokens_seen": 2484464,
      "step": 3795
    },
    {
      "epoch": 1.991614255765199,
      "grad_norm": 1.0854727029800415,
      "learning_rate": 4.977725366876311e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 2487312,
      "step": 3800
    },
    {
      "epoch": 1.9942348008385744,
      "grad_norm": 1.0891664028167725,
      "learning_rate": 4.984276729559749e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 2490288,
      "step": 3805
    },
    {
      "epoch": 1.9968553459119498,
      "grad_norm": 0.8696821331977844,
      "learning_rate": 4.9908280922431866e-05,
      "loss": 0.4912,
      "num_input_tokens_seen": 2494128,
      "step": 3810
    },
    {
      "epoch": 1.999475890985325,
      "grad_norm": 0.8102225065231323,
      "learning_rate": 4.997379454926625e-05,
      "loss": 0.4289,
      "num_input_tokens_seen": 2496848,
      "step": 3815
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.47038066387176514,
      "eval_runtime": 15.1757,
      "eval_samples_per_second": 55.879,
      "eval_steps_per_second": 13.97,
      "num_input_tokens_seen": 2497016,
      "step": 3816
    },
    {
      "epoch": 2.0020964360587,
      "grad_norm": 0.8429184556007385,
      "learning_rate": 4.999999905865133e-05,
      "loss": 0.3389,
      "num_input_tokens_seen": 2499256,
      "step": 3820
    },
    {
      "epoch": 2.0047169811320753,
      "grad_norm": 0.9440101385116577,
      "learning_rate": 4.999999330596526e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 2502200,
      "step": 3825
    },
    {
      "epoch": 2.007337526205451,
      "grad_norm": 1.2819377183914185,
      "learning_rate": 4.9999982323565806e-05,
      "loss": 0.5725,
      "num_input_tokens_seen": 2505080,
      "step": 3830
    },
    {
      "epoch": 2.009958071278826,
      "grad_norm": 0.9235860705375671,
      "learning_rate": 4.999996611145527e-05,
      "loss": 0.4507,
      "num_input_tokens_seen": 2508216,
      "step": 3835
    },
    {
      "epoch": 2.0125786163522013,
      "grad_norm": 0.798232913017273,
      "learning_rate": 4.9999944669637024e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 2511128,
      "step": 3840
    },
    {
      "epoch": 2.0151991614255764,
      "grad_norm": 0.8014755845069885,
      "learning_rate": 4.999991799811557e-05,
      "loss": 0.4649,
      "num_input_tokens_seen": 2514776,
      "step": 3845
    },
    {
      "epoch": 2.0178197064989516,
      "grad_norm": 1.5800092220306396,
      "learning_rate": 4.99998860968965e-05,
      "loss": 0.5657,
      "num_input_tokens_seen": 2517528,
      "step": 3850
    },
    {
      "epoch": 2.020440251572327,
      "grad_norm": 0.6533426642417908,
      "learning_rate": 4.999984896598646e-05,
      "loss": 0.4971,
      "num_input_tokens_seen": 2523032,
      "step": 3855
    },
    {
      "epoch": 2.0230607966457024,
      "grad_norm": 0.7745879292488098,
      "learning_rate": 4.999980660539323e-05,
      "loss": 0.4144,
      "num_input_tokens_seen": 2526200,
      "step": 3860
    },
    {
      "epoch": 2.0256813417190775,
      "grad_norm": 1.1020114421844482,
      "learning_rate": 4.999975901512567e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 2532760,
      "step": 3865
    },
    {
      "epoch": 2.0283018867924527,
      "grad_norm": 0.8689053058624268,
      "learning_rate": 4.999970619519374e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 2535640,
      "step": 3870
    },
    {
      "epoch": 2.030922431865828,
      "grad_norm": 0.7562381029129028,
      "learning_rate": 4.999964814560849e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 2539288,
      "step": 3875
    },
    {
      "epoch": 2.0335429769392035,
      "grad_norm": 1.4872406721115112,
      "learning_rate": 4.999958486638205e-05,
      "loss": 0.4996,
      "num_input_tokens_seen": 2541880,
      "step": 3880
    },
    {
      "epoch": 2.0361635220125787,
      "grad_norm": 0.8948652744293213,
      "learning_rate": 4.999951635752767e-05,
      "loss": 0.4264,
      "num_input_tokens_seen": 2545080,
      "step": 3885
    },
    {
      "epoch": 2.038784067085954,
      "grad_norm": 0.823993444442749,
      "learning_rate": 4.999944261905968e-05,
      "loss": 0.3304,
      "num_input_tokens_seen": 2548376,
      "step": 3890
    },
    {
      "epoch": 2.041404612159329,
      "grad_norm": 0.8851125836372375,
      "learning_rate": 4.99993636509935e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 2550616,
      "step": 3895
    },
    {
      "epoch": 2.0440251572327046,
      "grad_norm": 0.9696165323257446,
      "learning_rate": 4.999927945334565e-05,
      "loss": 0.5286,
      "num_input_tokens_seen": 2555000,
      "step": 3900
    },
    {
      "epoch": 2.04664570230608,
      "grad_norm": 1.4545605182647705,
      "learning_rate": 4.9999190026133745e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 2558328,
      "step": 3905
    },
    {
      "epoch": 2.049266247379455,
      "grad_norm": 1.112327218055725,
      "learning_rate": 4.9999095369376506e-05,
      "loss": 0.5278,
      "num_input_tokens_seen": 2561592,
      "step": 3910
    },
    {
      "epoch": 2.05188679245283,
      "grad_norm": 0.7287724018096924,
      "learning_rate": 4.99989954830937e-05,
      "loss": 0.3109,
      "num_input_tokens_seen": 2564472,
      "step": 3915
    },
    {
      "epoch": 2.0545073375262053,
      "grad_norm": 1.8660250902175903,
      "learning_rate": 4.9998890367306255e-05,
      "loss": 0.4202,
      "num_input_tokens_seen": 2567352,
      "step": 3920
    },
    {
      "epoch": 2.057127882599581,
      "grad_norm": 0.9334644079208374,
      "learning_rate": 4.999878002203614e-05,
      "loss": 0.3268,
      "num_input_tokens_seen": 2571448,
      "step": 3925
    },
    {
      "epoch": 2.059748427672956,
      "grad_norm": 1.4282631874084473,
      "learning_rate": 4.999866444730645e-05,
      "loss": 0.3625,
      "num_input_tokens_seen": 2574264,
      "step": 3930
    },
    {
      "epoch": 2.0623689727463312,
      "grad_norm": 1.337568998336792,
      "learning_rate": 4.999854364314136e-05,
      "loss": 0.5308,
      "num_input_tokens_seen": 2577592,
      "step": 3935
    },
    {
      "epoch": 2.0649895178197064,
      "grad_norm": 1.4006571769714355,
      "learning_rate": 4.999841760956614e-05,
      "loss": 0.4623,
      "num_input_tokens_seen": 2581624,
      "step": 3940
    },
    {
      "epoch": 2.0676100628930816,
      "grad_norm": 0.8792033791542053,
      "learning_rate": 4.999828634660715e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 2584888,
      "step": 3945
    },
    {
      "epoch": 2.070230607966457,
      "grad_norm": 1.7611385583877563,
      "learning_rate": 4.9998149854291856e-05,
      "loss": 0.4612,
      "num_input_tokens_seen": 2587672,
      "step": 3950
    },
    {
      "epoch": 2.0728511530398324,
      "grad_norm": 1.092012882232666,
      "learning_rate": 4.99980081326488e-05,
      "loss": 0.439,
      "num_input_tokens_seen": 2590136,
      "step": 3955
    },
    {
      "epoch": 2.0754716981132075,
      "grad_norm": 0.8344743847846985,
      "learning_rate": 4.999786118170763e-05,
      "loss": 0.428,
      "num_input_tokens_seen": 2593976,
      "step": 3960
    },
    {
      "epoch": 2.0780922431865827,
      "grad_norm": 1.1540946960449219,
      "learning_rate": 4.99977090014991e-05,
      "loss": 0.9195,
      "num_input_tokens_seen": 2597208,
      "step": 3965
    },
    {
      "epoch": 2.080712788259958,
      "grad_norm": 1.0468941926956177,
      "learning_rate": 4.9997551592055034e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 2600536,
      "step": 3970
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 0.851332426071167,
      "learning_rate": 4.999738895340836e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 2603608,
      "step": 3975
    },
    {
      "epoch": 2.0859538784067087,
      "grad_norm": 1.0525293350219727,
      "learning_rate": 4.999722108559311e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 2607288,
      "step": 3980
    },
    {
      "epoch": 2.088574423480084,
      "grad_norm": 0.9889012575149536,
      "learning_rate": 4.9997047988644385e-05,
      "loss": 0.5572,
      "num_input_tokens_seen": 2610392,
      "step": 3985
    },
    {
      "epoch": 2.091194968553459,
      "grad_norm": 0.808921217918396,
      "learning_rate": 4.99968696625984e-05,
      "loss": 0.5387,
      "num_input_tokens_seen": 2613208,
      "step": 3990
    },
    {
      "epoch": 2.0938155136268346,
      "grad_norm": 0.8648212552070618,
      "learning_rate": 4.999668610749246e-05,
      "loss": 0.5622,
      "num_input_tokens_seen": 2616472,
      "step": 3995
    },
    {
      "epoch": 2.0964360587002098,
      "grad_norm": 0.8454099297523499,
      "learning_rate": 4.999649732336498e-05,
      "loss": 0.4319,
      "num_input_tokens_seen": 2620504,
      "step": 4000
    },
    {
      "epoch": 2.099056603773585,
      "grad_norm": 0.975403368473053,
      "learning_rate": 4.999630331025542e-05,
      "loss": 0.5043,
      "num_input_tokens_seen": 2624632,
      "step": 4005
    },
    {
      "epoch": 2.10167714884696,
      "grad_norm": 1.1253719329833984,
      "learning_rate": 4.9996104068204385e-05,
      "loss": 0.4347,
      "num_input_tokens_seen": 2627608,
      "step": 4010
    },
    {
      "epoch": 2.1042976939203353,
      "grad_norm": 0.6056440472602844,
      "learning_rate": 4.999589959725355e-05,
      "loss": 0.3505,
      "num_input_tokens_seen": 2630232,
      "step": 4015
    },
    {
      "epoch": 2.106918238993711,
      "grad_norm": 0.7290267944335938,
      "learning_rate": 4.999568989744568e-05,
      "loss": 0.3729,
      "num_input_tokens_seen": 2633624,
      "step": 4020
    },
    {
      "epoch": 2.109538784067086,
      "grad_norm": 1.4998009204864502,
      "learning_rate": 4.999547496882466e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 2636280,
      "step": 4025
    },
    {
      "epoch": 2.1121593291404612,
      "grad_norm": 1.0073578357696533,
      "learning_rate": 4.999525481143543e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 2640152,
      "step": 4030
    },
    {
      "epoch": 2.1147798742138364,
      "grad_norm": 0.6814026832580566,
      "learning_rate": 4.999502942532406e-05,
      "loss": 0.4344,
      "num_input_tokens_seen": 2643800,
      "step": 4035
    },
    {
      "epoch": 2.1174004192872116,
      "grad_norm": 0.9525241255760193,
      "learning_rate": 4.999479881053769e-05,
      "loss": 0.4133,
      "num_input_tokens_seen": 2647608,
      "step": 4040
    },
    {
      "epoch": 2.120020964360587,
      "grad_norm": 0.9816739559173584,
      "learning_rate": 4.999456296712457e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 2650200,
      "step": 4045
    },
    {
      "epoch": 2.1226415094339623,
      "grad_norm": 0.9119827151298523,
      "learning_rate": 4.9994321895134025e-05,
      "loss": 0.5608,
      "num_input_tokens_seen": 2653784,
      "step": 4050
    },
    {
      "epoch": 2.1252620545073375,
      "grad_norm": 0.5650546550750732,
      "learning_rate": 4.999407559461649e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 2658104,
      "step": 4055
    },
    {
      "epoch": 2.1278825995807127,
      "grad_norm": 1.4439082145690918,
      "learning_rate": 4.999382406562349e-05,
      "loss": 0.3862,
      "num_input_tokens_seen": 2661080,
      "step": 4060
    },
    {
      "epoch": 2.130503144654088,
      "grad_norm": 0.823525607585907,
      "learning_rate": 4.999356730820765e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 2664312,
      "step": 4065
    },
    {
      "epoch": 2.1331236897274635,
      "grad_norm": 1.6031163930892944,
      "learning_rate": 4.999330532242266e-05,
      "loss": 0.4855,
      "num_input_tokens_seen": 2667704,
      "step": 4070
    },
    {
      "epoch": 2.1357442348008386,
      "grad_norm": 1.0796573162078857,
      "learning_rate": 4.999303810832333e-05,
      "loss": 0.364,
      "num_input_tokens_seen": 2671160,
      "step": 4075
    },
    {
      "epoch": 2.138364779874214,
      "grad_norm": 0.7700610756874084,
      "learning_rate": 4.999276566596557e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 2675544,
      "step": 4080
    },
    {
      "epoch": 2.140985324947589,
      "grad_norm": 1.204138994216919,
      "learning_rate": 4.9992487995406364e-05,
      "loss": 0.4501,
      "num_input_tokens_seen": 2678264,
      "step": 4085
    },
    {
      "epoch": 2.1436058700209646,
      "grad_norm": 1.090428352355957,
      "learning_rate": 4.9992205096703806e-05,
      "loss": 0.3918,
      "num_input_tokens_seen": 2684120,
      "step": 4090
    },
    {
      "epoch": 2.1462264150943398,
      "grad_norm": 1.5191525220870972,
      "learning_rate": 4.9991916969917054e-05,
      "loss": 0.6037,
      "num_input_tokens_seen": 2686872,
      "step": 4095
    },
    {
      "epoch": 2.148846960167715,
      "grad_norm": 0.966274619102478,
      "learning_rate": 4.99916236151064e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 2690456,
      "step": 4100
    },
    {
      "epoch": 2.15146750524109,
      "grad_norm": 0.9836441874504089,
      "learning_rate": 4.999132503233321e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 2692824,
      "step": 4105
    },
    {
      "epoch": 2.1540880503144653,
      "grad_norm": 1.1106511354446411,
      "learning_rate": 4.999102122165994e-05,
      "loss": 0.5383,
      "num_input_tokens_seen": 2695704,
      "step": 4110
    },
    {
      "epoch": 2.156708595387841,
      "grad_norm": 0.6694199442863464,
      "learning_rate": 4.9990712183150135e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 2699480,
      "step": 4115
    },
    {
      "epoch": 2.159329140461216,
      "grad_norm": 0.7963168025016785,
      "learning_rate": 4.9990397916868456e-05,
      "loss": 0.5051,
      "num_input_tokens_seen": 2703960,
      "step": 4120
    },
    {
      "epoch": 2.161949685534591,
      "grad_norm": 0.8698011636734009,
      "learning_rate": 4.999007842288064e-05,
      "loss": 0.386,
      "num_input_tokens_seen": 2706552,
      "step": 4125
    },
    {
      "epoch": 2.1645702306079664,
      "grad_norm": 0.9642242193222046,
      "learning_rate": 4.9989753701253514e-05,
      "loss": 0.4764,
      "num_input_tokens_seen": 2709368,
      "step": 4130
    },
    {
      "epoch": 2.1671907756813416,
      "grad_norm": 1.2504425048828125,
      "learning_rate": 4.998942375205502e-05,
      "loss": 0.4906,
      "num_input_tokens_seen": 2712248,
      "step": 4135
    },
    {
      "epoch": 2.169811320754717,
      "grad_norm": 0.9956188201904297,
      "learning_rate": 4.998908857535416e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 2715448,
      "step": 4140
    },
    {
      "epoch": 2.1724318658280923,
      "grad_norm": 0.5744521021842957,
      "learning_rate": 4.9988748171221065e-05,
      "loss": 0.3575,
      "num_input_tokens_seen": 2719064,
      "step": 4145
    },
    {
      "epoch": 2.1750524109014675,
      "grad_norm": 0.8713458776473999,
      "learning_rate": 4.998840253972695e-05,
      "loss": 0.4105,
      "num_input_tokens_seen": 2722488,
      "step": 4150
    },
    {
      "epoch": 2.1776729559748427,
      "grad_norm": 1.1698057651519775,
      "learning_rate": 4.9988051680944094e-05,
      "loss": 0.3513,
      "num_input_tokens_seen": 2725528,
      "step": 4155
    },
    {
      "epoch": 2.180293501048218,
      "grad_norm": 0.5877286791801453,
      "learning_rate": 4.998769559494591e-05,
      "loss": 0.3805,
      "num_input_tokens_seen": 2728824,
      "step": 4160
    },
    {
      "epoch": 2.1829140461215935,
      "grad_norm": 0.7740557789802551,
      "learning_rate": 4.998733428180689e-05,
      "loss": 0.5394,
      "num_input_tokens_seen": 2732472,
      "step": 4165
    },
    {
      "epoch": 2.1855345911949686,
      "grad_norm": 0.6815105080604553,
      "learning_rate": 4.998696774160259e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 2735544,
      "step": 4170
    },
    {
      "epoch": 2.188155136268344,
      "grad_norm": 0.8429138660430908,
      "learning_rate": 4.9986595974409714e-05,
      "loss": 0.4558,
      "num_input_tokens_seen": 2739032,
      "step": 4175
    },
    {
      "epoch": 2.190775681341719,
      "grad_norm": 1.1280319690704346,
      "learning_rate": 4.998621898030603e-05,
      "loss": 0.5546,
      "num_input_tokens_seen": 2742456,
      "step": 4180
    },
    {
      "epoch": 2.1933962264150946,
      "grad_norm": 1.42265784740448,
      "learning_rate": 4.998583675937039e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 2744888,
      "step": 4185
    },
    {
      "epoch": 2.1960167714884697,
      "grad_norm": 0.9691464304924011,
      "learning_rate": 4.998544931168274e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 2747512,
      "step": 4190
    },
    {
      "epoch": 2.198637316561845,
      "grad_norm": 0.8956053853034973,
      "learning_rate": 4.9985056637324154e-05,
      "loss": 0.4675,
      "num_input_tokens_seen": 2750648,
      "step": 4195
    },
    {
      "epoch": 2.20125786163522,
      "grad_norm": 1.2592849731445312,
      "learning_rate": 4.998465873637677e-05,
      "loss": 0.3467,
      "num_input_tokens_seen": 2753368,
      "step": 4200
    },
    {
      "epoch": 2.2038784067085953,
      "grad_norm": 1.059638500213623,
      "learning_rate": 4.9984255608923806e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 2756696,
      "step": 4205
    },
    {
      "epoch": 2.2064989517819704,
      "grad_norm": 1.6949458122253418,
      "learning_rate": 4.998384725504961e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 2759256,
      "step": 4210
    },
    {
      "epoch": 2.209119496855346,
      "grad_norm": 1.0745654106140137,
      "learning_rate": 4.9983433674839596e-05,
      "loss": 0.4922,
      "num_input_tokens_seen": 2762296,
      "step": 4215
    },
    {
      "epoch": 2.211740041928721,
      "grad_norm": 1.6838698387145996,
      "learning_rate": 4.998301486838028e-05,
      "loss": 0.4311,
      "num_input_tokens_seen": 2765400,
      "step": 4220
    },
    {
      "epoch": 2.2143605870020964,
      "grad_norm": 1.3699911832809448,
      "learning_rate": 4.998259083575929e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 2768120,
      "step": 4225
    },
    {
      "epoch": 2.2169811320754715,
      "grad_norm": 0.8958917856216431,
      "learning_rate": 4.99821615770653e-05,
      "loss": 0.4891,
      "num_input_tokens_seen": 2771096,
      "step": 4230
    },
    {
      "epoch": 2.219601677148847,
      "grad_norm": 1.301735520362854,
      "learning_rate": 4.998172709238813e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 2774328,
      "step": 4235
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.6917656660079956,
      "learning_rate": 4.9981287381818654e-05,
      "loss": 0.4986,
      "num_input_tokens_seen": 2778008,
      "step": 4240
    },
    {
      "epoch": 2.2248427672955975,
      "grad_norm": 2.5083141326904297,
      "learning_rate": 4.9980842445448856e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 2780216,
      "step": 4245
    },
    {
      "epoch": 2.2274633123689727,
      "grad_norm": 1.394303321838379,
      "learning_rate": 4.9980392283371815e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 2784088,
      "step": 4250
    },
    {
      "epoch": 2.230083857442348,
      "grad_norm": 0.9635797739028931,
      "learning_rate": 4.99799368956817e-05,
      "loss": 0.3528,
      "num_input_tokens_seen": 2786488,
      "step": 4255
    },
    {
      "epoch": 2.2327044025157234,
      "grad_norm": 1.2756898403167725,
      "learning_rate": 4.997947628247378e-05,
      "loss": 0.358,
      "num_input_tokens_seen": 2788920,
      "step": 4260
    },
    {
      "epoch": 2.2353249475890986,
      "grad_norm": 1.0811316967010498,
      "learning_rate": 4.9979010443844395e-05,
      "loss": 0.5976,
      "num_input_tokens_seen": 2791640,
      "step": 4265
    },
    {
      "epoch": 2.237945492662474,
      "grad_norm": 1.1904816627502441,
      "learning_rate": 4.9978539379891e-05,
      "loss": 0.3367,
      "num_input_tokens_seen": 2794904,
      "step": 4270
    },
    {
      "epoch": 2.240566037735849,
      "grad_norm": 0.9583600163459778,
      "learning_rate": 4.997806309071214e-05,
      "loss": 0.3873,
      "num_input_tokens_seen": 2798840,
      "step": 4275
    },
    {
      "epoch": 2.243186582809224,
      "grad_norm": 0.979896068572998,
      "learning_rate": 4.997758157640745e-05,
      "loss": 0.4851,
      "num_input_tokens_seen": 2802616,
      "step": 4280
    },
    {
      "epoch": 2.2458071278825997,
      "grad_norm": 1.4950098991394043,
      "learning_rate": 4.997709483707765e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 2805432,
      "step": 4285
    },
    {
      "epoch": 2.248427672955975,
      "grad_norm": 1.5229915380477905,
      "learning_rate": 4.997660287282456e-05,
      "loss": 0.4791,
      "num_input_tokens_seen": 2808696,
      "step": 4290
    },
    {
      "epoch": 2.25104821802935,
      "grad_norm": 0.8820854425430298,
      "learning_rate": 4.99761056837511e-05,
      "loss": 0.3892,
      "num_input_tokens_seen": 2811640,
      "step": 4295
    },
    {
      "epoch": 2.2536687631027252,
      "grad_norm": 0.9790735244750977,
      "learning_rate": 4.997560326996128e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 2815224,
      "step": 4300
    },
    {
      "epoch": 2.2562893081761004,
      "grad_norm": 0.9023936986923218,
      "learning_rate": 4.997509563156018e-05,
      "loss": 0.4639,
      "num_input_tokens_seen": 2817688,
      "step": 4305
    },
    {
      "epoch": 2.258909853249476,
      "grad_norm": 0.8386456966400146,
      "learning_rate": 4.997458276865401e-05,
      "loss": 0.3621,
      "num_input_tokens_seen": 2821016,
      "step": 4310
    },
    {
      "epoch": 2.261530398322851,
      "grad_norm": 1.1109496355056763,
      "learning_rate": 4.997406468135005e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 2823608,
      "step": 4315
    },
    {
      "epoch": 2.2641509433962264,
      "grad_norm": 1.0568325519561768,
      "learning_rate": 4.9973541369756686e-05,
      "loss": 0.3816,
      "num_input_tokens_seen": 2828248,
      "step": 4320
    },
    {
      "epoch": 2.2667714884696015,
      "grad_norm": 0.9341875314712524,
      "learning_rate": 4.9973012833983365e-05,
      "loss": 0.4859,
      "num_input_tokens_seen": 2832344,
      "step": 4325
    },
    {
      "epoch": 2.269392033542977,
      "grad_norm": 1.1064938306808472,
      "learning_rate": 4.9972479074140684e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 2836408,
      "step": 4330
    },
    {
      "epoch": 2.2720125786163523,
      "grad_norm": 0.8199834823608398,
      "learning_rate": 4.9971940090340276e-05,
      "loss": 0.4844,
      "num_input_tokens_seen": 2839320,
      "step": 4335
    },
    {
      "epoch": 2.2746331236897275,
      "grad_norm": 1.3928143978118896,
      "learning_rate": 4.997139588269489e-05,
      "loss": 0.4903,
      "num_input_tokens_seen": 2842456,
      "step": 4340
    },
    {
      "epoch": 2.2772536687631026,
      "grad_norm": 1.5273067951202393,
      "learning_rate": 4.997084645131838e-05,
      "loss": 0.558,
      "num_input_tokens_seen": 2845432,
      "step": 4345
    },
    {
      "epoch": 2.279874213836478,
      "grad_norm": 0.8181790113449097,
      "learning_rate": 4.9970291796325666e-05,
      "loss": 0.4951,
      "num_input_tokens_seen": 2848536,
      "step": 4350
    },
    {
      "epoch": 2.2824947589098534,
      "grad_norm": 0.8241934180259705,
      "learning_rate": 4.996973191783279e-05,
      "loss": 0.5409,
      "num_input_tokens_seen": 2851544,
      "step": 4355
    },
    {
      "epoch": 2.2851153039832286,
      "grad_norm": 1.0759936571121216,
      "learning_rate": 4.996916681595687e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 2854392,
      "step": 4360
    },
    {
      "epoch": 2.2877358490566038,
      "grad_norm": 0.7287266850471497,
      "learning_rate": 4.996859649081611e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 2858424,
      "step": 4365
    },
    {
      "epoch": 2.290356394129979,
      "grad_norm": 0.8690919280052185,
      "learning_rate": 4.996802094252983e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 2861368,
      "step": 4370
    },
    {
      "epoch": 2.2929769392033545,
      "grad_norm": 1.0349922180175781,
      "learning_rate": 4.996744017121841e-05,
      "loss": 0.4406,
      "num_input_tokens_seen": 2863960,
      "step": 4375
    },
    {
      "epoch": 2.2955974842767297,
      "grad_norm": 1.614778757095337,
      "learning_rate": 4.9966854177003355e-05,
      "loss": 0.5256,
      "num_input_tokens_seen": 2867352,
      "step": 4380
    },
    {
      "epoch": 2.298218029350105,
      "grad_norm": 1.1178346872329712,
      "learning_rate": 4.996626296000724e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 2871992,
      "step": 4385
    },
    {
      "epoch": 2.30083857442348,
      "grad_norm": 1.2045910358428955,
      "learning_rate": 4.996566652035374e-05,
      "loss": 0.4011,
      "num_input_tokens_seen": 2874680,
      "step": 4390
    },
    {
      "epoch": 2.3034591194968552,
      "grad_norm": 1.7131117582321167,
      "learning_rate": 4.996506485816763e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 2877688,
      "step": 4395
    },
    {
      "epoch": 2.3060796645702304,
      "grad_norm": 0.8357512950897217,
      "learning_rate": 4.996445797357477e-05,
      "loss": 0.5187,
      "num_input_tokens_seen": 2885528,
      "step": 4400
    },
    {
      "epoch": 2.308700209643606,
      "grad_norm": 0.8940435647964478,
      "learning_rate": 4.996384586670211e-05,
      "loss": 0.5683,
      "num_input_tokens_seen": 2889464,
      "step": 4405
    },
    {
      "epoch": 2.311320754716981,
      "grad_norm": 1.1085188388824463,
      "learning_rate": 4.99632285376777e-05,
      "loss": 0.3406,
      "num_input_tokens_seen": 2893464,
      "step": 4410
    },
    {
      "epoch": 2.3139412997903563,
      "grad_norm": 1.3269397020339966,
      "learning_rate": 4.9962605986630674e-05,
      "loss": 0.4799,
      "num_input_tokens_seen": 2896792,
      "step": 4415
    },
    {
      "epoch": 2.3165618448637315,
      "grad_norm": 1.0326014757156372,
      "learning_rate": 4.996197821369126e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 2900664,
      "step": 4420
    },
    {
      "epoch": 2.319182389937107,
      "grad_norm": 0.9346234202384949,
      "learning_rate": 4.9961345218990796e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 2904152,
      "step": 4425
    },
    {
      "epoch": 2.3218029350104823,
      "grad_norm": 0.9099321961402893,
      "learning_rate": 4.996070700266168e-05,
      "loss": 0.4479,
      "num_input_tokens_seen": 2906776,
      "step": 4430
    },
    {
      "epoch": 2.3244234800838575,
      "grad_norm": 0.7889841794967651,
      "learning_rate": 4.9960063564837424e-05,
      "loss": 0.4619,
      "num_input_tokens_seen": 2909720,
      "step": 4435
    },
    {
      "epoch": 2.3270440251572326,
      "grad_norm": 1.4740393161773682,
      "learning_rate": 4.995941490565262e-05,
      "loss": 0.3218,
      "num_input_tokens_seen": 2916536,
      "step": 4440
    },
    {
      "epoch": 2.329664570230608,
      "grad_norm": 1.4618377685546875,
      "learning_rate": 4.995876102524299e-05,
      "loss": 0.51,
      "num_input_tokens_seen": 2919608,
      "step": 4445
    },
    {
      "epoch": 2.3322851153039834,
      "grad_norm": 0.8530322313308716,
      "learning_rate": 4.995810192374529e-05,
      "loss": 0.5754,
      "num_input_tokens_seen": 2923352,
      "step": 4450
    },
    {
      "epoch": 2.3349056603773586,
      "grad_norm": 1.6213737726211548,
      "learning_rate": 4.99574376012974e-05,
      "loss": 0.4214,
      "num_input_tokens_seen": 2926232,
      "step": 4455
    },
    {
      "epoch": 2.3375262054507338,
      "grad_norm": 0.9508211016654968,
      "learning_rate": 4.995676805803829e-05,
      "loss": 0.4127,
      "num_input_tokens_seen": 2929656,
      "step": 4460
    },
    {
      "epoch": 2.340146750524109,
      "grad_norm": 0.9139396548271179,
      "learning_rate": 4.995609329410804e-05,
      "loss": 0.5086,
      "num_input_tokens_seen": 2932344,
      "step": 4465
    },
    {
      "epoch": 2.342767295597484,
      "grad_norm": 0.8153642416000366,
      "learning_rate": 4.9955413309647776e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 2936152,
      "step": 4470
    },
    {
      "epoch": 2.3453878406708597,
      "grad_norm": 0.8099355697631836,
      "learning_rate": 4.9954728104799756e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 2939224,
      "step": 4475
    },
    {
      "epoch": 2.348008385744235,
      "grad_norm": 0.8453266620635986,
      "learning_rate": 4.995403767970731e-05,
      "loss": 0.3561,
      "num_input_tokens_seen": 2941912,
      "step": 4480
    },
    {
      "epoch": 2.35062893081761,
      "grad_norm": 0.8667944073677063,
      "learning_rate": 4.995334203451488e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 2944568,
      "step": 4485
    },
    {
      "epoch": 2.353249475890985,
      "grad_norm": 0.8178995847702026,
      "learning_rate": 4.995264116936798e-05,
      "loss": 0.3407,
      "num_input_tokens_seen": 2947480,
      "step": 4490
    },
    {
      "epoch": 2.3558700209643604,
      "grad_norm": 1.3570282459259033,
      "learning_rate": 4.995193508441322e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 2951192,
      "step": 4495
    },
    {
      "epoch": 2.358490566037736,
      "grad_norm": 1.1334937810897827,
      "learning_rate": 4.99512237797983e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 2954648,
      "step": 4500
    },
    {
      "epoch": 2.361111111111111,
      "grad_norm": 0.6134249567985535,
      "learning_rate": 4.995050725567203e-05,
      "loss": 0.3605,
      "num_input_tokens_seen": 2958552,
      "step": 4505
    },
    {
      "epoch": 2.3637316561844863,
      "grad_norm": 0.6999497413635254,
      "learning_rate": 4.994978551218429e-05,
      "loss": 0.3739,
      "num_input_tokens_seen": 2962040,
      "step": 4510
    },
    {
      "epoch": 2.3663522012578615,
      "grad_norm": 0.7942072749137878,
      "learning_rate": 4.9949058549486066e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 2964824,
      "step": 4515
    },
    {
      "epoch": 2.368972746331237,
      "grad_norm": 0.7014560103416443,
      "learning_rate": 4.994832636772943e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 2967448,
      "step": 4520
    },
    {
      "epoch": 2.3715932914046123,
      "grad_norm": 0.9759536981582642,
      "learning_rate": 4.994758896706754e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 2971000,
      "step": 4525
    },
    {
      "epoch": 2.3742138364779874,
      "grad_norm": 1.654792070388794,
      "learning_rate": 4.994684634765465e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 2973592,
      "step": 4530
    },
    {
      "epoch": 2.3768343815513626,
      "grad_norm": 0.8894294500350952,
      "learning_rate": 4.9946098509646125e-05,
      "loss": 0.5156,
      "num_input_tokens_seen": 2977560,
      "step": 4535
    },
    {
      "epoch": 2.379454926624738,
      "grad_norm": 0.9547339677810669,
      "learning_rate": 4.994534545319838e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 2980280,
      "step": 4540
    },
    {
      "epoch": 2.3820754716981134,
      "grad_norm": 1.007703423500061,
      "learning_rate": 4.994458717846896e-05,
      "loss": 0.4071,
      "num_input_tokens_seen": 2983352,
      "step": 4545
    },
    {
      "epoch": 2.3846960167714886,
      "grad_norm": 0.940125584602356,
      "learning_rate": 4.9943823685616485e-05,
      "loss": 0.4867,
      "num_input_tokens_seen": 2986648,
      "step": 4550
    },
    {
      "epoch": 2.3873165618448637,
      "grad_norm": 1.337336540222168,
      "learning_rate": 4.994305497480067e-05,
      "loss": 0.4157,
      "num_input_tokens_seen": 2989880,
      "step": 4555
    },
    {
      "epoch": 2.389937106918239,
      "grad_norm": 1.1791070699691772,
      "learning_rate": 4.994228104618233e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 2992536,
      "step": 4560
    },
    {
      "epoch": 2.392557651991614,
      "grad_norm": 1.7307493686676025,
      "learning_rate": 4.994150189992334e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 2996152,
      "step": 4565
    },
    {
      "epoch": 2.3951781970649897,
      "grad_norm": 2.431165933609009,
      "learning_rate": 4.99407175361867e-05,
      "loss": 0.4263,
      "num_input_tokens_seen": 2999288,
      "step": 4570
    },
    {
      "epoch": 2.397798742138365,
      "grad_norm": 1.1340893507003784,
      "learning_rate": 4.99399279551365e-05,
      "loss": 0.3369,
      "num_input_tokens_seen": 3002520,
      "step": 4575
    },
    {
      "epoch": 2.40041928721174,
      "grad_norm": 1.0937410593032837,
      "learning_rate": 4.9939133156937886e-05,
      "loss": 0.5275,
      "num_input_tokens_seen": 3005752,
      "step": 4580
    },
    {
      "epoch": 2.403039832285115,
      "grad_norm": 0.9767884612083435,
      "learning_rate": 4.993833314175716e-05,
      "loss": 0.4477,
      "num_input_tokens_seen": 3009432,
      "step": 4585
    },
    {
      "epoch": 2.4056603773584904,
      "grad_norm": 1.0828862190246582,
      "learning_rate": 4.993752790976163e-05,
      "loss": 0.4782,
      "num_input_tokens_seen": 3014136,
      "step": 4590
    },
    {
      "epoch": 2.408280922431866,
      "grad_norm": 0.8273671865463257,
      "learning_rate": 4.993671746111977e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 3017144,
      "step": 4595
    },
    {
      "epoch": 2.410901467505241,
      "grad_norm": 1.8383989334106445,
      "learning_rate": 4.993590179600111e-05,
      "loss": 0.4822,
      "num_input_tokens_seen": 3019448,
      "step": 4600
    },
    {
      "epoch": 2.4135220125786163,
      "grad_norm": 0.8419188857078552,
      "learning_rate": 4.9935080914576284e-05,
      "loss": 0.3375,
      "num_input_tokens_seen": 3022104,
      "step": 4605
    },
    {
      "epoch": 2.4161425576519915,
      "grad_norm": 1.1976885795593262,
      "learning_rate": 4.9934254817016995e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 3025688,
      "step": 4610
    },
    {
      "epoch": 2.418763102725367,
      "grad_norm": 0.8790993094444275,
      "learning_rate": 4.993342350349607e-05,
      "loss": 0.3226,
      "num_input_tokens_seen": 3029816,
      "step": 4615
    },
    {
      "epoch": 2.4213836477987423,
      "grad_norm": 0.6937848925590515,
      "learning_rate": 4.9932586974187404e-05,
      "loss": 0.3179,
      "num_input_tokens_seen": 3032824,
      "step": 4620
    },
    {
      "epoch": 2.4240041928721174,
      "grad_norm": 0.9883701205253601,
      "learning_rate": 4.9931745229265986e-05,
      "loss": 0.5591,
      "num_input_tokens_seen": 3036632,
      "step": 4625
    },
    {
      "epoch": 2.4266247379454926,
      "grad_norm": 1.2346781492233276,
      "learning_rate": 4.99308982689079e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 3039384,
      "step": 4630
    },
    {
      "epoch": 2.4292452830188678,
      "grad_norm": 0.7027636766433716,
      "learning_rate": 4.993004609329033e-05,
      "loss": 0.4353,
      "num_input_tokens_seen": 3042296,
      "step": 4635
    },
    {
      "epoch": 2.431865828092243,
      "grad_norm": 0.8931586742401123,
      "learning_rate": 4.992918870259153e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 3045304,
      "step": 4640
    },
    {
      "epoch": 2.4344863731656186,
      "grad_norm": 1.161683440208435,
      "learning_rate": 4.992832609699086e-05,
      "loss": 0.3549,
      "num_input_tokens_seen": 3048824,
      "step": 4645
    },
    {
      "epoch": 2.4371069182389937,
      "grad_norm": 0.7877296805381775,
      "learning_rate": 4.992745827666877e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 3052120,
      "step": 4650
    },
    {
      "epoch": 2.439727463312369,
      "grad_norm": 0.8748054504394531,
      "learning_rate": 4.99265852418068e-05,
      "loss": 0.4332,
      "num_input_tokens_seen": 3055384,
      "step": 4655
    },
    {
      "epoch": 2.442348008385744,
      "grad_norm": 0.6292238831520081,
      "learning_rate": 4.992570699258757e-05,
      "loss": 0.3382,
      "num_input_tokens_seen": 3058328,
      "step": 4660
    },
    {
      "epoch": 2.4449685534591197,
      "grad_norm": 0.6316247582435608,
      "learning_rate": 4.99248235291948e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 3063512,
      "step": 4665
    },
    {
      "epoch": 2.447589098532495,
      "grad_norm": 1.0266984701156616,
      "learning_rate": 4.9923934851813314e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 3068056,
      "step": 4670
    },
    {
      "epoch": 2.45020964360587,
      "grad_norm": 0.8972869515419006,
      "learning_rate": 4.9923040960628996e-05,
      "loss": 0.3915,
      "num_input_tokens_seen": 3071608,
      "step": 4675
    },
    {
      "epoch": 2.452830188679245,
      "grad_norm": 1.841019868850708,
      "learning_rate": 4.9922141855828856e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 3074872,
      "step": 4680
    },
    {
      "epoch": 2.4554507337526204,
      "grad_norm": 1.3335187435150146,
      "learning_rate": 4.992123753760096e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 3077720,
      "step": 4685
    },
    {
      "epoch": 2.458071278825996,
      "grad_norm": 1.1303366422653198,
      "learning_rate": 4.9920328006134484e-05,
      "loss": 0.5298,
      "num_input_tokens_seen": 3080664,
      "step": 4690
    },
    {
      "epoch": 2.460691823899371,
      "grad_norm": 1.20348060131073,
      "learning_rate": 4.991941326161971e-05,
      "loss": 0.4358,
      "num_input_tokens_seen": 3084408,
      "step": 4695
    },
    {
      "epoch": 2.4633123689727463,
      "grad_norm": 0.8186688423156738,
      "learning_rate": 4.991849330424796e-05,
      "loss": 0.4825,
      "num_input_tokens_seen": 3087640,
      "step": 4700
    },
    {
      "epoch": 2.4659329140461215,
      "grad_norm": 0.9811978936195374,
      "learning_rate": 4.9917568134211703e-05,
      "loss": 0.4955,
      "num_input_tokens_seen": 3090520,
      "step": 4705
    },
    {
      "epoch": 2.468553459119497,
      "grad_norm": 0.8562654256820679,
      "learning_rate": 4.991663775170447e-05,
      "loss": 0.3951,
      "num_input_tokens_seen": 3093624,
      "step": 4710
    },
    {
      "epoch": 2.4711740041928723,
      "grad_norm": 0.7299864292144775,
      "learning_rate": 4.991570215692088e-05,
      "loss": 0.4022,
      "num_input_tokens_seen": 3096184,
      "step": 4715
    },
    {
      "epoch": 2.4737945492662474,
      "grad_norm": 1.2840220928192139,
      "learning_rate": 4.991476135005665e-05,
      "loss": 0.5277,
      "num_input_tokens_seen": 3099928,
      "step": 4720
    },
    {
      "epoch": 2.4764150943396226,
      "grad_norm": 1.1632888317108154,
      "learning_rate": 4.991381533130859e-05,
      "loss": 0.4247,
      "num_input_tokens_seen": 3102456,
      "step": 4725
    },
    {
      "epoch": 2.4790356394129978,
      "grad_norm": 2.453960657119751,
      "learning_rate": 4.99128641008746e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 3104728,
      "step": 4730
    },
    {
      "epoch": 2.481656184486373,
      "grad_norm": 1.7334926128387451,
      "learning_rate": 4.9911907658953664e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 3107704,
      "step": 4735
    },
    {
      "epoch": 2.4842767295597485,
      "grad_norm": 0.8063175678253174,
      "learning_rate": 4.991094600574584e-05,
      "loss": 0.3769,
      "num_input_tokens_seen": 3110328,
      "step": 4740
    },
    {
      "epoch": 2.4868972746331237,
      "grad_norm": 0.9002105593681335,
      "learning_rate": 4.9909979141452324e-05,
      "loss": 0.4645,
      "num_input_tokens_seen": 3113464,
      "step": 4745
    },
    {
      "epoch": 2.489517819706499,
      "grad_norm": 0.7292216420173645,
      "learning_rate": 4.9909007066275346e-05,
      "loss": 0.3848,
      "num_input_tokens_seen": 3117112,
      "step": 4750
    },
    {
      "epoch": 2.492138364779874,
      "grad_norm": 1.0988832712173462,
      "learning_rate": 4.990802978041828e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 3120280,
      "step": 4755
    },
    {
      "epoch": 2.4947589098532497,
      "grad_norm": 0.8403136730194092,
      "learning_rate": 4.990704728408555e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 3123928,
      "step": 4760
    },
    {
      "epoch": 2.497379454926625,
      "grad_norm": 0.9639562964439392,
      "learning_rate": 4.990605957748268e-05,
      "loss": 0.3691,
      "num_input_tokens_seen": 3126584,
      "step": 4765
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.3065717220306396,
      "learning_rate": 4.990506666081628e-05,
      "loss": 0.5382,
      "num_input_tokens_seen": 3129368,
      "step": 4770
    },
    {
      "epoch": 2.502620545073375,
      "grad_norm": 1.9460973739624023,
      "learning_rate": 4.990406853429408e-05,
      "loss": 0.5259,
      "num_input_tokens_seen": 3132248,
      "step": 4775
    },
    {
      "epoch": 2.5052410901467503,
      "grad_norm": 0.6801477670669556,
      "learning_rate": 4.990306519812485e-05,
      "loss": 0.3254,
      "num_input_tokens_seen": 3135320,
      "step": 4780
    },
    {
      "epoch": 2.507861635220126,
      "grad_norm": 1.313228726387024,
      "learning_rate": 4.99020566525185e-05,
      "loss": 0.6554,
      "num_input_tokens_seen": 3139032,
      "step": 4785
    },
    {
      "epoch": 2.510482180293501,
      "grad_norm": 1.124787449836731,
      "learning_rate": 4.9901042897685996e-05,
      "loss": 0.5826,
      "num_input_tokens_seen": 3141944,
      "step": 4790
    },
    {
      "epoch": 2.5131027253668763,
      "grad_norm": 1.5207223892211914,
      "learning_rate": 4.99000239338394e-05,
      "loss": 0.435,
      "num_input_tokens_seen": 3145400,
      "step": 4795
    },
    {
      "epoch": 2.5157232704402515,
      "grad_norm": 0.6984018683433533,
      "learning_rate": 4.989899976119187e-05,
      "loss": 0.5247,
      "num_input_tokens_seen": 3149496,
      "step": 4800
    },
    {
      "epoch": 2.518343815513627,
      "grad_norm": 0.9024474620819092,
      "learning_rate": 4.989797037995766e-05,
      "loss": 0.5381,
      "num_input_tokens_seen": 3152344,
      "step": 4805
    },
    {
      "epoch": 2.5209643605870022,
      "grad_norm": 0.7873271107673645,
      "learning_rate": 4.989693579035209e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 3155192,
      "step": 4810
    },
    {
      "epoch": 2.5235849056603774,
      "grad_norm": 1.1532708406448364,
      "learning_rate": 4.989589599259159e-05,
      "loss": 0.3534,
      "num_input_tokens_seen": 3157688,
      "step": 4815
    },
    {
      "epoch": 2.5262054507337526,
      "grad_norm": 0.5361354947090149,
      "learning_rate": 4.9894850986893685e-05,
      "loss": 0.4294,
      "num_input_tokens_seen": 3161592,
      "step": 4820
    },
    {
      "epoch": 2.5288259958071277,
      "grad_norm": 1.5745389461517334,
      "learning_rate": 4.9893800773476965e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 3164024,
      "step": 4825
    },
    {
      "epoch": 2.531446540880503,
      "grad_norm": 1.1866437196731567,
      "learning_rate": 4.9892745352561134e-05,
      "loss": 0.3493,
      "num_input_tokens_seen": 3166424,
      "step": 4830
    },
    {
      "epoch": 2.5340670859538785,
      "grad_norm": 1.0896060466766357,
      "learning_rate": 4.9891684724366955e-05,
      "loss": 0.6008,
      "num_input_tokens_seen": 3169240,
      "step": 4835
    },
    {
      "epoch": 2.5366876310272537,
      "grad_norm": 1.0861070156097412,
      "learning_rate": 4.989061888911631e-05,
      "loss": 0.4471,
      "num_input_tokens_seen": 3172088,
      "step": 4840
    },
    {
      "epoch": 2.539308176100629,
      "grad_norm": 1.072611689567566,
      "learning_rate": 4.988954784703217e-05,
      "loss": 0.4462,
      "num_input_tokens_seen": 3175480,
      "step": 4845
    },
    {
      "epoch": 2.541928721174004,
      "grad_norm": 0.7668029069900513,
      "learning_rate": 4.9888471598338574e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 3179480,
      "step": 4850
    },
    {
      "epoch": 2.5445492662473796,
      "grad_norm": 0.7206979393959045,
      "learning_rate": 4.988739014326066e-05,
      "loss": 0.4175,
      "num_input_tokens_seen": 3182520,
      "step": 4855
    },
    {
      "epoch": 2.547169811320755,
      "grad_norm": 0.9272150993347168,
      "learning_rate": 4.9886303482024656e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 3185560,
      "step": 4860
    },
    {
      "epoch": 2.54979035639413,
      "grad_norm": 0.9358075261116028,
      "learning_rate": 4.988521161485789e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 3189592,
      "step": 4865
    },
    {
      "epoch": 2.552410901467505,
      "grad_norm": 2.1239211559295654,
      "learning_rate": 4.9884114541988747e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 3192760,
      "step": 4870
    },
    {
      "epoch": 2.5550314465408803,
      "grad_norm": 1.3913785219192505,
      "learning_rate": 4.988301226364674e-05,
      "loss": 0.3692,
      "num_input_tokens_seen": 3195992,
      "step": 4875
    },
    {
      "epoch": 2.5576519916142555,
      "grad_norm": 1.299209713935852,
      "learning_rate": 4.988190478006245e-05,
      "loss": 0.449,
      "num_input_tokens_seen": 3198200,
      "step": 4880
    },
    {
      "epoch": 2.560272536687631,
      "grad_norm": 0.6559023857116699,
      "learning_rate": 4.988079209146754e-05,
      "loss": 0.4249,
      "num_input_tokens_seen": 3201848,
      "step": 4885
    },
    {
      "epoch": 2.5628930817610063,
      "grad_norm": 2.4413156509399414,
      "learning_rate": 4.9879674198094794e-05,
      "loss": 0.6002,
      "num_input_tokens_seen": 3204504,
      "step": 4890
    },
    {
      "epoch": 2.5655136268343814,
      "grad_norm": 1.0521197319030762,
      "learning_rate": 4.987855110017804e-05,
      "loss": 0.5539,
      "num_input_tokens_seen": 3207320,
      "step": 4895
    },
    {
      "epoch": 2.568134171907757,
      "grad_norm": 0.9547411799430847,
      "learning_rate": 4.9877422797952224e-05,
      "loss": 0.4039,
      "num_input_tokens_seen": 3210584,
      "step": 4900
    },
    {
      "epoch": 2.5707547169811322,
      "grad_norm": 0.9833049774169922,
      "learning_rate": 4.987628929165338e-05,
      "loss": 0.334,
      "num_input_tokens_seen": 3213496,
      "step": 4905
    },
    {
      "epoch": 2.5733752620545074,
      "grad_norm": 0.6294229030609131,
      "learning_rate": 4.9875150581518615e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 3217016,
      "step": 4910
    },
    {
      "epoch": 2.5759958071278826,
      "grad_norm": 0.724772572517395,
      "learning_rate": 4.987400666778614e-05,
      "loss": 0.5422,
      "num_input_tokens_seen": 3219800,
      "step": 4915
    },
    {
      "epoch": 2.5786163522012577,
      "grad_norm": 1.0680897235870361,
      "learning_rate": 4.987285755069524e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 3222424,
      "step": 4920
    },
    {
      "epoch": 2.581236897274633,
      "grad_norm": 1.3875783681869507,
      "learning_rate": 4.987170323048632e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 3225272,
      "step": 4925
    },
    {
      "epoch": 2.5838574423480085,
      "grad_norm": 0.7660390734672546,
      "learning_rate": 4.987054370740083e-05,
      "loss": 0.3968,
      "num_input_tokens_seen": 3228088,
      "step": 4930
    },
    {
      "epoch": 2.5864779874213837,
      "grad_norm": 0.7820528149604797,
      "learning_rate": 4.986937898168134e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 3231096,
      "step": 4935
    },
    {
      "epoch": 2.589098532494759,
      "grad_norm": 1.6510859727859497,
      "learning_rate": 4.986820905357148e-05,
      "loss": 0.5202,
      "num_input_tokens_seen": 3234552,
      "step": 4940
    },
    {
      "epoch": 2.591719077568134,
      "grad_norm": 1.8995813131332397,
      "learning_rate": 4.9867033923315995e-05,
      "loss": 0.358,
      "num_input_tokens_seen": 3238072,
      "step": 4945
    },
    {
      "epoch": 2.5943396226415096,
      "grad_norm": 1.1122084856033325,
      "learning_rate": 4.9865853591160724e-05,
      "loss": 0.4983,
      "num_input_tokens_seen": 3241624,
      "step": 4950
    },
    {
      "epoch": 2.596960167714885,
      "grad_norm": 1.0678510665893555,
      "learning_rate": 4.9864668057352556e-05,
      "loss": 0.7068,
      "num_input_tokens_seen": 3245080,
      "step": 4955
    },
    {
      "epoch": 2.59958071278826,
      "grad_norm": 0.936063289642334,
      "learning_rate": 4.9863477322139517e-05,
      "loss": 0.4869,
      "num_input_tokens_seen": 3247800,
      "step": 4960
    },
    {
      "epoch": 2.602201257861635,
      "grad_norm": 0.6619852185249329,
      "learning_rate": 4.986228138577067e-05,
      "loss": 0.3364,
      "num_input_tokens_seen": 3251000,
      "step": 4965
    },
    {
      "epoch": 2.6048218029350103,
      "grad_norm": 0.8489488959312439,
      "learning_rate": 4.986108024849619e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 3254872,
      "step": 4970
    },
    {
      "epoch": 2.6074423480083855,
      "grad_norm": 0.972998321056366,
      "learning_rate": 4.985987391056737e-05,
      "loss": 0.468,
      "num_input_tokens_seen": 3258520,
      "step": 4975
    },
    {
      "epoch": 2.610062893081761,
      "grad_norm": 1.0234451293945312,
      "learning_rate": 4.9858662372236544e-05,
      "loss": 0.4843,
      "num_input_tokens_seen": 3261944,
      "step": 4980
    },
    {
      "epoch": 2.6126834381551363,
      "grad_norm": 0.8305946588516235,
      "learning_rate": 4.985744563375714e-05,
      "loss": 0.4995,
      "num_input_tokens_seen": 3265432,
      "step": 4985
    },
    {
      "epoch": 2.6153039832285114,
      "grad_norm": 1.5783718824386597,
      "learning_rate": 4.985622369538371e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 3268632,
      "step": 4990
    },
    {
      "epoch": 2.617924528301887,
      "grad_norm": 1.5278154611587524,
      "learning_rate": 4.985499655737185e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 3271416,
      "step": 4995
    },
    {
      "epoch": 2.620545073375262,
      "grad_norm": 1.0186734199523926,
      "learning_rate": 4.985376421997827e-05,
      "loss": 0.3272,
      "num_input_tokens_seen": 3274424,
      "step": 5000
    },
    {
      "epoch": 2.6231656184486374,
      "grad_norm": 1.1219056844711304,
      "learning_rate": 4.985252668346076e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 3277688,
      "step": 5005
    },
    {
      "epoch": 2.6257861635220126,
      "grad_norm": 0.5762096643447876,
      "learning_rate": 4.9851283948078206e-05,
      "loss": 0.3491,
      "num_input_tokens_seen": 3280440,
      "step": 5010
    },
    {
      "epoch": 2.6284067085953877,
      "grad_norm": 1.458986520767212,
      "learning_rate": 4.985003601409057e-05,
      "loss": 0.6238,
      "num_input_tokens_seen": 3283704,
      "step": 5015
    },
    {
      "epoch": 2.631027253668763,
      "grad_norm": 0.8948667049407959,
      "learning_rate": 4.9848782881758906e-05,
      "loss": 0.4068,
      "num_input_tokens_seen": 3286808,
      "step": 5020
    },
    {
      "epoch": 2.6336477987421385,
      "grad_norm": 0.7779580950737,
      "learning_rate": 4.9847524551345345e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 3289592,
      "step": 5025
    },
    {
      "epoch": 2.6362683438155137,
      "grad_norm": 0.8737650513648987,
      "learning_rate": 4.984626102311313e-05,
      "loss": 0.45,
      "num_input_tokens_seen": 3292696,
      "step": 5030
    },
    {
      "epoch": 2.638888888888889,
      "grad_norm": 0.8664358258247375,
      "learning_rate": 4.984499229732656e-05,
      "loss": 0.3045,
      "num_input_tokens_seen": 3296632,
      "step": 5035
    },
    {
      "epoch": 2.641509433962264,
      "grad_norm": 1.371038794517517,
      "learning_rate": 4.984371837425106e-05,
      "loss": 0.3675,
      "num_input_tokens_seen": 3300536,
      "step": 5040
    },
    {
      "epoch": 2.6441299790356396,
      "grad_norm": 1.0012391805648804,
      "learning_rate": 4.9842439254153106e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 3304728,
      "step": 5045
    },
    {
      "epoch": 2.646750524109015,
      "grad_norm": 1.2487449645996094,
      "learning_rate": 4.984115493730027e-05,
      "loss": 0.4884,
      "num_input_tokens_seen": 3307928,
      "step": 5050
    },
    {
      "epoch": 2.64937106918239,
      "grad_norm": 1.1153794527053833,
      "learning_rate": 4.983986542396123e-05,
      "loss": 0.5561,
      "num_input_tokens_seen": 3310552,
      "step": 5055
    },
    {
      "epoch": 2.651991614255765,
      "grad_norm": 0.978253185749054,
      "learning_rate": 4.983857071440573e-05,
      "loss": 0.5572,
      "num_input_tokens_seen": 3314904,
      "step": 5060
    },
    {
      "epoch": 2.6546121593291403,
      "grad_norm": 0.8714599013328552,
      "learning_rate": 4.9837270808904616e-05,
      "loss": 0.4288,
      "num_input_tokens_seen": 3320184,
      "step": 5065
    },
    {
      "epoch": 2.6572327044025155,
      "grad_norm": 1.4069757461547852,
      "learning_rate": 4.9835965707729807e-05,
      "loss": 0.3961,
      "num_input_tokens_seen": 3322904,
      "step": 5070
    },
    {
      "epoch": 2.659853249475891,
      "grad_norm": 0.8109891414642334,
      "learning_rate": 4.983465541115431e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 3325784,
      "step": 5075
    },
    {
      "epoch": 2.6624737945492662,
      "grad_norm": 0.8539614081382751,
      "learning_rate": 4.983333991945223e-05,
      "loss": 0.4909,
      "num_input_tokens_seen": 3328760,
      "step": 5080
    },
    {
      "epoch": 2.6650943396226414,
      "grad_norm": 1.3139861822128296,
      "learning_rate": 4.983201923289875e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 3331576,
      "step": 5085
    },
    {
      "epoch": 2.667714884696017,
      "grad_norm": 1.4376450777053833,
      "learning_rate": 4.983069335177014e-05,
      "loss": 0.5024,
      "num_input_tokens_seen": 3333880,
      "step": 5090
    },
    {
      "epoch": 2.670335429769392,
      "grad_norm": 1.3314241170883179,
      "learning_rate": 4.982936227634378e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 3337880,
      "step": 5095
    },
    {
      "epoch": 2.6729559748427674,
      "grad_norm": 0.7745426297187805,
      "learning_rate": 4.982802600689809e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 3340920,
      "step": 5100
    },
    {
      "epoch": 2.6755765199161425,
      "grad_norm": 1.0145207643508911,
      "learning_rate": 4.982668454371261e-05,
      "loss": 0.4077,
      "num_input_tokens_seen": 3344312,
      "step": 5105
    },
    {
      "epoch": 2.6781970649895177,
      "grad_norm": 0.8223521113395691,
      "learning_rate": 4.9825337887067966e-05,
      "loss": 0.3881,
      "num_input_tokens_seen": 3347672,
      "step": 5110
    },
    {
      "epoch": 2.680817610062893,
      "grad_norm": 0.8266567587852478,
      "learning_rate": 4.982398603724586e-05,
      "loss": 0.4637,
      "num_input_tokens_seen": 3351032,
      "step": 5115
    },
    {
      "epoch": 2.6834381551362685,
      "grad_norm": 1.165482521057129,
      "learning_rate": 4.982262899452909e-05,
      "loss": 0.5954,
      "num_input_tokens_seen": 3353752,
      "step": 5120
    },
    {
      "epoch": 2.6860587002096437,
      "grad_norm": 1.0625375509262085,
      "learning_rate": 4.982126675920151e-05,
      "loss": 0.4021,
      "num_input_tokens_seen": 3356600,
      "step": 5125
    },
    {
      "epoch": 2.688679245283019,
      "grad_norm": 1.253237009048462,
      "learning_rate": 4.98198993315481e-05,
      "loss": 0.5022,
      "num_input_tokens_seen": 3360280,
      "step": 5130
    },
    {
      "epoch": 2.691299790356394,
      "grad_norm": 0.94330233335495,
      "learning_rate": 4.981852671185492e-05,
      "loss": 0.5653,
      "num_input_tokens_seen": 3363768,
      "step": 5135
    },
    {
      "epoch": 2.6939203354297696,
      "grad_norm": 0.8090857863426208,
      "learning_rate": 4.9817148900409093e-05,
      "loss": 0.3921,
      "num_input_tokens_seen": 3367480,
      "step": 5140
    },
    {
      "epoch": 2.6965408805031448,
      "grad_norm": 1.9015811681747437,
      "learning_rate": 4.981576589749885e-05,
      "loss": 0.4467,
      "num_input_tokens_seen": 3370520,
      "step": 5145
    },
    {
      "epoch": 2.69916142557652,
      "grad_norm": 1.3386646509170532,
      "learning_rate": 4.981437770341348e-05,
      "loss": 0.3898,
      "num_input_tokens_seen": 3372984,
      "step": 5150
    },
    {
      "epoch": 2.701781970649895,
      "grad_norm": 1.7751637697219849,
      "learning_rate": 4.9812984318443414e-05,
      "loss": 0.4428,
      "num_input_tokens_seen": 3375288,
      "step": 5155
    },
    {
      "epoch": 2.7044025157232703,
      "grad_norm": 1.0934585332870483,
      "learning_rate": 4.98115857428801e-05,
      "loss": 0.3542,
      "num_input_tokens_seen": 3378488,
      "step": 5160
    },
    {
      "epoch": 2.7070230607966455,
      "grad_norm": 0.8715333938598633,
      "learning_rate": 4.9810181977016104e-05,
      "loss": 0.3907,
      "num_input_tokens_seen": 3381368,
      "step": 5165
    },
    {
      "epoch": 2.709643605870021,
      "grad_norm": 2.0910489559173584,
      "learning_rate": 4.9808773021145096e-05,
      "loss": 0.4584,
      "num_input_tokens_seen": 3385016,
      "step": 5170
    },
    {
      "epoch": 2.7122641509433962,
      "grad_norm": 0.6804368495941162,
      "learning_rate": 4.98073588755618e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 3387992,
      "step": 5175
    },
    {
      "epoch": 2.7148846960167714,
      "grad_norm": 0.7627809643745422,
      "learning_rate": 4.980593954056205e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 3390872,
      "step": 5180
    },
    {
      "epoch": 2.717505241090147,
      "grad_norm": 1.317960262298584,
      "learning_rate": 4.9804515016442754e-05,
      "loss": 0.4775,
      "num_input_tokens_seen": 3393496,
      "step": 5185
    },
    {
      "epoch": 2.720125786163522,
      "grad_norm": 1.2220566272735596,
      "learning_rate": 4.9803085303501896e-05,
      "loss": 0.6623,
      "num_input_tokens_seen": 3395992,
      "step": 5190
    },
    {
      "epoch": 2.7227463312368974,
      "grad_norm": 1.1885526180267334,
      "learning_rate": 4.9801650402038555e-05,
      "loss": 0.5397,
      "num_input_tokens_seen": 3399160,
      "step": 5195
    },
    {
      "epoch": 2.7253668763102725,
      "grad_norm": 1.082427978515625,
      "learning_rate": 4.980021031235291e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 3402072,
      "step": 5200
    },
    {
      "epoch": 2.7279874213836477,
      "grad_norm": 1.249182105064392,
      "learning_rate": 4.9798765034746206e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 3404344,
      "step": 5205
    },
    {
      "epoch": 2.730607966457023,
      "grad_norm": 0.9153455495834351,
      "learning_rate": 4.979731456952077e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 3407640,
      "step": 5210
    },
    {
      "epoch": 2.7332285115303985,
      "grad_norm": 1.0273938179016113,
      "learning_rate": 4.9795858916980034e-05,
      "loss": 0.3483,
      "num_input_tokens_seen": 3410200,
      "step": 5215
    },
    {
      "epoch": 2.7358490566037736,
      "grad_norm": 0.979881227016449,
      "learning_rate": 4.9794398077428495e-05,
      "loss": 0.3636,
      "num_input_tokens_seen": 3412856,
      "step": 5220
    },
    {
      "epoch": 2.738469601677149,
      "grad_norm": 0.5403985381126404,
      "learning_rate": 4.979293205117175e-05,
      "loss": 0.4296,
      "num_input_tokens_seen": 3415864,
      "step": 5225
    },
    {
      "epoch": 2.741090146750524,
      "grad_norm": 0.6778361797332764,
      "learning_rate": 4.979146083851647e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 3419224,
      "step": 5230
    },
    {
      "epoch": 2.7437106918238996,
      "grad_norm": 1.4220987558364868,
      "learning_rate": 4.978998443977042e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 3422520,
      "step": 5235
    },
    {
      "epoch": 2.7463312368972748,
      "grad_norm": 1.089937448501587,
      "learning_rate": 4.9788502855242436e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 3425976,
      "step": 5240
    },
    {
      "epoch": 2.74895178197065,
      "grad_norm": 0.8441261053085327,
      "learning_rate": 4.978701608524247e-05,
      "loss": 0.4508,
      "num_input_tokens_seen": 3429368,
      "step": 5245
    },
    {
      "epoch": 2.751572327044025,
      "grad_norm": 1.1600626707077026,
      "learning_rate": 4.978552413008152e-05,
      "loss": 0.4941,
      "num_input_tokens_seen": 3434680,
      "step": 5250
    },
    {
      "epoch": 2.7541928721174003,
      "grad_norm": 1.5682578086853027,
      "learning_rate": 4.9784026990071687e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 3437752,
      "step": 5255
    },
    {
      "epoch": 2.7568134171907754,
      "grad_norm": 0.6980395913124084,
      "learning_rate": 4.978252466552615e-05,
      "loss": 0.5331,
      "num_input_tokens_seen": 3440760,
      "step": 5260
    },
    {
      "epoch": 2.759433962264151,
      "grad_norm": 1.3447577953338623,
      "learning_rate": 4.9781017156759205e-05,
      "loss": 0.4741,
      "num_input_tokens_seen": 3443736,
      "step": 5265
    },
    {
      "epoch": 2.762054507337526,
      "grad_norm": 1.210381269454956,
      "learning_rate": 4.977950446408618e-05,
      "loss": 0.5603,
      "num_input_tokens_seen": 3446904,
      "step": 5270
    },
    {
      "epoch": 2.7646750524109014,
      "grad_norm": 1.1257424354553223,
      "learning_rate": 4.9777986587823523e-05,
      "loss": 0.3821,
      "num_input_tokens_seen": 3450104,
      "step": 5275
    },
    {
      "epoch": 2.767295597484277,
      "grad_norm": 0.6903092265129089,
      "learning_rate": 4.977646352828875e-05,
      "loss": 0.5846,
      "num_input_tokens_seen": 3453592,
      "step": 5280
    },
    {
      "epoch": 2.769916142557652,
      "grad_norm": 1.3636537790298462,
      "learning_rate": 4.9774935285800476e-05,
      "loss": 0.4386,
      "num_input_tokens_seen": 3456824,
      "step": 5285
    },
    {
      "epoch": 2.7725366876310273,
      "grad_norm": 1.4441767930984497,
      "learning_rate": 4.977340186067838e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 3460792,
      "step": 5290
    },
    {
      "epoch": 2.7751572327044025,
      "grad_norm": 0.8986649513244629,
      "learning_rate": 4.977186325324325e-05,
      "loss": 0.4896,
      "num_input_tokens_seen": 3464280,
      "step": 5295
    },
    {
      "epoch": 2.7777777777777777,
      "grad_norm": 2.3286004066467285,
      "learning_rate": 4.977031946381694e-05,
      "loss": 0.5583,
      "num_input_tokens_seen": 3467320,
      "step": 5300
    },
    {
      "epoch": 2.780398322851153,
      "grad_norm": 1.155515432357788,
      "learning_rate": 4.9768770492722394e-05,
      "loss": 0.4221,
      "num_input_tokens_seen": 3470520,
      "step": 5305
    },
    {
      "epoch": 2.7830188679245285,
      "grad_norm": 0.9925292134284973,
      "learning_rate": 4.976721634028363e-05,
      "loss": 0.4543,
      "num_input_tokens_seen": 3474008,
      "step": 5310
    },
    {
      "epoch": 2.7856394129979036,
      "grad_norm": 1.2621760368347168,
      "learning_rate": 4.976565700682578e-05,
      "loss": 0.3872,
      "num_input_tokens_seen": 3477208,
      "step": 5315
    },
    {
      "epoch": 2.788259958071279,
      "grad_norm": 0.9712128639221191,
      "learning_rate": 4.976409249267502e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 3480088,
      "step": 5320
    },
    {
      "epoch": 2.790880503144654,
      "grad_norm": 0.6206348538398743,
      "learning_rate": 4.976252279815863e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 3484504,
      "step": 5325
    },
    {
      "epoch": 2.7935010482180296,
      "grad_norm": 1.0001192092895508,
      "learning_rate": 4.976094792360498e-05,
      "loss": 0.3993,
      "num_input_tokens_seen": 3487000,
      "step": 5330
    },
    {
      "epoch": 2.7961215932914047,
      "grad_norm": 0.9135286211967468,
      "learning_rate": 4.975936786934351e-05,
      "loss": 0.3937,
      "num_input_tokens_seen": 3490712,
      "step": 5335
    },
    {
      "epoch": 2.79874213836478,
      "grad_norm": 0.870322048664093,
      "learning_rate": 4.975778263570475e-05,
      "loss": 0.4735,
      "num_input_tokens_seen": 3494296,
      "step": 5340
    },
    {
      "epoch": 2.801362683438155,
      "grad_norm": 0.7400071620941162,
      "learning_rate": 4.975619222302032e-05,
      "loss": 0.3891,
      "num_input_tokens_seen": 3498232,
      "step": 5345
    },
    {
      "epoch": 2.8039832285115303,
      "grad_norm": 1.6717215776443481,
      "learning_rate": 4.975459663162291e-05,
      "loss": 0.434,
      "num_input_tokens_seen": 3501080,
      "step": 5350
    },
    {
      "epoch": 2.8066037735849054,
      "grad_norm": 1.1250576972961426,
      "learning_rate": 4.9752995861846296e-05,
      "loss": 0.3474,
      "num_input_tokens_seen": 3504024,
      "step": 5355
    },
    {
      "epoch": 2.809224318658281,
      "grad_norm": 0.9073261022567749,
      "learning_rate": 4.975138991402535e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 3507256,
      "step": 5360
    },
    {
      "epoch": 2.811844863731656,
      "grad_norm": 0.7139402031898499,
      "learning_rate": 4.974977878849601e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 3510392,
      "step": 5365
    },
    {
      "epoch": 2.8144654088050314,
      "grad_norm": 1.061522126197815,
      "learning_rate": 4.9748162485595303e-05,
      "loss": 0.481,
      "num_input_tokens_seen": 3513272,
      "step": 5370
    },
    {
      "epoch": 2.8170859538784065,
      "grad_norm": 0.6802165508270264,
      "learning_rate": 4.974654100566135e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 3516312,
      "step": 5375
    },
    {
      "epoch": 2.819706498951782,
      "grad_norm": 1.214333415031433,
      "learning_rate": 4.974491434903335e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 3519032,
      "step": 5380
    },
    {
      "epoch": 2.8223270440251573,
      "grad_norm": 1.3036080598831177,
      "learning_rate": 4.9743282516051556e-05,
      "loss": 0.3452,
      "num_input_tokens_seen": 3521720,
      "step": 5385
    },
    {
      "epoch": 2.8249475890985325,
      "grad_norm": 1.2469098567962646,
      "learning_rate": 4.974164550705736e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 3524888,
      "step": 5390
    },
    {
      "epoch": 2.8275681341719077,
      "grad_norm": 0.5505228638648987,
      "learning_rate": 4.974000332239319e-05,
      "loss": 0.3753,
      "num_input_tokens_seen": 3529688,
      "step": 5395
    },
    {
      "epoch": 2.830188679245283,
      "grad_norm": 0.8741199970245361,
      "learning_rate": 4.973835596240257e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 3532440,
      "step": 5400
    },
    {
      "epoch": 2.832809224318658,
      "grad_norm": 0.9883480668067932,
      "learning_rate": 4.9736703427430114e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 3535288,
      "step": 5405
    },
    {
      "epoch": 2.8354297693920336,
      "grad_norm": 1.0739392042160034,
      "learning_rate": 4.9735045717821514e-05,
      "loss": 0.4647,
      "num_input_tokens_seen": 3538840,
      "step": 5410
    },
    {
      "epoch": 2.838050314465409,
      "grad_norm": 0.6862120032310486,
      "learning_rate": 4.973338283392354e-05,
      "loss": 0.3331,
      "num_input_tokens_seen": 3541880,
      "step": 5415
    },
    {
      "epoch": 2.840670859538784,
      "grad_norm": 0.9104738235473633,
      "learning_rate": 4.9731714776084046e-05,
      "loss": 0.4807,
      "num_input_tokens_seen": 3545208,
      "step": 5420
    },
    {
      "epoch": 2.8432914046121596,
      "grad_norm": 1.4339609146118164,
      "learning_rate": 4.973004154465198e-05,
      "loss": 0.3838,
      "num_input_tokens_seen": 3548120,
      "step": 5425
    },
    {
      "epoch": 2.8459119496855347,
      "grad_norm": 0.800495445728302,
      "learning_rate": 4.972836313997736e-05,
      "loss": 0.4421,
      "num_input_tokens_seen": 3551512,
      "step": 5430
    },
    {
      "epoch": 2.84853249475891,
      "grad_norm": 1.6909878253936768,
      "learning_rate": 4.9726679562411285e-05,
      "loss": 0.3668,
      "num_input_tokens_seen": 3554520,
      "step": 5435
    },
    {
      "epoch": 2.851153039832285,
      "grad_norm": 0.8565283417701721,
      "learning_rate": 4.972499081230595e-05,
      "loss": 0.5025,
      "num_input_tokens_seen": 3557208,
      "step": 5440
    },
    {
      "epoch": 2.8537735849056602,
      "grad_norm": 1.7085037231445312,
      "learning_rate": 4.9723296890014605e-05,
      "loss": 0.6034,
      "num_input_tokens_seen": 3560600,
      "step": 5445
    },
    {
      "epoch": 2.8563941299790354,
      "grad_norm": 0.6000375151634216,
      "learning_rate": 4.9721597795891605e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 3565752,
      "step": 5450
    },
    {
      "epoch": 2.859014675052411,
      "grad_norm": 0.783115804195404,
      "learning_rate": 4.9719893530292396e-05,
      "loss": 0.3677,
      "num_input_tokens_seen": 3569400,
      "step": 5455
    },
    {
      "epoch": 2.861635220125786,
      "grad_norm": 1.4003881216049194,
      "learning_rate": 4.9718184093573475e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 3572088,
      "step": 5460
    },
    {
      "epoch": 2.8642557651991614,
      "grad_norm": 0.6787271499633789,
      "learning_rate": 4.9716469486092435e-05,
      "loss": 0.5252,
      "num_input_tokens_seen": 3575032,
      "step": 5465
    },
    {
      "epoch": 2.8668763102725365,
      "grad_norm": 1.1111884117126465,
      "learning_rate": 4.971474970820797e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 3578232,
      "step": 5470
    },
    {
      "epoch": 2.869496855345912,
      "grad_norm": 1.0108882188796997,
      "learning_rate": 4.971302476027982e-05,
      "loss": 0.4599,
      "num_input_tokens_seen": 3581304,
      "step": 5475
    },
    {
      "epoch": 2.8721174004192873,
      "grad_norm": 0.8265075087547302,
      "learning_rate": 4.971129464266884e-05,
      "loss": 0.3561,
      "num_input_tokens_seen": 3584664,
      "step": 5480
    },
    {
      "epoch": 2.8747379454926625,
      "grad_norm": 0.6670494675636292,
      "learning_rate": 4.970955935573693e-05,
      "loss": 0.5177,
      "num_input_tokens_seen": 3587544,
      "step": 5485
    },
    {
      "epoch": 2.8773584905660377,
      "grad_norm": 1.2708839178085327,
      "learning_rate": 4.9707818899847106e-05,
      "loss": 0.4722,
      "num_input_tokens_seen": 3590360,
      "step": 5490
    },
    {
      "epoch": 2.879979035639413,
      "grad_norm": 0.854418933391571,
      "learning_rate": 4.970607327536346e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 3597208,
      "step": 5495
    },
    {
      "epoch": 2.882599580712788,
      "grad_norm": 1.3330143690109253,
      "learning_rate": 4.970432248265114e-05,
      "loss": 0.3601,
      "num_input_tokens_seen": 3600568,
      "step": 5500
    },
    {
      "epoch": 2.8852201257861636,
      "grad_norm": 2.350616693496704,
      "learning_rate": 4.97025665220764e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 3603480,
      "step": 5505
    },
    {
      "epoch": 2.8878406708595388,
      "grad_norm": 0.8330750465393066,
      "learning_rate": 4.970080539400656e-05,
      "loss": 0.4958,
      "num_input_tokens_seen": 3606360,
      "step": 5510
    },
    {
      "epoch": 2.890461215932914,
      "grad_norm": 1.5194817781448364,
      "learning_rate": 4.969903909881004e-05,
      "loss": 0.4084,
      "num_input_tokens_seen": 3608856,
      "step": 5515
    },
    {
      "epoch": 2.8930817610062896,
      "grad_norm": 0.7303959727287292,
      "learning_rate": 4.9697267636856317e-05,
      "loss": 0.3629,
      "num_input_tokens_seen": 3611384,
      "step": 5520
    },
    {
      "epoch": 2.8957023060796647,
      "grad_norm": 0.9123006463050842,
      "learning_rate": 4.969549100851597e-05,
      "loss": 0.6605,
      "num_input_tokens_seen": 3614488,
      "step": 5525
    },
    {
      "epoch": 2.89832285115304,
      "grad_norm": 1.631321907043457,
      "learning_rate": 4.9693709214160645e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 3617080,
      "step": 5530
    },
    {
      "epoch": 2.900943396226415,
      "grad_norm": 1.3066174983978271,
      "learning_rate": 4.969192225416307e-05,
      "loss": 0.5431,
      "num_input_tokens_seen": 3620728,
      "step": 5535
    },
    {
      "epoch": 2.9035639412997902,
      "grad_norm": 0.8833510279655457,
      "learning_rate": 4.969013012889706e-05,
      "loss": 0.4894,
      "num_input_tokens_seen": 3624184,
      "step": 5540
    },
    {
      "epoch": 2.9061844863731654,
      "grad_norm": 0.5202651619911194,
      "learning_rate": 4.9688332838737504e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 3628856,
      "step": 5545
    },
    {
      "epoch": 2.908805031446541,
      "grad_norm": 0.691677451133728,
      "learning_rate": 4.9686530384060386e-05,
      "loss": 0.5035,
      "num_input_tokens_seen": 3631800,
      "step": 5550
    },
    {
      "epoch": 2.911425576519916,
      "grad_norm": 1.8115389347076416,
      "learning_rate": 4.968472276524275e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 3634680,
      "step": 5555
    },
    {
      "epoch": 2.9140461215932913,
      "grad_norm": 0.8552885055541992,
      "learning_rate": 4.9682909982662725e-05,
      "loss": 0.371,
      "num_input_tokens_seen": 3638264,
      "step": 5560
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 1.031092643737793,
      "learning_rate": 4.968109203669953e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 3641240,
      "step": 5565
    },
    {
      "epoch": 2.919287211740042,
      "grad_norm": 0.6744198799133301,
      "learning_rate": 4.967926892773346e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 3643640,
      "step": 5570
    },
    {
      "epoch": 2.9219077568134173,
      "grad_norm": 0.6631165742874146,
      "learning_rate": 4.967744065614588e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 3648088,
      "step": 5575
    },
    {
      "epoch": 2.9245283018867925,
      "grad_norm": 0.5809226632118225,
      "learning_rate": 4.9675607222319255e-05,
      "loss": 0.4027,
      "num_input_tokens_seen": 3651352,
      "step": 5580
    },
    {
      "epoch": 2.9271488469601676,
      "grad_norm": 0.7530726194381714,
      "learning_rate": 4.967376862663711e-05,
      "loss": 0.4693,
      "num_input_tokens_seen": 3655288,
      "step": 5585
    },
    {
      "epoch": 2.929769392033543,
      "grad_norm": 0.9450503587722778,
      "learning_rate": 4.9671924869484064e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 3657624,
      "step": 5590
    },
    {
      "epoch": 2.932389937106918,
      "grad_norm": 1.2568992376327515,
      "learning_rate": 4.967007595124581e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 3660472,
      "step": 5595
    },
    {
      "epoch": 2.9350104821802936,
      "grad_norm": 0.8384485840797424,
      "learning_rate": 4.966822187230911e-05,
      "loss": 0.5024,
      "num_input_tokens_seen": 3663992,
      "step": 5600
    },
    {
      "epoch": 2.9376310272536688,
      "grad_norm": 1.1213929653167725,
      "learning_rate": 4.966636263306183e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 3666424,
      "step": 5605
    },
    {
      "epoch": 2.940251572327044,
      "grad_norm": 1.203474760055542,
      "learning_rate": 4.9664498233892895e-05,
      "loss": 0.5109,
      "num_input_tokens_seen": 3669400,
      "step": 5610
    },
    {
      "epoch": 2.9428721174004195,
      "grad_norm": 0.9240164160728455,
      "learning_rate": 4.966262867519232e-05,
      "loss": 0.5768,
      "num_input_tokens_seen": 3672440,
      "step": 5615
    },
    {
      "epoch": 2.9454926624737947,
      "grad_norm": 1.5098598003387451,
      "learning_rate": 4.966075395735119e-05,
      "loss": 0.3352,
      "num_input_tokens_seen": 3675704,
      "step": 5620
    },
    {
      "epoch": 2.94811320754717,
      "grad_norm": 1.2156813144683838,
      "learning_rate": 4.965887408076167e-05,
      "loss": 0.5829,
      "num_input_tokens_seen": 3679160,
      "step": 5625
    },
    {
      "epoch": 2.950733752620545,
      "grad_norm": 0.7599944472312927,
      "learning_rate": 4.965698904581703e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 3682328,
      "step": 5630
    },
    {
      "epoch": 2.95335429769392,
      "grad_norm": 0.9643664360046387,
      "learning_rate": 4.965509885291157e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 3685240,
      "step": 5635
    },
    {
      "epoch": 2.9559748427672954,
      "grad_norm": 1.7181671857833862,
      "learning_rate": 4.9653203502440716e-05,
      "loss": 0.5825,
      "num_input_tokens_seen": 3687960,
      "step": 5640
    },
    {
      "epoch": 2.958595387840671,
      "grad_norm": 0.8275561928749084,
      "learning_rate": 4.965130299480095e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 3691704,
      "step": 5645
    },
    {
      "epoch": 2.961215932914046,
      "grad_norm": 1.3291679620742798,
      "learning_rate": 4.9649397330389834e-05,
      "loss": 0.4874,
      "num_input_tokens_seen": 3695416,
      "step": 5650
    },
    {
      "epoch": 2.9638364779874213,
      "grad_norm": 1.1810803413391113,
      "learning_rate": 4.964748650960601e-05,
      "loss": 0.4225,
      "num_input_tokens_seen": 3698168,
      "step": 5655
    },
    {
      "epoch": 2.9664570230607965,
      "grad_norm": 1.8437060117721558,
      "learning_rate": 4.96455705328492e-05,
      "loss": 0.5698,
      "num_input_tokens_seen": 3700984,
      "step": 5660
    },
    {
      "epoch": 2.969077568134172,
      "grad_norm": 0.980815052986145,
      "learning_rate": 4.964364940052021e-05,
      "loss": 0.6134,
      "num_input_tokens_seen": 3704472,
      "step": 5665
    },
    {
      "epoch": 2.9716981132075473,
      "grad_norm": 0.9091605544090271,
      "learning_rate": 4.964172311302091e-05,
      "loss": 0.3615,
      "num_input_tokens_seen": 3707960,
      "step": 5670
    },
    {
      "epoch": 2.9743186582809225,
      "grad_norm": 1.2400377988815308,
      "learning_rate": 4.9639791670754265e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 3710520,
      "step": 5675
    },
    {
      "epoch": 2.9769392033542976,
      "grad_norm": 0.9985286593437195,
      "learning_rate": 4.9637855074124305e-05,
      "loss": 0.455,
      "num_input_tokens_seen": 3713688,
      "step": 5680
    },
    {
      "epoch": 2.979559748427673,
      "grad_norm": 0.8987990617752075,
      "learning_rate": 4.963591332353615e-05,
      "loss": 0.4238,
      "num_input_tokens_seen": 3717304,
      "step": 5685
    },
    {
      "epoch": 2.982180293501048,
      "grad_norm": 0.8297590017318726,
      "learning_rate": 4.9633966419396e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 3720888,
      "step": 5690
    },
    {
      "epoch": 2.9848008385744236,
      "grad_norm": 0.9446940422058105,
      "learning_rate": 4.96320143621111e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 3724888,
      "step": 5695
    },
    {
      "epoch": 2.9874213836477987,
      "grad_norm": 1.2296825647354126,
      "learning_rate": 4.963005715208981e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 3727288,
      "step": 5700
    },
    {
      "epoch": 2.990041928721174,
      "grad_norm": 1.2910012006759644,
      "learning_rate": 4.962809478974157e-05,
      "loss": 0.4945,
      "num_input_tokens_seen": 3730328,
      "step": 5705
    },
    {
      "epoch": 2.9926624737945495,
      "grad_norm": 1.0771610736846924,
      "learning_rate": 4.962612727547686e-05,
      "loss": 0.3857,
      "num_input_tokens_seen": 3733912,
      "step": 5710
    },
    {
      "epoch": 2.9952830188679247,
      "grad_norm": 0.9960704445838928,
      "learning_rate": 4.962415460970728e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 3737464,
      "step": 5715
    },
    {
      "epoch": 2.9979035639413,
      "grad_norm": 0.7147172689437866,
      "learning_rate": 4.962217679284548e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 3740152,
      "step": 5720
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.4691515564918518,
      "eval_runtime": 11.633,
      "eval_samples_per_second": 72.896,
      "eval_steps_per_second": 18.224,
      "num_input_tokens_seen": 3742552,
      "step": 5724
    },
    {
      "epoch": 3.000524109014675,
      "grad_norm": 1.4770927429199219,
      "learning_rate": 4.962019382530521e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 3742968,
      "step": 5725
    },
    {
      "epoch": 3.00314465408805,
      "grad_norm": 0.7010685801506042,
      "learning_rate": 4.9618205707501256e-05,
      "loss": 0.4309,
      "num_input_tokens_seen": 3745880,
      "step": 5730
    },
    {
      "epoch": 3.0057651991614254,
      "grad_norm": 0.679942786693573,
      "learning_rate": 4.961621243984954e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 3749464,
      "step": 5735
    },
    {
      "epoch": 3.008385744234801,
      "grad_norm": 1.348645567893982,
      "learning_rate": 4.9614214022767014e-05,
      "loss": 0.3473,
      "num_input_tokens_seen": 3752248,
      "step": 5740
    },
    {
      "epoch": 3.011006289308176,
      "grad_norm": 1.0565544366836548,
      "learning_rate": 4.9612210456671735e-05,
      "loss": 0.4173,
      "num_input_tokens_seen": 3755480,
      "step": 5745
    },
    {
      "epoch": 3.0136268343815513,
      "grad_norm": 0.7455442547798157,
      "learning_rate": 4.961020174198281e-05,
      "loss": 0.3776,
      "num_input_tokens_seen": 3758040,
      "step": 5750
    },
    {
      "epoch": 3.0162473794549265,
      "grad_norm": 1.5643376111984253,
      "learning_rate": 4.9608187879120454e-05,
      "loss": 0.3944,
      "num_input_tokens_seen": 3761368,
      "step": 5755
    },
    {
      "epoch": 3.018867924528302,
      "grad_norm": 1.1249891519546509,
      "learning_rate": 4.9606168868505936e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 3764824,
      "step": 5760
    },
    {
      "epoch": 3.0214884696016773,
      "grad_norm": 0.6320895552635193,
      "learning_rate": 4.9604144710561616e-05,
      "loss": 0.3229,
      "num_input_tokens_seen": 3768056,
      "step": 5765
    },
    {
      "epoch": 3.0241090146750524,
      "grad_norm": 1.2398571968078613,
      "learning_rate": 4.960211540571092e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 3770168,
      "step": 5770
    },
    {
      "epoch": 3.0267295597484276,
      "grad_norm": 0.713738739490509,
      "learning_rate": 4.960008095437836e-05,
      "loss": 0.4203,
      "num_input_tokens_seen": 3773112,
      "step": 5775
    },
    {
      "epoch": 3.029350104821803,
      "grad_norm": 0.7344282269477844,
      "learning_rate": 4.959804135698951e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 3776568,
      "step": 5780
    },
    {
      "epoch": 3.0319706498951784,
      "grad_norm": 0.6748427748680115,
      "learning_rate": 4.959599661397104e-05,
      "loss": 0.5663,
      "num_input_tokens_seen": 3780920,
      "step": 5785
    },
    {
      "epoch": 3.0345911949685536,
      "grad_norm": 0.995612382888794,
      "learning_rate": 4.959394672575068e-05,
      "loss": 0.4622,
      "num_input_tokens_seen": 3783960,
      "step": 5790
    },
    {
      "epoch": 3.0372117400419287,
      "grad_norm": 1.129480004310608,
      "learning_rate": 4.959189169275725e-05,
      "loss": 0.3683,
      "num_input_tokens_seen": 3786968,
      "step": 5795
    },
    {
      "epoch": 3.039832285115304,
      "grad_norm": 0.6761322617530823,
      "learning_rate": 4.958983151542065e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 3790808,
      "step": 5800
    },
    {
      "epoch": 3.042452830188679,
      "grad_norm": 0.7117992639541626,
      "learning_rate": 4.958776619417182e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 3794264,
      "step": 5805
    },
    {
      "epoch": 3.0450733752620547,
      "grad_norm": 0.8598412871360779,
      "learning_rate": 4.9585695729442814e-05,
      "loss": 0.4381,
      "num_input_tokens_seen": 3797496,
      "step": 5810
    },
    {
      "epoch": 3.04769392033543,
      "grad_norm": 1.2457610368728638,
      "learning_rate": 4.9583620121666754e-05,
      "loss": 0.5496,
      "num_input_tokens_seen": 3800888,
      "step": 5815
    },
    {
      "epoch": 3.050314465408805,
      "grad_norm": 0.8220885992050171,
      "learning_rate": 4.958153937127783e-05,
      "loss": 0.3588,
      "num_input_tokens_seen": 3803512,
      "step": 5820
    },
    {
      "epoch": 3.05293501048218,
      "grad_norm": 1.3170619010925293,
      "learning_rate": 4.95794534787113e-05,
      "loss": 0.4768,
      "num_input_tokens_seen": 3806552,
      "step": 5825
    },
    {
      "epoch": 3.0555555555555554,
      "grad_norm": 0.5071884393692017,
      "learning_rate": 4.957736244440353e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 3810648,
      "step": 5830
    },
    {
      "epoch": 3.058176100628931,
      "grad_norm": 0.8427143096923828,
      "learning_rate": 4.957526626879193e-05,
      "loss": 0.345,
      "num_input_tokens_seen": 3813976,
      "step": 5835
    },
    {
      "epoch": 3.060796645702306,
      "grad_norm": 2.4423635005950928,
      "learning_rate": 4.957316495231499e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 3816824,
      "step": 5840
    },
    {
      "epoch": 3.0634171907756813,
      "grad_norm": 0.7100880146026611,
      "learning_rate": 4.9571058495412294e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 3820376,
      "step": 5845
    },
    {
      "epoch": 3.0660377358490565,
      "grad_norm": 1.09442138671875,
      "learning_rate": 4.9568946898524485e-05,
      "loss": 0.3612,
      "num_input_tokens_seen": 3823736,
      "step": 5850
    },
    {
      "epoch": 3.068658280922432,
      "grad_norm": 0.9921320080757141,
      "learning_rate": 4.956683016209327e-05,
      "loss": 0.5631,
      "num_input_tokens_seen": 3826104,
      "step": 5855
    },
    {
      "epoch": 3.0712788259958073,
      "grad_norm": 1.0051078796386719,
      "learning_rate": 4.956470828656147e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 3828632,
      "step": 5860
    },
    {
      "epoch": 3.0738993710691824,
      "grad_norm": 0.8631693720817566,
      "learning_rate": 4.956258127237294e-05,
      "loss": 0.2636,
      "num_input_tokens_seen": 3831576,
      "step": 5865
    },
    {
      "epoch": 3.0765199161425576,
      "grad_norm": 0.7383410930633545,
      "learning_rate": 4.956044911997263e-05,
      "loss": 0.3099,
      "num_input_tokens_seen": 3834392,
      "step": 5870
    },
    {
      "epoch": 3.0791404612159328,
      "grad_norm": 1.3220387697219849,
      "learning_rate": 4.955831182980657e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 3837144,
      "step": 5875
    },
    {
      "epoch": 3.0817610062893084,
      "grad_norm": 0.9474557638168335,
      "learning_rate": 4.955616940232185e-05,
      "loss": 0.4522,
      "num_input_tokens_seen": 3840504,
      "step": 5880
    },
    {
      "epoch": 3.0843815513626835,
      "grad_norm": 0.7037832140922546,
      "learning_rate": 4.9554021837966636e-05,
      "loss": 0.4518,
      "num_input_tokens_seen": 3844408,
      "step": 5885
    },
    {
      "epoch": 3.0870020964360587,
      "grad_norm": 1.1962401866912842,
      "learning_rate": 4.955186913719019e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 3847736,
      "step": 5890
    },
    {
      "epoch": 3.089622641509434,
      "grad_norm": 0.6811770796775818,
      "learning_rate": 4.954971130044281e-05,
      "loss": 0.4123,
      "num_input_tokens_seen": 3850360,
      "step": 5895
    },
    {
      "epoch": 3.092243186582809,
      "grad_norm": 0.7133316397666931,
      "learning_rate": 4.954754832817592e-05,
      "loss": 0.3925,
      "num_input_tokens_seen": 3853624,
      "step": 5900
    },
    {
      "epoch": 3.0948637316561847,
      "grad_norm": 1.013593077659607,
      "learning_rate": 4.954538022084196e-05,
      "loss": 0.3107,
      "num_input_tokens_seen": 3856664,
      "step": 5905
    },
    {
      "epoch": 3.09748427672956,
      "grad_norm": 1.1806660890579224,
      "learning_rate": 4.9543206978894496e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 3859640,
      "step": 5910
    },
    {
      "epoch": 3.100104821802935,
      "grad_norm": 1.2220680713653564,
      "learning_rate": 4.954102860278813e-05,
      "loss": 0.3291,
      "num_input_tokens_seen": 3862712,
      "step": 5915
    },
    {
      "epoch": 3.10272536687631,
      "grad_norm": 0.9284286499023438,
      "learning_rate": 4.9538845092978567e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 3866328,
      "step": 5920
    },
    {
      "epoch": 3.1053459119496853,
      "grad_norm": 0.8521385788917542,
      "learning_rate": 4.953665644992256e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 3869592,
      "step": 5925
    },
    {
      "epoch": 3.107966457023061,
      "grad_norm": 1.376509428024292,
      "learning_rate": 4.953446267407795e-05,
      "loss": 0.3228,
      "num_input_tokens_seen": 3875032,
      "step": 5930
    },
    {
      "epoch": 3.110587002096436,
      "grad_norm": 0.6410566568374634,
      "learning_rate": 4.953226376590365e-05,
      "loss": 0.465,
      "num_input_tokens_seen": 3879288,
      "step": 5935
    },
    {
      "epoch": 3.1132075471698113,
      "grad_norm": 0.7748413681983948,
      "learning_rate": 4.953005972585965e-05,
      "loss": 0.2998,
      "num_input_tokens_seen": 3882936,
      "step": 5940
    },
    {
      "epoch": 3.1158280922431865,
      "grad_norm": 0.6489357352256775,
      "learning_rate": 4.952785055440701e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 3886296,
      "step": 5945
    },
    {
      "epoch": 3.1184486373165616,
      "grad_norm": 1.6252607107162476,
      "learning_rate": 4.9525636252007864e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 3889816,
      "step": 5950
    },
    {
      "epoch": 3.1210691823899372,
      "grad_norm": 0.8534907698631287,
      "learning_rate": 4.9523416819125425e-05,
      "loss": 0.3279,
      "num_input_tokens_seen": 3892888,
      "step": 5955
    },
    {
      "epoch": 3.1236897274633124,
      "grad_norm": 2.764716148376465,
      "learning_rate": 4.952119225622396e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 3896152,
      "step": 5960
    },
    {
      "epoch": 3.1263102725366876,
      "grad_norm": 0.46571052074432373,
      "learning_rate": 4.951896256376882e-05,
      "loss": 0.2892,
      "num_input_tokens_seen": 3898776,
      "step": 5965
    },
    {
      "epoch": 3.1289308176100628,
      "grad_norm": 1.1511560678482056,
      "learning_rate": 4.9516727742226444e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 3901912,
      "step": 5970
    },
    {
      "epoch": 3.131551362683438,
      "grad_norm": 0.7660908102989197,
      "learning_rate": 4.9514487792064323e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 3905688,
      "step": 5975
    },
    {
      "epoch": 3.1341719077568135,
      "grad_norm": 1.037286400794983,
      "learning_rate": 4.951224271375104e-05,
      "loss": 0.4183,
      "num_input_tokens_seen": 3908888,
      "step": 5980
    },
    {
      "epoch": 3.1367924528301887,
      "grad_norm": 1.4576787948608398,
      "learning_rate": 4.950999250775623e-05,
      "loss": 0.3486,
      "num_input_tokens_seen": 3912280,
      "step": 5985
    },
    {
      "epoch": 3.139412997903564,
      "grad_norm": 2.1368067264556885,
      "learning_rate": 4.950773717455061e-05,
      "loss": 0.363,
      "num_input_tokens_seen": 3915256,
      "step": 5990
    },
    {
      "epoch": 3.142033542976939,
      "grad_norm": 0.9620910882949829,
      "learning_rate": 4.950547671460596e-05,
      "loss": 0.3716,
      "num_input_tokens_seen": 3917912,
      "step": 5995
    },
    {
      "epoch": 3.1446540880503147,
      "grad_norm": 0.9549274444580078,
      "learning_rate": 4.950321112839517e-05,
      "loss": 0.4402,
      "num_input_tokens_seen": 3920824,
      "step": 6000
    },
    {
      "epoch": 3.14727463312369,
      "grad_norm": 1.274253249168396,
      "learning_rate": 4.9500940416392156e-05,
      "loss": 0.3489,
      "num_input_tokens_seen": 3923736,
      "step": 6005
    },
    {
      "epoch": 3.149895178197065,
      "grad_norm": 0.7367783188819885,
      "learning_rate": 4.949866457907193e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 3926584,
      "step": 6010
    },
    {
      "epoch": 3.15251572327044,
      "grad_norm": 1.0366238355636597,
      "learning_rate": 4.9496383616910565e-05,
      "loss": 0.5519,
      "num_input_tokens_seen": 3929432,
      "step": 6015
    },
    {
      "epoch": 3.1551362683438153,
      "grad_norm": 0.9166523814201355,
      "learning_rate": 4.949409753038522e-05,
      "loss": 0.4338,
      "num_input_tokens_seen": 3932088,
      "step": 6020
    },
    {
      "epoch": 3.157756813417191,
      "grad_norm": 1.1643489599227905,
      "learning_rate": 4.949180631997411e-05,
      "loss": 0.3634,
      "num_input_tokens_seen": 3935384,
      "step": 6025
    },
    {
      "epoch": 3.160377358490566,
      "grad_norm": 1.0934761762619019,
      "learning_rate": 4.948950998615654e-05,
      "loss": 0.3952,
      "num_input_tokens_seen": 3938424,
      "step": 6030
    },
    {
      "epoch": 3.1629979035639413,
      "grad_norm": 1.1742889881134033,
      "learning_rate": 4.9487208529412876e-05,
      "loss": 0.4435,
      "num_input_tokens_seen": 3940984,
      "step": 6035
    },
    {
      "epoch": 3.1656184486373165,
      "grad_norm": 1.05732262134552,
      "learning_rate": 4.9484901950224546e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 3945144,
      "step": 6040
    },
    {
      "epoch": 3.1682389937106916,
      "grad_norm": 0.705045223236084,
      "learning_rate": 4.948259024907407e-05,
      "loss": 0.3984,
      "num_input_tokens_seen": 3948344,
      "step": 6045
    },
    {
      "epoch": 3.1708595387840672,
      "grad_norm": 1.0089046955108643,
      "learning_rate": 4.948027342644502e-05,
      "loss": 0.4742,
      "num_input_tokens_seen": 3951000,
      "step": 6050
    },
    {
      "epoch": 3.1734800838574424,
      "grad_norm": 1.0154316425323486,
      "learning_rate": 4.947795148282206e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 3954584,
      "step": 6055
    },
    {
      "epoch": 3.1761006289308176,
      "grad_norm": 0.8111104965209961,
      "learning_rate": 4.947562441869091e-05,
      "loss": 0.271,
      "num_input_tokens_seen": 3957944,
      "step": 6060
    },
    {
      "epoch": 3.1787211740041927,
      "grad_norm": 0.7711851000785828,
      "learning_rate": 4.947329223453836e-05,
      "loss": 0.3544,
      "num_input_tokens_seen": 3961816,
      "step": 6065
    },
    {
      "epoch": 3.181341719077568,
      "grad_norm": 0.7113642692565918,
      "learning_rate": 4.947095493085229e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 3964824,
      "step": 6070
    },
    {
      "epoch": 3.1839622641509435,
      "grad_norm": 1.2744046449661255,
      "learning_rate": 4.9468612508121616e-05,
      "loss": 0.3228,
      "num_input_tokens_seen": 3968056,
      "step": 6075
    },
    {
      "epoch": 3.1865828092243187,
      "grad_norm": 1.002604603767395,
      "learning_rate": 4.9466264966836365e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 3970424,
      "step": 6080
    },
    {
      "epoch": 3.189203354297694,
      "grad_norm": 1.4560989141464233,
      "learning_rate": 4.94639123074876e-05,
      "loss": 0.3751,
      "num_input_tokens_seen": 3973208,
      "step": 6085
    },
    {
      "epoch": 3.191823899371069,
      "grad_norm": 1.0693790912628174,
      "learning_rate": 4.946155453056749e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 3976216,
      "step": 6090
    },
    {
      "epoch": 3.1944444444444446,
      "grad_norm": 1.3012446165084839,
      "learning_rate": 4.945919163656923e-05,
      "loss": 0.4785,
      "num_input_tokens_seen": 3980824,
      "step": 6095
    },
    {
      "epoch": 3.19706498951782,
      "grad_norm": 1.0188781023025513,
      "learning_rate": 4.945682362598715e-05,
      "loss": 0.4607,
      "num_input_tokens_seen": 3987480,
      "step": 6100
    },
    {
      "epoch": 3.199685534591195,
      "grad_norm": 1.1343061923980713,
      "learning_rate": 4.945445049931656e-05,
      "loss": 0.3952,
      "num_input_tokens_seen": 3991288,
      "step": 6105
    },
    {
      "epoch": 3.20230607966457,
      "grad_norm": 0.8765153288841248,
      "learning_rate": 4.945207225705393e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 3993848,
      "step": 6110
    },
    {
      "epoch": 3.2049266247379453,
      "grad_norm": 0.9571018218994141,
      "learning_rate": 4.944968889969674e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 3997848,
      "step": 6115
    },
    {
      "epoch": 3.207547169811321,
      "grad_norm": 1.2624537944793701,
      "learning_rate": 4.944730042774358e-05,
      "loss": 0.3466,
      "num_input_tokens_seen": 4001784,
      "step": 6120
    },
    {
      "epoch": 3.210167714884696,
      "grad_norm": 0.7199388146400452,
      "learning_rate": 4.944490684169408e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 4005048,
      "step": 6125
    },
    {
      "epoch": 3.2127882599580713,
      "grad_norm": 1.2392172813415527,
      "learning_rate": 4.9442508142048945e-05,
      "loss": 0.4376,
      "num_input_tokens_seen": 4007864,
      "step": 6130
    },
    {
      "epoch": 3.2154088050314464,
      "grad_norm": 0.9141506552696228,
      "learning_rate": 4.9440104329309956e-05,
      "loss": 0.4366,
      "num_input_tokens_seen": 4011832,
      "step": 6135
    },
    {
      "epoch": 3.2180293501048216,
      "grad_norm": 1.1276228427886963,
      "learning_rate": 4.943769540397998e-05,
      "loss": 0.3964,
      "num_input_tokens_seen": 4015032,
      "step": 6140
    },
    {
      "epoch": 3.220649895178197,
      "grad_norm": 1.022176742553711,
      "learning_rate": 4.943528136656292e-05,
      "loss": 0.548,
      "num_input_tokens_seen": 4017432,
      "step": 6145
    },
    {
      "epoch": 3.2232704402515724,
      "grad_norm": 1.5977028608322144,
      "learning_rate": 4.943286221756377e-05,
      "loss": 0.3397,
      "num_input_tokens_seen": 4020216,
      "step": 6150
    },
    {
      "epoch": 3.2258909853249476,
      "grad_norm": 1.0155227184295654,
      "learning_rate": 4.9430437957488586e-05,
      "loss": 0.4598,
      "num_input_tokens_seen": 4023096,
      "step": 6155
    },
    {
      "epoch": 3.2285115303983227,
      "grad_norm": 1.7813668251037598,
      "learning_rate": 4.94280085868445e-05,
      "loss": 0.288,
      "num_input_tokens_seen": 4026712,
      "step": 6160
    },
    {
      "epoch": 3.231132075471698,
      "grad_norm": 1.3633891344070435,
      "learning_rate": 4.942557410613972e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 4030872,
      "step": 6165
    },
    {
      "epoch": 3.2337526205450735,
      "grad_norm": 1.1039844751358032,
      "learning_rate": 4.942313451588348e-05,
      "loss": 0.3841,
      "num_input_tokens_seen": 4034744,
      "step": 6170
    },
    {
      "epoch": 3.2363731656184487,
      "grad_norm": 1.700437307357788,
      "learning_rate": 4.942068981658614e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 4037752,
      "step": 6175
    },
    {
      "epoch": 3.238993710691824,
      "grad_norm": 1.6665750741958618,
      "learning_rate": 4.9418240008759097e-05,
      "loss": 0.3363,
      "num_input_tokens_seen": 4041048,
      "step": 6180
    },
    {
      "epoch": 3.241614255765199,
      "grad_norm": 1.3957469463348389,
      "learning_rate": 4.9415785092914826e-05,
      "loss": 0.4944,
      "num_input_tokens_seen": 4043832,
      "step": 6185
    },
    {
      "epoch": 3.2442348008385746,
      "grad_norm": 1.047013521194458,
      "learning_rate": 4.941332506956686e-05,
      "loss": 0.354,
      "num_input_tokens_seen": 4047512,
      "step": 6190
    },
    {
      "epoch": 3.24685534591195,
      "grad_norm": 0.9742141366004944,
      "learning_rate": 4.941085993922981e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 4050584,
      "step": 6195
    },
    {
      "epoch": 3.249475890985325,
      "grad_norm": 0.8751751184463501,
      "learning_rate": 4.9408389702419354e-05,
      "loss": 0.3208,
      "num_input_tokens_seen": 4053752,
      "step": 6200
    },
    {
      "epoch": 3.2520964360587,
      "grad_norm": 1.4667249917984009,
      "learning_rate": 4.940591435965224e-05,
      "loss": 0.4292,
      "num_input_tokens_seen": 4056824,
      "step": 6205
    },
    {
      "epoch": 3.2547169811320753,
      "grad_norm": 2.2152256965637207,
      "learning_rate": 4.9403433911446285e-05,
      "loss": 0.5357,
      "num_input_tokens_seen": 4059864,
      "step": 6210
    },
    {
      "epoch": 3.257337526205451,
      "grad_norm": 1.4078154563903809,
      "learning_rate": 4.9400948358320356e-05,
      "loss": 0.3757,
      "num_input_tokens_seen": 4063416,
      "step": 6215
    },
    {
      "epoch": 3.259958071278826,
      "grad_norm": 0.5578177571296692,
      "learning_rate": 4.939845770079442e-05,
      "loss": 0.354,
      "num_input_tokens_seen": 4066264,
      "step": 6220
    },
    {
      "epoch": 3.2625786163522013,
      "grad_norm": 4.674698352813721,
      "learning_rate": 4.9395961939389475e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 4068664,
      "step": 6225
    },
    {
      "epoch": 3.2651991614255764,
      "grad_norm": 0.9690462350845337,
      "learning_rate": 4.939346107462762e-05,
      "loss": 0.6004,
      "num_input_tokens_seen": 4072088,
      "step": 6230
    },
    {
      "epoch": 3.2678197064989516,
      "grad_norm": 2.2050063610076904,
      "learning_rate": 4.9390955107032e-05,
      "loss": 0.5474,
      "num_input_tokens_seen": 4075064,
      "step": 6235
    },
    {
      "epoch": 3.270440251572327,
      "grad_norm": 1.220058798789978,
      "learning_rate": 4.938844403712685e-05,
      "loss": 0.4212,
      "num_input_tokens_seen": 4077752,
      "step": 6240
    },
    {
      "epoch": 3.2730607966457024,
      "grad_norm": 1.0979998111724854,
      "learning_rate": 4.938592786543743e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 4080472,
      "step": 6245
    },
    {
      "epoch": 3.2756813417190775,
      "grad_norm": 1.0081955194473267,
      "learning_rate": 4.938340659249012e-05,
      "loss": 0.317,
      "num_input_tokens_seen": 4083384,
      "step": 6250
    },
    {
      "epoch": 3.2783018867924527,
      "grad_norm": 0.8592813014984131,
      "learning_rate": 4.938088021881233e-05,
      "loss": 0.4453,
      "num_input_tokens_seen": 4086488,
      "step": 6255
    },
    {
      "epoch": 3.280922431865828,
      "grad_norm": 0.882614016532898,
      "learning_rate": 4.937834874493255e-05,
      "loss": 0.2871,
      "num_input_tokens_seen": 4089464,
      "step": 6260
    },
    {
      "epoch": 3.2835429769392035,
      "grad_norm": 1.3703263998031616,
      "learning_rate": 4.9375812171380334e-05,
      "loss": 0.4575,
      "num_input_tokens_seen": 4092088,
      "step": 6265
    },
    {
      "epoch": 3.2861635220125787,
      "grad_norm": 1.2647825479507446,
      "learning_rate": 4.9373270498686306e-05,
      "loss": 0.3867,
      "num_input_tokens_seen": 4095544,
      "step": 6270
    },
    {
      "epoch": 3.288784067085954,
      "grad_norm": 0.7467767000198364,
      "learning_rate": 4.937072372738216e-05,
      "loss": 0.4252,
      "num_input_tokens_seen": 4098264,
      "step": 6275
    },
    {
      "epoch": 3.291404612159329,
      "grad_norm": 0.8326416611671448,
      "learning_rate": 4.9368171858000635e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 4100792,
      "step": 6280
    },
    {
      "epoch": 3.2940251572327046,
      "grad_norm": 0.9533650875091553,
      "learning_rate": 4.9365614891075576e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 4103896,
      "step": 6285
    },
    {
      "epoch": 3.29664570230608,
      "grad_norm": 0.7458609938621521,
      "learning_rate": 4.936305282714185e-05,
      "loss": 0.3177,
      "num_input_tokens_seen": 4106968,
      "step": 6290
    },
    {
      "epoch": 3.299266247379455,
      "grad_norm": 1.0000035762786865,
      "learning_rate": 4.936048566673542e-05,
      "loss": 0.475,
      "num_input_tokens_seen": 4110040,
      "step": 6295
    },
    {
      "epoch": 3.30188679245283,
      "grad_norm": 0.823353111743927,
      "learning_rate": 4.935791341039332e-05,
      "loss": 0.3578,
      "num_input_tokens_seen": 4112664,
      "step": 6300
    },
    {
      "epoch": 3.3045073375262053,
      "grad_norm": 1.6502560377120972,
      "learning_rate": 4.935533605865361e-05,
      "loss": 0.3982,
      "num_input_tokens_seen": 4115832,
      "step": 6305
    },
    {
      "epoch": 3.307127882599581,
      "grad_norm": 1.0610933303833008,
      "learning_rate": 4.935275361205547e-05,
      "loss": 0.3742,
      "num_input_tokens_seen": 4119160,
      "step": 6310
    },
    {
      "epoch": 3.309748427672956,
      "grad_norm": 2.1820929050445557,
      "learning_rate": 4.93501660711391e-05,
      "loss": 0.7292,
      "num_input_tokens_seen": 4122520,
      "step": 6315
    },
    {
      "epoch": 3.3123689727463312,
      "grad_norm": 1.7687222957611084,
      "learning_rate": 4.934757343644578e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 4124888,
      "step": 6320
    },
    {
      "epoch": 3.3149895178197064,
      "grad_norm": 1.16593337059021,
      "learning_rate": 4.934497570851788e-05,
      "loss": 0.4502,
      "num_input_tokens_seen": 4127832,
      "step": 6325
    },
    {
      "epoch": 3.3176100628930816,
      "grad_norm": 1.208551287651062,
      "learning_rate": 4.9342372887898804e-05,
      "loss": 0.5136,
      "num_input_tokens_seen": 4130232,
      "step": 6330
    },
    {
      "epoch": 3.320230607966457,
      "grad_norm": 1.3114453554153442,
      "learning_rate": 4.9339764975133026e-05,
      "loss": 0.3492,
      "num_input_tokens_seen": 4133080,
      "step": 6335
    },
    {
      "epoch": 3.3228511530398324,
      "grad_norm": 0.7902427315711975,
      "learning_rate": 4.93371519707661e-05,
      "loss": 0.3495,
      "num_input_tokens_seen": 4136536,
      "step": 6340
    },
    {
      "epoch": 3.3254716981132075,
      "grad_norm": 1.059352993965149,
      "learning_rate": 4.933453387534464e-05,
      "loss": 0.3661,
      "num_input_tokens_seen": 4139608,
      "step": 6345
    },
    {
      "epoch": 3.3280922431865827,
      "grad_norm": 2.1869497299194336,
      "learning_rate": 4.933191068941631e-05,
      "loss": 0.5084,
      "num_input_tokens_seen": 4142200,
      "step": 6350
    },
    {
      "epoch": 3.330712788259958,
      "grad_norm": 1.5502902269363403,
      "learning_rate": 4.932928241352987e-05,
      "loss": 0.7439,
      "num_input_tokens_seen": 4146200,
      "step": 6355
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.8472700715065002,
      "learning_rate": 4.9326649048235095e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 4149464,
      "step": 6360
    },
    {
      "epoch": 3.3359538784067087,
      "grad_norm": 0.8009471297264099,
      "learning_rate": 4.932401059408287e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 4151992,
      "step": 6365
    },
    {
      "epoch": 3.338574423480084,
      "grad_norm": 0.7408495545387268,
      "learning_rate": 4.932136705162514e-05,
      "loss": 0.4256,
      "num_input_tokens_seen": 4155576,
      "step": 6370
    },
    {
      "epoch": 3.341194968553459,
      "grad_norm": 1.2845807075500488,
      "learning_rate": 4.93187184214149e-05,
      "loss": 0.5054,
      "num_input_tokens_seen": 4159672,
      "step": 6375
    },
    {
      "epoch": 3.3438155136268346,
      "grad_norm": 0.8976457118988037,
      "learning_rate": 4.9316064704006205e-05,
      "loss": 0.3153,
      "num_input_tokens_seen": 4162360,
      "step": 6380
    },
    {
      "epoch": 3.3464360587002098,
      "grad_norm": 1.1488622426986694,
      "learning_rate": 4.931340589995419e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 4165304,
      "step": 6385
    },
    {
      "epoch": 3.349056603773585,
      "grad_norm": 0.7284955978393555,
      "learning_rate": 4.9310742009815035e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 4168728,
      "step": 6390
    },
    {
      "epoch": 3.35167714884696,
      "grad_norm": 1.5408676862716675,
      "learning_rate": 4.930807303414601e-05,
      "loss": 0.314,
      "num_input_tokens_seen": 4171960,
      "step": 6395
    },
    {
      "epoch": 3.3542976939203353,
      "grad_norm": 0.8286621570587158,
      "learning_rate": 4.930539897350542e-05,
      "loss": 0.3934,
      "num_input_tokens_seen": 4175544,
      "step": 6400
    },
    {
      "epoch": 3.3569182389937104,
      "grad_norm": 0.7095407247543335,
      "learning_rate": 4.930271982845266e-05,
      "loss": 0.3132,
      "num_input_tokens_seen": 4178264,
      "step": 6405
    },
    {
      "epoch": 3.359538784067086,
      "grad_norm": 1.254093050956726,
      "learning_rate": 4.930003559954817e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 4180984,
      "step": 6410
    },
    {
      "epoch": 3.3621593291404612,
      "grad_norm": 1.1733587980270386,
      "learning_rate": 4.929734628735346e-05,
      "loss": 0.488,
      "num_input_tokens_seen": 4183768,
      "step": 6415
    },
    {
      "epoch": 3.3647798742138364,
      "grad_norm": 1.241513729095459,
      "learning_rate": 4.9294651892431115e-05,
      "loss": 0.3503,
      "num_input_tokens_seen": 4187000,
      "step": 6420
    },
    {
      "epoch": 3.3674004192872116,
      "grad_norm": 1.5705666542053223,
      "learning_rate": 4.929195241534476e-05,
      "loss": 0.4009,
      "num_input_tokens_seen": 4189304,
      "step": 6425
    },
    {
      "epoch": 3.370020964360587,
      "grad_norm": 0.6689044833183289,
      "learning_rate": 4.9289247856659095e-05,
      "loss": 0.4293,
      "num_input_tokens_seen": 4193080,
      "step": 6430
    },
    {
      "epoch": 3.3726415094339623,
      "grad_norm": 1.1286836862564087,
      "learning_rate": 4.9286538216939883e-05,
      "loss": 0.4196,
      "num_input_tokens_seen": 4195608,
      "step": 6435
    },
    {
      "epoch": 3.3752620545073375,
      "grad_norm": 1.3013557195663452,
      "learning_rate": 4.9283823496753944e-05,
      "loss": 0.3269,
      "num_input_tokens_seen": 4198744,
      "step": 6440
    },
    {
      "epoch": 3.3778825995807127,
      "grad_norm": 1.5227488279342651,
      "learning_rate": 4.928110369666918e-05,
      "loss": 0.6344,
      "num_input_tokens_seen": 4202680,
      "step": 6445
    },
    {
      "epoch": 3.380503144654088,
      "grad_norm": 1.303765058517456,
      "learning_rate": 4.927837881725455e-05,
      "loss": 0.3989,
      "num_input_tokens_seen": 4205144,
      "step": 6450
    },
    {
      "epoch": 3.3831236897274635,
      "grad_norm": 1.1148513555526733,
      "learning_rate": 4.927564885908003e-05,
      "loss": 0.3733,
      "num_input_tokens_seen": 4209112,
      "step": 6455
    },
    {
      "epoch": 3.3857442348008386,
      "grad_norm": 0.9269514083862305,
      "learning_rate": 4.9272913822716735e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 4213048,
      "step": 6460
    },
    {
      "epoch": 3.388364779874214,
      "grad_norm": 0.6931104063987732,
      "learning_rate": 4.927017370873678e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 4216472,
      "step": 6465
    },
    {
      "epoch": 3.390985324947589,
      "grad_norm": 1.0566164255142212,
      "learning_rate": 4.926742851771338e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 4219672,
      "step": 6470
    },
    {
      "epoch": 3.3936058700209646,
      "grad_norm": 1.1621488332748413,
      "learning_rate": 4.926467825022079e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 4222872,
      "step": 6475
    },
    {
      "epoch": 3.3962264150943398,
      "grad_norm": 0.9229390025138855,
      "learning_rate": 4.926192290683434e-05,
      "loss": 0.4628,
      "num_input_tokens_seen": 4225528,
      "step": 6480
    },
    {
      "epoch": 3.398846960167715,
      "grad_norm": 1.7468476295471191,
      "learning_rate": 4.9259162488130404e-05,
      "loss": 0.3622,
      "num_input_tokens_seen": 4229144,
      "step": 6485
    },
    {
      "epoch": 3.40146750524109,
      "grad_norm": 0.9047337770462036,
      "learning_rate": 4.925639699468645e-05,
      "loss": 0.4356,
      "num_input_tokens_seen": 4232792,
      "step": 6490
    },
    {
      "epoch": 3.4040880503144653,
      "grad_norm": 0.8875569105148315,
      "learning_rate": 4.925362642708097e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 4236600,
      "step": 6495
    },
    {
      "epoch": 3.4067085953878404,
      "grad_norm": 1.14535391330719,
      "learning_rate": 4.9250850785893536e-05,
      "loss": 0.5304,
      "num_input_tokens_seen": 4239896,
      "step": 6500
    },
    {
      "epoch": 3.409329140461216,
      "grad_norm": 0.6103781461715698,
      "learning_rate": 4.924807007170479e-05,
      "loss": 0.4421,
      "num_input_tokens_seen": 4243640,
      "step": 6505
    },
    {
      "epoch": 3.411949685534591,
      "grad_norm": 0.9674192070960999,
      "learning_rate": 4.924528428509643e-05,
      "loss": 0.4698,
      "num_input_tokens_seen": 4247160,
      "step": 6510
    },
    {
      "epoch": 3.4145702306079664,
      "grad_norm": 0.935106098651886,
      "learning_rate": 4.9242493426651196e-05,
      "loss": 0.5015,
      "num_input_tokens_seen": 4251032,
      "step": 6515
    },
    {
      "epoch": 3.4171907756813416,
      "grad_norm": 1.3191595077514648,
      "learning_rate": 4.92396974969529e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 4254680,
      "step": 6520
    },
    {
      "epoch": 3.419811320754717,
      "grad_norm": 0.9108496904373169,
      "learning_rate": 4.923689649658644e-05,
      "loss": 0.3531,
      "num_input_tokens_seen": 4257720,
      "step": 6525
    },
    {
      "epoch": 3.4224318658280923,
      "grad_norm": 0.7642751336097717,
      "learning_rate": 4.923409042613774e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 4260696,
      "step": 6530
    },
    {
      "epoch": 3.4250524109014675,
      "grad_norm": 0.6801159977912903,
      "learning_rate": 4.92312792861938e-05,
      "loss": 0.3607,
      "num_input_tokens_seen": 4263416,
      "step": 6535
    },
    {
      "epoch": 3.4276729559748427,
      "grad_norm": 1.0453122854232788,
      "learning_rate": 4.922846307734268e-05,
      "loss": 0.5026,
      "num_input_tokens_seen": 4266232,
      "step": 6540
    },
    {
      "epoch": 3.430293501048218,
      "grad_norm": 1.696795105934143,
      "learning_rate": 4.9225641800173496e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 4269400,
      "step": 6545
    },
    {
      "epoch": 3.4329140461215935,
      "grad_norm": 1.1105738878250122,
      "learning_rate": 4.922281545527643e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 4271896,
      "step": 6550
    },
    {
      "epoch": 3.4355345911949686,
      "grad_norm": 0.913668692111969,
      "learning_rate": 4.9219984043242715e-05,
      "loss": 0.5797,
      "num_input_tokens_seen": 4275416,
      "step": 6555
    },
    {
      "epoch": 3.438155136268344,
      "grad_norm": 0.7191050052642822,
      "learning_rate": 4.921714756466466e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 4278936,
      "step": 6560
    },
    {
      "epoch": 3.440775681341719,
      "grad_norm": 1.0937926769256592,
      "learning_rate": 4.921430602013561e-05,
      "loss": 0.3977,
      "num_input_tokens_seen": 4282808,
      "step": 6565
    },
    {
      "epoch": 3.4433962264150946,
      "grad_norm": 0.8788013458251953,
      "learning_rate": 4.9211459410249995e-05,
      "loss": 0.4348,
      "num_input_tokens_seen": 4285400,
      "step": 6570
    },
    {
      "epoch": 3.4460167714884697,
      "grad_norm": 1.0953500270843506,
      "learning_rate": 4.92086077356033e-05,
      "loss": 0.3247,
      "num_input_tokens_seen": 4288280,
      "step": 6575
    },
    {
      "epoch": 3.448637316561845,
      "grad_norm": 0.767535388469696,
      "learning_rate": 4.920575099679204e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 4291576,
      "step": 6580
    },
    {
      "epoch": 3.45125786163522,
      "grad_norm": 1.0040791034698486,
      "learning_rate": 4.920288919441383e-05,
      "loss": 0.3889,
      "num_input_tokens_seen": 4295288,
      "step": 6585
    },
    {
      "epoch": 3.4538784067085953,
      "grad_norm": 1.0339033603668213,
      "learning_rate": 4.9200022329067316e-05,
      "loss": 0.4433,
      "num_input_tokens_seen": 4298808,
      "step": 6590
    },
    {
      "epoch": 3.4564989517819704,
      "grad_norm": 1.158129334449768,
      "learning_rate": 4.9197150401352224e-05,
      "loss": 0.3306,
      "num_input_tokens_seen": 4304280,
      "step": 6595
    },
    {
      "epoch": 3.459119496855346,
      "grad_norm": 1.755755066871643,
      "learning_rate": 4.919427341186933e-05,
      "loss": 0.5716,
      "num_input_tokens_seen": 4306616,
      "step": 6600
    },
    {
      "epoch": 3.461740041928721,
      "grad_norm": 1.6793385744094849,
      "learning_rate": 4.919139136122045e-05,
      "loss": 0.4066,
      "num_input_tokens_seen": 4309112,
      "step": 6605
    },
    {
      "epoch": 3.4643605870020964,
      "grad_norm": 1.8056482076644897,
      "learning_rate": 4.918850425000849e-05,
      "loss": 0.4827,
      "num_input_tokens_seen": 4311832,
      "step": 6610
    },
    {
      "epoch": 3.4669811320754715,
      "grad_norm": 0.2527998387813568,
      "learning_rate": 4.918561207883741e-05,
      "loss": 0.3836,
      "num_input_tokens_seen": 4318328,
      "step": 6615
    },
    {
      "epoch": 3.469601677148847,
      "grad_norm": 1.7736222743988037,
      "learning_rate": 4.9182714848312195e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 4321848,
      "step": 6620
    },
    {
      "epoch": 3.4722222222222223,
      "grad_norm": 0.9234317541122437,
      "learning_rate": 4.917981255903893e-05,
      "loss": 0.2934,
      "num_input_tokens_seen": 4324984,
      "step": 6625
    },
    {
      "epoch": 3.4748427672955975,
      "grad_norm": 1.2282058000564575,
      "learning_rate": 4.917690521162474e-05,
      "loss": 0.3577,
      "num_input_tokens_seen": 4327704,
      "step": 6630
    },
    {
      "epoch": 3.4774633123689727,
      "grad_norm": 0.794272243976593,
      "learning_rate": 4.9173992806677796e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 4330904,
      "step": 6635
    },
    {
      "epoch": 3.480083857442348,
      "grad_norm": 1.0858047008514404,
      "learning_rate": 4.917107534480735e-05,
      "loss": 0.4546,
      "num_input_tokens_seen": 4333592,
      "step": 6640
    },
    {
      "epoch": 3.4827044025157234,
      "grad_norm": 1.416876196861267,
      "learning_rate": 4.916815282662369e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 4336696,
      "step": 6645
    },
    {
      "epoch": 3.4853249475890986,
      "grad_norm": 2.611642837524414,
      "learning_rate": 4.91652252527382e-05,
      "loss": 0.6161,
      "num_input_tokens_seen": 4340088,
      "step": 6650
    },
    {
      "epoch": 3.487945492662474,
      "grad_norm": 1.2022171020507812,
      "learning_rate": 4.9162292623763275e-05,
      "loss": 0.4818,
      "num_input_tokens_seen": 4343640,
      "step": 6655
    },
    {
      "epoch": 3.490566037735849,
      "grad_norm": 0.831134557723999,
      "learning_rate": 4.9159354940312386e-05,
      "loss": 0.3988,
      "num_input_tokens_seen": 4348568,
      "step": 6660
    },
    {
      "epoch": 3.4931865828092246,
      "grad_norm": 0.9427507519721985,
      "learning_rate": 4.915641220300007e-05,
      "loss": 0.3676,
      "num_input_tokens_seen": 4352376,
      "step": 6665
    },
    {
      "epoch": 3.4958071278825997,
      "grad_norm": 0.9272551536560059,
      "learning_rate": 4.9153464412441915e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 4356120,
      "step": 6670
    },
    {
      "epoch": 3.498427672955975,
      "grad_norm": 1.0241420269012451,
      "learning_rate": 4.915051156925456e-05,
      "loss": 0.3968,
      "num_input_tokens_seen": 4359448,
      "step": 6675
    },
    {
      "epoch": 3.50104821802935,
      "grad_norm": 0.8697819709777832,
      "learning_rate": 4.91475536740557e-05,
      "loss": 0.2927,
      "num_input_tokens_seen": 4362936,
      "step": 6680
    },
    {
      "epoch": 3.5036687631027252,
      "grad_norm": 1.1291390657424927,
      "learning_rate": 4.914459072746411e-05,
      "loss": 0.3627,
      "num_input_tokens_seen": 4366968,
      "step": 6685
    },
    {
      "epoch": 3.5062893081761004,
      "grad_norm": 0.9144736528396606,
      "learning_rate": 4.91416227300996e-05,
      "loss": 0.3103,
      "num_input_tokens_seen": 4370264,
      "step": 6690
    },
    {
      "epoch": 3.508909853249476,
      "grad_norm": 1.0676838159561157,
      "learning_rate": 4.913864968258303e-05,
      "loss": 0.2771,
      "num_input_tokens_seen": 4373432,
      "step": 6695
    },
    {
      "epoch": 3.511530398322851,
      "grad_norm": 0.8867568969726562,
      "learning_rate": 4.913567158553634e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 4376824,
      "step": 6700
    },
    {
      "epoch": 3.5141509433962264,
      "grad_norm": 0.8970465064048767,
      "learning_rate": 4.9132688439582505e-05,
      "loss": 0.4578,
      "num_input_tokens_seen": 4380504,
      "step": 6705
    },
    {
      "epoch": 3.5167714884696015,
      "grad_norm": 0.9824775457382202,
      "learning_rate": 4.9129700245345565e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 4383704,
      "step": 6710
    },
    {
      "epoch": 3.519392033542977,
      "grad_norm": 1.4917669296264648,
      "learning_rate": 4.912670700345063e-05,
      "loss": 0.5407,
      "num_input_tokens_seen": 4387128,
      "step": 6715
    },
    {
      "epoch": 3.5220125786163523,
      "grad_norm": 0.9443346261978149,
      "learning_rate": 4.912370871452383e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 4390296,
      "step": 6720
    },
    {
      "epoch": 3.5246331236897275,
      "grad_norm": 1.7107806205749512,
      "learning_rate": 4.91207053791924e-05,
      "loss": 0.4621,
      "num_input_tokens_seen": 4393624,
      "step": 6725
    },
    {
      "epoch": 3.5272536687631026,
      "grad_norm": 1.0839745998382568,
      "learning_rate": 4.911769699808457e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 4397944,
      "step": 6730
    },
    {
      "epoch": 3.529874213836478,
      "grad_norm": 1.0365018844604492,
      "learning_rate": 4.911468357182969e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 4400696,
      "step": 6735
    },
    {
      "epoch": 3.532494758909853,
      "grad_norm": 1.597683072090149,
      "learning_rate": 4.9111665101058116e-05,
      "loss": 0.4048,
      "num_input_tokens_seen": 4403416,
      "step": 6740
    },
    {
      "epoch": 3.5351153039832286,
      "grad_norm": 0.6632439494132996,
      "learning_rate": 4.910864158640128e-05,
      "loss": 0.3632,
      "num_input_tokens_seen": 4407320,
      "step": 6745
    },
    {
      "epoch": 3.5377358490566038,
      "grad_norm": 1.2090425491333008,
      "learning_rate": 4.910561302849168e-05,
      "loss": 0.3561,
      "num_input_tokens_seen": 4409848,
      "step": 6750
    },
    {
      "epoch": 3.540356394129979,
      "grad_norm": 0.9071605205535889,
      "learning_rate": 4.910257942796284e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 4412920,
      "step": 6755
    },
    {
      "epoch": 3.5429769392033545,
      "grad_norm": 1.0767334699630737,
      "learning_rate": 4.909954078544936e-05,
      "loss": 0.3168,
      "num_input_tokens_seen": 4415384,
      "step": 6760
    },
    {
      "epoch": 3.5455974842767297,
      "grad_norm": 1.0099319219589233,
      "learning_rate": 4.909649710158689e-05,
      "loss": 0.3186,
      "num_input_tokens_seen": 4418040,
      "step": 6765
    },
    {
      "epoch": 3.548218029350105,
      "grad_norm": 0.8827725648880005,
      "learning_rate": 4.909344837701214e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 4421464,
      "step": 6770
    },
    {
      "epoch": 3.55083857442348,
      "grad_norm": 1.0453975200653076,
      "learning_rate": 4.909039461236285e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 4423672,
      "step": 6775
    },
    {
      "epoch": 3.5534591194968552,
      "grad_norm": 1.094611644744873,
      "learning_rate": 4.908733580827786e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 4426648,
      "step": 6780
    },
    {
      "epoch": 3.5560796645702304,
      "grad_norm": 2.206169605255127,
      "learning_rate": 4.9084271965397014e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 4429752,
      "step": 6785
    },
    {
      "epoch": 3.558700209643606,
      "grad_norm": 1.0361583232879639,
      "learning_rate": 4.908120308436124e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 4433784,
      "step": 6790
    },
    {
      "epoch": 3.561320754716981,
      "grad_norm": 1.5098739862442017,
      "learning_rate": 4.907812916581251e-05,
      "loss": 0.3853,
      "num_input_tokens_seen": 4436376,
      "step": 6795
    },
    {
      "epoch": 3.5639412997903563,
      "grad_norm": 1.2422913312911987,
      "learning_rate": 4.907505021039386e-05,
      "loss": 0.3428,
      "num_input_tokens_seen": 4439384,
      "step": 6800
    },
    {
      "epoch": 3.5665618448637315,
      "grad_norm": 1.500923991203308,
      "learning_rate": 4.907196621874937e-05,
      "loss": 0.4275,
      "num_input_tokens_seen": 4442712,
      "step": 6805
    },
    {
      "epoch": 3.569182389937107,
      "grad_norm": 1.104443907737732,
      "learning_rate": 4.9068877191524176e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 4446936,
      "step": 6810
    },
    {
      "epoch": 3.5718029350104823,
      "grad_norm": 0.869310200214386,
      "learning_rate": 4.906578312936446e-05,
      "loss": 0.4695,
      "num_input_tokens_seen": 4450328,
      "step": 6815
    },
    {
      "epoch": 3.5744234800838575,
      "grad_norm": 0.8154891133308411,
      "learning_rate": 4.906268403291747e-05,
      "loss": 0.341,
      "num_input_tokens_seen": 4453816,
      "step": 6820
    },
    {
      "epoch": 3.5770440251572326,
      "grad_norm": 0.6796230673789978,
      "learning_rate": 4.905957990283151e-05,
      "loss": 0.339,
      "num_input_tokens_seen": 4458072,
      "step": 6825
    },
    {
      "epoch": 3.579664570230608,
      "grad_norm": 0.8579254150390625,
      "learning_rate": 4.905647073975592e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 4461144,
      "step": 6830
    },
    {
      "epoch": 3.582285115303983,
      "grad_norm": 0.8039872050285339,
      "learning_rate": 4.9053356544341095e-05,
      "loss": 0.2973,
      "num_input_tokens_seen": 4463928,
      "step": 6835
    },
    {
      "epoch": 3.5849056603773586,
      "grad_norm": 1.2537533044815063,
      "learning_rate": 4.90502373172385e-05,
      "loss": 0.3969,
      "num_input_tokens_seen": 4466488,
      "step": 6840
    },
    {
      "epoch": 3.5875262054507338,
      "grad_norm": 0.7446708679199219,
      "learning_rate": 4.9047113059100635e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 4470104,
      "step": 6845
    },
    {
      "epoch": 3.590146750524109,
      "grad_norm": 1.1371793746948242,
      "learning_rate": 4.904398377058106e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 4472184,
      "step": 6850
    },
    {
      "epoch": 3.5927672955974845,
      "grad_norm": 1.0135655403137207,
      "learning_rate": 4.90408494523344e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 4474840,
      "step": 6855
    },
    {
      "epoch": 3.5953878406708597,
      "grad_norm": 1.7349276542663574,
      "learning_rate": 4.90377101050163e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 4477816,
      "step": 6860
    },
    {
      "epoch": 3.598008385744235,
      "grad_norm": 0.9788726568222046,
      "learning_rate": 4.903456572928348e-05,
      "loss": 0.4855,
      "num_input_tokens_seen": 4481816,
      "step": 6865
    },
    {
      "epoch": 3.60062893081761,
      "grad_norm": 1.147194743156433,
      "learning_rate": 4.903141632579371e-05,
      "loss": 0.3266,
      "num_input_tokens_seen": 4485464,
      "step": 6870
    },
    {
      "epoch": 3.603249475890985,
      "grad_norm": 0.7556443810462952,
      "learning_rate": 4.902826189520581e-05,
      "loss": 0.3361,
      "num_input_tokens_seen": 4490392,
      "step": 6875
    },
    {
      "epoch": 3.6058700209643604,
      "grad_norm": 1.3191825151443481,
      "learning_rate": 4.9025102438179656e-05,
      "loss": 0.417,
      "num_input_tokens_seen": 4494712,
      "step": 6880
    },
    {
      "epoch": 3.608490566037736,
      "grad_norm": 0.9364436864852905,
      "learning_rate": 4.902193795537616e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 4497528,
      "step": 6885
    },
    {
      "epoch": 3.611111111111111,
      "grad_norm": 0.8439273238182068,
      "learning_rate": 4.90187684474573e-05,
      "loss": 0.3159,
      "num_input_tokens_seen": 4500536,
      "step": 6890
    },
    {
      "epoch": 3.6137316561844863,
      "grad_norm": 1.2053357362747192,
      "learning_rate": 4.901559391508611e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 4504056,
      "step": 6895
    },
    {
      "epoch": 3.6163522012578615,
      "grad_norm": 1.1504745483398438,
      "learning_rate": 4.9012414358926644e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 4506936,
      "step": 6900
    },
    {
      "epoch": 3.618972746331237,
      "grad_norm": 0.7138338088989258,
      "learning_rate": 4.900922977964405e-05,
      "loss": 0.319,
      "num_input_tokens_seen": 4509560,
      "step": 6905
    },
    {
      "epoch": 3.6215932914046123,
      "grad_norm": 1.122328281402588,
      "learning_rate": 4.90060401779045e-05,
      "loss": 0.4013,
      "num_input_tokens_seen": 4513240,
      "step": 6910
    },
    {
      "epoch": 3.6242138364779874,
      "grad_norm": 1.1168872117996216,
      "learning_rate": 4.900284555437521e-05,
      "loss": 0.3945,
      "num_input_tokens_seen": 4515640,
      "step": 6915
    },
    {
      "epoch": 3.6268343815513626,
      "grad_norm": 0.830141544342041,
      "learning_rate": 4.899964590972448e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 4520824,
      "step": 6920
    },
    {
      "epoch": 3.629454926624738,
      "grad_norm": 1.1952943801879883,
      "learning_rate": 4.899644124462163e-05,
      "loss": 0.4331,
      "num_input_tokens_seen": 4523896,
      "step": 6925
    },
    {
      "epoch": 3.632075471698113,
      "grad_norm": 1.4916858673095703,
      "learning_rate": 4.8993231559737026e-05,
      "loss": 0.5143,
      "num_input_tokens_seen": 4527000,
      "step": 6930
    },
    {
      "epoch": 3.6346960167714886,
      "grad_norm": 1.5683764219284058,
      "learning_rate": 4.8990016855742114e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 4529912,
      "step": 6935
    },
    {
      "epoch": 3.6373165618448637,
      "grad_norm": 1.3899452686309814,
      "learning_rate": 4.898679713330937e-05,
      "loss": 0.356,
      "num_input_tokens_seen": 4533336,
      "step": 6940
    },
    {
      "epoch": 3.639937106918239,
      "grad_norm": 0.830420970916748,
      "learning_rate": 4.898357239311232e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 4536344,
      "step": 6945
    },
    {
      "epoch": 3.6425576519916145,
      "grad_norm": 0.8101591467857361,
      "learning_rate": 4.898034263582554e-05,
      "loss": 0.3353,
      "num_input_tokens_seen": 4540152,
      "step": 6950
    },
    {
      "epoch": 3.6451781970649897,
      "grad_norm": 1.2478569746017456,
      "learning_rate": 4.897710786212467e-05,
      "loss": 0.3132,
      "num_input_tokens_seen": 4543032,
      "step": 6955
    },
    {
      "epoch": 3.647798742138365,
      "grad_norm": 0.8612096905708313,
      "learning_rate": 4.897386807268638e-05,
      "loss": 0.3857,
      "num_input_tokens_seen": 4547064,
      "step": 6960
    },
    {
      "epoch": 3.65041928721174,
      "grad_norm": 0.9237677454948425,
      "learning_rate": 4.897062326818839e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 4551000,
      "step": 6965
    },
    {
      "epoch": 3.653039832285115,
      "grad_norm": 0.743744432926178,
      "learning_rate": 4.8967373449309495e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 4555000,
      "step": 6970
    },
    {
      "epoch": 3.6556603773584904,
      "grad_norm": 0.6542105674743652,
      "learning_rate": 4.89641186167295e-05,
      "loss": 0.3556,
      "num_input_tokens_seen": 4558680,
      "step": 6975
    },
    {
      "epoch": 3.658280922431866,
      "grad_norm": 0.8431555032730103,
      "learning_rate": 4.8960858771129294e-05,
      "loss": 0.3831,
      "num_input_tokens_seen": 4561336,
      "step": 6980
    },
    {
      "epoch": 3.660901467505241,
      "grad_norm": 1.1184678077697754,
      "learning_rate": 4.895759391319079e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 4564568,
      "step": 6985
    },
    {
      "epoch": 3.6635220125786163,
      "grad_norm": 1.042493462562561,
      "learning_rate": 4.8954324043596954e-05,
      "loss": 0.383,
      "num_input_tokens_seen": 4567448,
      "step": 6990
    },
    {
      "epoch": 3.6661425576519915,
      "grad_norm": 0.7103968858718872,
      "learning_rate": 4.8951049163031826e-05,
      "loss": 0.4665,
      "num_input_tokens_seen": 4570136,
      "step": 6995
    },
    {
      "epoch": 3.668763102725367,
      "grad_norm": 0.972148060798645,
      "learning_rate": 4.894776927218045e-05,
      "loss": 0.4093,
      "num_input_tokens_seen": 4573624,
      "step": 7000
    },
    {
      "epoch": 3.6713836477987423,
      "grad_norm": 2.53896427154541,
      "learning_rate": 4.894448437172896e-05,
      "loss": 0.3498,
      "num_input_tokens_seen": 4577048,
      "step": 7005
    },
    {
      "epoch": 3.6740041928721174,
      "grad_norm": 1.0121194124221802,
      "learning_rate": 4.894119446236452e-05,
      "loss": 0.432,
      "num_input_tokens_seen": 4580088,
      "step": 7010
    },
    {
      "epoch": 3.6766247379454926,
      "grad_norm": 0.9318395853042603,
      "learning_rate": 4.893789954477532e-05,
      "loss": 0.3451,
      "num_input_tokens_seen": 4583480,
      "step": 7015
    },
    {
      "epoch": 3.6792452830188678,
      "grad_norm": 1.2297399044036865,
      "learning_rate": 4.893459961965063e-05,
      "loss": 0.5333,
      "num_input_tokens_seen": 4586168,
      "step": 7020
    },
    {
      "epoch": 3.681865828092243,
      "grad_norm": 1.254990577697754,
      "learning_rate": 4.8931294687680775e-05,
      "loss": 0.3462,
      "num_input_tokens_seen": 4589592,
      "step": 7025
    },
    {
      "epoch": 3.6844863731656186,
      "grad_norm": 1.1304959058761597,
      "learning_rate": 4.8927984749557086e-05,
      "loss": 0.4498,
      "num_input_tokens_seen": 4592440,
      "step": 7030
    },
    {
      "epoch": 3.6871069182389937,
      "grad_norm": 0.8838256001472473,
      "learning_rate": 4.892466980597198e-05,
      "loss": 0.3004,
      "num_input_tokens_seen": 4594904,
      "step": 7035
    },
    {
      "epoch": 3.689727463312369,
      "grad_norm": 0.9947354197502136,
      "learning_rate": 4.892134985761889e-05,
      "loss": 0.4361,
      "num_input_tokens_seen": 4597624,
      "step": 7040
    },
    {
      "epoch": 3.6923480083857445,
      "grad_norm": 1.6695979833602905,
      "learning_rate": 4.891802490519232e-05,
      "loss": 0.4839,
      "num_input_tokens_seen": 4599928,
      "step": 7045
    },
    {
      "epoch": 3.6949685534591197,
      "grad_norm": 2.051722764968872,
      "learning_rate": 4.891469494938781e-05,
      "loss": 0.4114,
      "num_input_tokens_seen": 4602328,
      "step": 7050
    },
    {
      "epoch": 3.697589098532495,
      "grad_norm": 0.7126214504241943,
      "learning_rate": 4.891135999090195e-05,
      "loss": 0.4193,
      "num_input_tokens_seen": 4606328,
      "step": 7055
    },
    {
      "epoch": 3.70020964360587,
      "grad_norm": 0.8453258275985718,
      "learning_rate": 4.890802003043238e-05,
      "loss": 0.3344,
      "num_input_tokens_seen": 4609176,
      "step": 7060
    },
    {
      "epoch": 3.702830188679245,
      "grad_norm": 0.9442159533500671,
      "learning_rate": 4.890467506867777e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 4612056,
      "step": 7065
    },
    {
      "epoch": 3.7054507337526204,
      "grad_norm": 1.2338637113571167,
      "learning_rate": 4.890132510633786e-05,
      "loss": 0.5097,
      "num_input_tokens_seen": 4615320,
      "step": 7070
    },
    {
      "epoch": 3.708071278825996,
      "grad_norm": 1.1111750602722168,
      "learning_rate": 4.8897970144113416e-05,
      "loss": 0.4571,
      "num_input_tokens_seen": 4618200,
      "step": 7075
    },
    {
      "epoch": 3.710691823899371,
      "grad_norm": 1.3653851747512817,
      "learning_rate": 4.889461018270626e-05,
      "loss": 0.2902,
      "num_input_tokens_seen": 4621368,
      "step": 7080
    },
    {
      "epoch": 3.7133123689727463,
      "grad_norm": 1.2205305099487305,
      "learning_rate": 4.889124522281926e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 4624088,
      "step": 7085
    },
    {
      "epoch": 3.7159329140461215,
      "grad_norm": 0.9996815919876099,
      "learning_rate": 4.888787526515632e-05,
      "loss": 0.5462,
      "num_input_tokens_seen": 4628568,
      "step": 7090
    },
    {
      "epoch": 3.718553459119497,
      "grad_norm": 0.9820882081985474,
      "learning_rate": 4.8884500310422406e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 4631288,
      "step": 7095
    },
    {
      "epoch": 3.7211740041928723,
      "grad_norm": 0.8371535539627075,
      "learning_rate": 4.8881120359323516e-05,
      "loss": 0.3402,
      "num_input_tokens_seen": 4635224,
      "step": 7100
    },
    {
      "epoch": 3.7237945492662474,
      "grad_norm": 1.0049054622650146,
      "learning_rate": 4.887773541256669e-05,
      "loss": 0.5845,
      "num_input_tokens_seen": 4638648,
      "step": 7105
    },
    {
      "epoch": 3.7264150943396226,
      "grad_norm": 1.6516706943511963,
      "learning_rate": 4.887434547086004e-05,
      "loss": 0.3181,
      "num_input_tokens_seen": 4641848,
      "step": 7110
    },
    {
      "epoch": 3.7290356394129978,
      "grad_norm": 1.0425766706466675,
      "learning_rate": 4.887095053491269e-05,
      "loss": 0.4089,
      "num_input_tokens_seen": 4644408,
      "step": 7115
    },
    {
      "epoch": 3.731656184486373,
      "grad_norm": 0.7379213571548462,
      "learning_rate": 4.8867550605434806e-05,
      "loss": 0.5302,
      "num_input_tokens_seen": 4647096,
      "step": 7120
    },
    {
      "epoch": 3.7342767295597485,
      "grad_norm": 1.0534827709197998,
      "learning_rate": 4.8864145683137644e-05,
      "loss": 0.4189,
      "num_input_tokens_seen": 4649176,
      "step": 7125
    },
    {
      "epoch": 3.7368972746331237,
      "grad_norm": 1.4477002620697021,
      "learning_rate": 4.886073576873346e-05,
      "loss": 0.4384,
      "num_input_tokens_seen": 4653144,
      "step": 7130
    },
    {
      "epoch": 3.739517819706499,
      "grad_norm": 1.2855682373046875,
      "learning_rate": 4.885732086293556e-05,
      "loss": 0.2894,
      "num_input_tokens_seen": 4655864,
      "step": 7135
    },
    {
      "epoch": 3.742138364779874,
      "grad_norm": 0.7629873752593994,
      "learning_rate": 4.8853900966458335e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 4659000,
      "step": 7140
    },
    {
      "epoch": 3.7447589098532497,
      "grad_norm": 0.8276227712631226,
      "learning_rate": 4.8850476080017146e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 4661720,
      "step": 7145
    },
    {
      "epoch": 3.747379454926625,
      "grad_norm": 1.200677514076233,
      "learning_rate": 4.884704620432848e-05,
      "loss": 0.4025,
      "num_input_tokens_seen": 4666584,
      "step": 7150
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.1558220386505127,
      "learning_rate": 4.8843611340109794e-05,
      "loss": 0.3756,
      "num_input_tokens_seen": 4669272,
      "step": 7155
    },
    {
      "epoch": 3.752620545073375,
      "grad_norm": 1.0291029214859009,
      "learning_rate": 4.884017148807965e-05,
      "loss": 0.414,
      "num_input_tokens_seen": 4672280,
      "step": 7160
    },
    {
      "epoch": 3.7552410901467503,
      "grad_norm": 1.3015202283859253,
      "learning_rate": 4.883672664895761e-05,
      "loss": 0.3233,
      "num_input_tokens_seen": 4675032,
      "step": 7165
    },
    {
      "epoch": 3.757861635220126,
      "grad_norm": 1.2713797092437744,
      "learning_rate": 4.8833276823464296e-05,
      "loss": 0.3855,
      "num_input_tokens_seen": 4678360,
      "step": 7170
    },
    {
      "epoch": 3.760482180293501,
      "grad_norm": 2.4212024211883545,
      "learning_rate": 4.882982201232138e-05,
      "loss": 0.4053,
      "num_input_tokens_seen": 4681208,
      "step": 7175
    },
    {
      "epoch": 3.7631027253668763,
      "grad_norm": 1.0874637365341187,
      "learning_rate": 4.882636221625156e-05,
      "loss": 0.4204,
      "num_input_tokens_seen": 4683768,
      "step": 7180
    },
    {
      "epoch": 3.7657232704402515,
      "grad_norm": 1.4910860061645508,
      "learning_rate": 4.8822897435978586e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 4686872,
      "step": 7185
    },
    {
      "epoch": 3.768343815513627,
      "grad_norm": 0.9689028263092041,
      "learning_rate": 4.8819427672227264e-05,
      "loss": 0.4003,
      "num_input_tokens_seen": 4690424,
      "step": 7190
    },
    {
      "epoch": 3.7709643605870022,
      "grad_norm": 1.9200360774993896,
      "learning_rate": 4.881595292572341e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 4692824,
      "step": 7195
    },
    {
      "epoch": 3.7735849056603774,
      "grad_norm": 0.9907832145690918,
      "learning_rate": 4.881247319719392e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 4696440,
      "step": 7200
    },
    {
      "epoch": 3.7762054507337526,
      "grad_norm": 0.9860627055168152,
      "learning_rate": 4.88089884873667e-05,
      "loss": 0.3708,
      "num_input_tokens_seen": 4700440,
      "step": 7205
    },
    {
      "epoch": 3.7788259958071277,
      "grad_norm": 0.9143077731132507,
      "learning_rate": 4.880549879697071e-05,
      "loss": 0.3679,
      "num_input_tokens_seen": 4703320,
      "step": 7210
    },
    {
      "epoch": 3.781446540880503,
      "grad_norm": 1.1299515962600708,
      "learning_rate": 4.8802004126735965e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 4706552,
      "step": 7215
    },
    {
      "epoch": 3.7840670859538785,
      "grad_norm": 0.5914731025695801,
      "learning_rate": 4.87985044773935e-05,
      "loss": 0.3384,
      "num_input_tokens_seen": 4709592,
      "step": 7220
    },
    {
      "epoch": 3.7866876310272537,
      "grad_norm": 1.2709351778030396,
      "learning_rate": 4.879499984967541e-05,
      "loss": 0.4831,
      "num_input_tokens_seen": 4712344,
      "step": 7225
    },
    {
      "epoch": 3.789308176100629,
      "grad_norm": 1.2898249626159668,
      "learning_rate": 4.8791490244314824e-05,
      "loss": 0.4243,
      "num_input_tokens_seen": 4715000,
      "step": 7230
    },
    {
      "epoch": 3.791928721174004,
      "grad_norm": 0.8999462127685547,
      "learning_rate": 4.8787975662045904e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 4718904,
      "step": 7235
    },
    {
      "epoch": 3.7945492662473796,
      "grad_norm": 1.337198257446289,
      "learning_rate": 4.878445610360386e-05,
      "loss": 0.4805,
      "num_input_tokens_seen": 4721656,
      "step": 7240
    },
    {
      "epoch": 3.797169811320755,
      "grad_norm": 2.173952341079712,
      "learning_rate": 4.878093156972494e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 4726264,
      "step": 7245
    },
    {
      "epoch": 3.79979035639413,
      "grad_norm": 1.0181993246078491,
      "learning_rate": 4.877740206114645e-05,
      "loss": 0.3399,
      "num_input_tokens_seen": 4729112,
      "step": 7250
    },
    {
      "epoch": 3.802410901467505,
      "grad_norm": 1.0462204217910767,
      "learning_rate": 4.8773867578606716e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 4731832,
      "step": 7255
    },
    {
      "epoch": 3.8050314465408803,
      "grad_norm": 1.4482824802398682,
      "learning_rate": 4.877032812284511e-05,
      "loss": 0.4882,
      "num_input_tokens_seen": 4738136,
      "step": 7260
    },
    {
      "epoch": 3.8076519916142555,
      "grad_norm": 0.852252185344696,
      "learning_rate": 4.876678369460205e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 4741080,
      "step": 7265
    },
    {
      "epoch": 3.810272536687631,
      "grad_norm": 0.9033676385879517,
      "learning_rate": 4.8763234294618985e-05,
      "loss": 0.3004,
      "num_input_tokens_seen": 4744472,
      "step": 7270
    },
    {
      "epoch": 3.8128930817610063,
      "grad_norm": 1.0593374967575073,
      "learning_rate": 4.875967992363841e-05,
      "loss": 0.3479,
      "num_input_tokens_seen": 4748664,
      "step": 7275
    },
    {
      "epoch": 3.8155136268343814,
      "grad_norm": 2.069556474685669,
      "learning_rate": 4.875612058240386e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 4751704,
      "step": 7280
    },
    {
      "epoch": 3.818134171907757,
      "grad_norm": 2.6545393466949463,
      "learning_rate": 4.875255627165991e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 4754104,
      "step": 7285
    },
    {
      "epoch": 3.8207547169811322,
      "grad_norm": 1.042280912399292,
      "learning_rate": 4.874898699215217e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 4757528,
      "step": 7290
    },
    {
      "epoch": 3.8233752620545074,
      "grad_norm": 0.707555890083313,
      "learning_rate": 4.8745412744627296e-05,
      "loss": 0.5335,
      "num_input_tokens_seen": 4761208,
      "step": 7295
    },
    {
      "epoch": 3.8259958071278826,
      "grad_norm": 1.104135513305664,
      "learning_rate": 4.874183352983297e-05,
      "loss": 0.3948,
      "num_input_tokens_seen": 4764088,
      "step": 7300
    },
    {
      "epoch": 3.8286163522012577,
      "grad_norm": 0.8122856020927429,
      "learning_rate": 4.873824934851794e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 4767352,
      "step": 7305
    },
    {
      "epoch": 3.831236897274633,
      "grad_norm": 1.0965662002563477,
      "learning_rate": 4.873466020143196e-05,
      "loss": 0.4121,
      "num_input_tokens_seen": 4770648,
      "step": 7310
    },
    {
      "epoch": 3.8338574423480085,
      "grad_norm": 0.6972967386245728,
      "learning_rate": 4.873106608932585e-05,
      "loss": 0.3929,
      "num_input_tokens_seen": 4777624,
      "step": 7315
    },
    {
      "epoch": 3.8364779874213837,
      "grad_norm": 1.588997483253479,
      "learning_rate": 4.872746701295145e-05,
      "loss": 0.491,
      "num_input_tokens_seen": 4780376,
      "step": 7320
    },
    {
      "epoch": 3.839098532494759,
      "grad_norm": 1.0644282102584839,
      "learning_rate": 4.872386297306165e-05,
      "loss": 0.3859,
      "num_input_tokens_seen": 4783096,
      "step": 7325
    },
    {
      "epoch": 3.841719077568134,
      "grad_norm": 0.837413489818573,
      "learning_rate": 4.872025397041037e-05,
      "loss": 0.2904,
      "num_input_tokens_seen": 4787000,
      "step": 7330
    },
    {
      "epoch": 3.8443396226415096,
      "grad_norm": 1.8508553504943848,
      "learning_rate": 4.871664000575258e-05,
      "loss": 0.4253,
      "num_input_tokens_seen": 4790168,
      "step": 7335
    },
    {
      "epoch": 3.846960167714885,
      "grad_norm": 0.9845690131187439,
      "learning_rate": 4.871302107984427e-05,
      "loss": 0.464,
      "num_input_tokens_seen": 4793944,
      "step": 7340
    },
    {
      "epoch": 3.84958071278826,
      "grad_norm": 1.1035288572311401,
      "learning_rate": 4.8709397193442474e-05,
      "loss": 0.3362,
      "num_input_tokens_seen": 4797016,
      "step": 7345
    },
    {
      "epoch": 3.852201257861635,
      "grad_norm": 0.8041729927062988,
      "learning_rate": 4.870576834730529e-05,
      "loss": 0.4235,
      "num_input_tokens_seen": 4800216,
      "step": 7350
    },
    {
      "epoch": 3.8548218029350103,
      "grad_norm": 1.0343341827392578,
      "learning_rate": 4.8702134542191816e-05,
      "loss": 0.326,
      "num_input_tokens_seen": 4803992,
      "step": 7355
    },
    {
      "epoch": 3.8574423480083855,
      "grad_norm": 1.2063922882080078,
      "learning_rate": 4.86984957788622e-05,
      "loss": 0.3882,
      "num_input_tokens_seen": 4806872,
      "step": 7360
    },
    {
      "epoch": 3.860062893081761,
      "grad_norm": 0.7907710075378418,
      "learning_rate": 4.869485205807763e-05,
      "loss": 0.263,
      "num_input_tokens_seen": 4809912,
      "step": 7365
    },
    {
      "epoch": 3.8626834381551363,
      "grad_norm": 0.717073380947113,
      "learning_rate": 4.869120338060034e-05,
      "loss": 0.2857,
      "num_input_tokens_seen": 4816984,
      "step": 7370
    },
    {
      "epoch": 3.8653039832285114,
      "grad_norm": 0.9717546105384827,
      "learning_rate": 4.8687549747193585e-05,
      "loss": 0.4676,
      "num_input_tokens_seen": 4819608,
      "step": 7375
    },
    {
      "epoch": 3.867924528301887,
      "grad_norm": 0.5792561173439026,
      "learning_rate": 4.868389115862166e-05,
      "loss": 0.3482,
      "num_input_tokens_seen": 4823448,
      "step": 7380
    },
    {
      "epoch": 3.870545073375262,
      "grad_norm": 0.9384453296661377,
      "learning_rate": 4.868022761564991e-05,
      "loss": 0.4186,
      "num_input_tokens_seen": 4826328,
      "step": 7385
    },
    {
      "epoch": 3.8731656184486374,
      "grad_norm": 1.606200098991394,
      "learning_rate": 4.8676559119044696e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 4829048,
      "step": 7390
    },
    {
      "epoch": 3.8757861635220126,
      "grad_norm": 1.0419594049453735,
      "learning_rate": 4.867288566957344e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 4833048,
      "step": 7395
    },
    {
      "epoch": 3.8784067085953877,
      "grad_norm": 0.8029802441596985,
      "learning_rate": 4.866920726800457e-05,
      "loss": 0.354,
      "num_input_tokens_seen": 4836184,
      "step": 7400
    },
    {
      "epoch": 3.881027253668763,
      "grad_norm": 1.841908574104309,
      "learning_rate": 4.8665523915107576e-05,
      "loss": 0.4082,
      "num_input_tokens_seen": 4839896,
      "step": 7405
    },
    {
      "epoch": 3.8836477987421385,
      "grad_norm": 1.4229681491851807,
      "learning_rate": 4.866183561165296e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 4842936,
      "step": 7410
    },
    {
      "epoch": 3.8862683438155137,
      "grad_norm": 1.5437281131744385,
      "learning_rate": 4.86581423584123e-05,
      "loss": 0.5048,
      "num_input_tokens_seen": 4846200,
      "step": 7415
    },
    {
      "epoch": 3.888888888888889,
      "grad_norm": 1.0466258525848389,
      "learning_rate": 4.8654444156158154e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 4849336,
      "step": 7420
    },
    {
      "epoch": 3.891509433962264,
      "grad_norm": 1.2517642974853516,
      "learning_rate": 4.865074100566415e-05,
      "loss": 0.4052,
      "num_input_tokens_seen": 4852632,
      "step": 7425
    },
    {
      "epoch": 3.8941299790356396,
      "grad_norm": 1.1274046897888184,
      "learning_rate": 4.864703290770495e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 4855384,
      "step": 7430
    },
    {
      "epoch": 3.896750524109015,
      "grad_norm": 0.822266697883606,
      "learning_rate": 4.864331986305626e-05,
      "loss": 0.442,
      "num_input_tokens_seen": 4858936,
      "step": 7435
    },
    {
      "epoch": 3.89937106918239,
      "grad_norm": 1.283467411994934,
      "learning_rate": 4.8639601872494775e-05,
      "loss": 0.4409,
      "num_input_tokens_seen": 4862488,
      "step": 7440
    },
    {
      "epoch": 3.901991614255765,
      "grad_norm": 0.8804435729980469,
      "learning_rate": 4.8635878936798287e-05,
      "loss": 0.3731,
      "num_input_tokens_seen": 4865240,
      "step": 7445
    },
    {
      "epoch": 3.9046121593291403,
      "grad_norm": 0.9166920185089111,
      "learning_rate": 4.863215105674557e-05,
      "loss": 0.3726,
      "num_input_tokens_seen": 4868152,
      "step": 7450
    },
    {
      "epoch": 3.9072327044025155,
      "grad_norm": 0.961549699306488,
      "learning_rate": 4.8628418233116457e-05,
      "loss": 0.2984,
      "num_input_tokens_seen": 4870808,
      "step": 7455
    },
    {
      "epoch": 3.909853249475891,
      "grad_norm": 0.7203589677810669,
      "learning_rate": 4.8624680466691824e-05,
      "loss": 0.377,
      "num_input_tokens_seen": 4873944,
      "step": 7460
    },
    {
      "epoch": 3.9124737945492662,
      "grad_norm": 1.168992280960083,
      "learning_rate": 4.862093775825356e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 4877528,
      "step": 7465
    },
    {
      "epoch": 3.9150943396226414,
      "grad_norm": 1.368825912475586,
      "learning_rate": 4.86171901085846e-05,
      "loss": 0.3559,
      "num_input_tokens_seen": 4880504,
      "step": 7470
    },
    {
      "epoch": 3.917714884696017,
      "grad_norm": 1.1566725969314575,
      "learning_rate": 4.861343751846891e-05,
      "loss": 0.3706,
      "num_input_tokens_seen": 4884312,
      "step": 7475
    },
    {
      "epoch": 3.920335429769392,
      "grad_norm": 1.395579218864441,
      "learning_rate": 4.860967998869148e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 4887032,
      "step": 7480
    },
    {
      "epoch": 3.9229559748427674,
      "grad_norm": 0.9612168073654175,
      "learning_rate": 4.860591752003836e-05,
      "loss": 0.462,
      "num_input_tokens_seen": 4889976,
      "step": 7485
    },
    {
      "epoch": 3.9255765199161425,
      "grad_norm": 0.7642187476158142,
      "learning_rate": 4.8602150113296596e-05,
      "loss": 0.5028,
      "num_input_tokens_seen": 4892600,
      "step": 7490
    },
    {
      "epoch": 3.9281970649895177,
      "grad_norm": 1.1520556211471558,
      "learning_rate": 4.8598377769254305e-05,
      "loss": 0.3919,
      "num_input_tokens_seen": 4895640,
      "step": 7495
    },
    {
      "epoch": 3.930817610062893,
      "grad_norm": 0.9350207448005676,
      "learning_rate": 4.85946004887006e-05,
      "loss": 0.3326,
      "num_input_tokens_seen": 4898200,
      "step": 7500
    },
    {
      "epoch": 3.9334381551362685,
      "grad_norm": 1.4286421537399292,
      "learning_rate": 4.859081827242566e-05,
      "loss": 0.3464,
      "num_input_tokens_seen": 4901080,
      "step": 7505
    },
    {
      "epoch": 3.9360587002096437,
      "grad_norm": 1.0492887496948242,
      "learning_rate": 4.858703112122068e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 4904280,
      "step": 7510
    },
    {
      "epoch": 3.938679245283019,
      "grad_norm": 0.9320124983787537,
      "learning_rate": 4.8583239035877876e-05,
      "loss": 0.5917,
      "num_input_tokens_seen": 4907832,
      "step": 7515
    },
    {
      "epoch": 3.941299790356394,
      "grad_norm": 1.3985347747802734,
      "learning_rate": 4.8579442017190526e-05,
      "loss": 0.5315,
      "num_input_tokens_seen": 4911320,
      "step": 7520
    },
    {
      "epoch": 3.9439203354297696,
      "grad_norm": 0.6675195097923279,
      "learning_rate": 4.8575640065952905e-05,
      "loss": 0.3883,
      "num_input_tokens_seen": 4914072,
      "step": 7525
    },
    {
      "epoch": 3.9465408805031448,
      "grad_norm": 1.1345549821853638,
      "learning_rate": 4.857183318296035e-05,
      "loss": 0.4361,
      "num_input_tokens_seen": 4918584,
      "step": 7530
    },
    {
      "epoch": 3.94916142557652,
      "grad_norm": 1.3349409103393555,
      "learning_rate": 4.856802136900922e-05,
      "loss": 0.3783,
      "num_input_tokens_seen": 4921496,
      "step": 7535
    },
    {
      "epoch": 3.951781970649895,
      "grad_norm": 0.8914067149162292,
      "learning_rate": 4.856420462489689e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 4924600,
      "step": 7540
    },
    {
      "epoch": 3.9544025157232703,
      "grad_norm": 0.6207623481750488,
      "learning_rate": 4.856038295142179e-05,
      "loss": 0.4034,
      "num_input_tokens_seen": 4927512,
      "step": 7545
    },
    {
      "epoch": 3.9570230607966455,
      "grad_norm": 1.068329095840454,
      "learning_rate": 4.855655634938337e-05,
      "loss": 0.396,
      "num_input_tokens_seen": 4930360,
      "step": 7550
    },
    {
      "epoch": 3.959643605870021,
      "grad_norm": 0.7535086274147034,
      "learning_rate": 4.8552724819582106e-05,
      "loss": 0.3059,
      "num_input_tokens_seen": 4933496,
      "step": 7555
    },
    {
      "epoch": 3.9622641509433962,
      "grad_norm": 1.1700550317764282,
      "learning_rate": 4.8548888362819515e-05,
      "loss": 0.332,
      "num_input_tokens_seen": 4936152,
      "step": 7560
    },
    {
      "epoch": 3.9648846960167714,
      "grad_norm": 1.3468430042266846,
      "learning_rate": 4.854504697989814e-05,
      "loss": 0.591,
      "num_input_tokens_seen": 4939832,
      "step": 7565
    },
    {
      "epoch": 3.967505241090147,
      "grad_norm": 0.8822637796401978,
      "learning_rate": 4.8541200671621554e-05,
      "loss": 0.4302,
      "num_input_tokens_seen": 4943192,
      "step": 7570
    },
    {
      "epoch": 3.970125786163522,
      "grad_norm": 1.564089298248291,
      "learning_rate": 4.8537349438794354e-05,
      "loss": 0.3902,
      "num_input_tokens_seen": 4946392,
      "step": 7575
    },
    {
      "epoch": 3.9727463312368974,
      "grad_norm": 0.9020586013793945,
      "learning_rate": 4.853349328222219e-05,
      "loss": 0.3188,
      "num_input_tokens_seen": 4949464,
      "step": 7580
    },
    {
      "epoch": 3.9753668763102725,
      "grad_norm": 0.9050703048706055,
      "learning_rate": 4.8529632202711706e-05,
      "loss": 0.3333,
      "num_input_tokens_seen": 4955288,
      "step": 7585
    },
    {
      "epoch": 3.9779874213836477,
      "grad_norm": 0.9194561243057251,
      "learning_rate": 4.852576620107061e-05,
      "loss": 0.3249,
      "num_input_tokens_seen": 4958008,
      "step": 7590
    },
    {
      "epoch": 3.980607966457023,
      "grad_norm": 0.9457065463066101,
      "learning_rate": 4.8521895278107615e-05,
      "loss": 0.3146,
      "num_input_tokens_seen": 4961400,
      "step": 7595
    },
    {
      "epoch": 3.9832285115303985,
      "grad_norm": 1.2911486625671387,
      "learning_rate": 4.8518019434632484e-05,
      "loss": 0.4976,
      "num_input_tokens_seen": 4964536,
      "step": 7600
    },
    {
      "epoch": 3.9858490566037736,
      "grad_norm": 1.0688034296035767,
      "learning_rate": 4.8514138671455995e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 4968280,
      "step": 7605
    },
    {
      "epoch": 3.988469601677149,
      "grad_norm": 0.7904596328735352,
      "learning_rate": 4.851025298938996e-05,
      "loss": 0.4843,
      "num_input_tokens_seen": 4971640,
      "step": 7610
    },
    {
      "epoch": 3.991090146750524,
      "grad_norm": 0.9205315709114075,
      "learning_rate": 4.850636238924722e-05,
      "loss": 0.4221,
      "num_input_tokens_seen": 4975512,
      "step": 7615
    },
    {
      "epoch": 3.9937106918238996,
      "grad_norm": 1.005888819694519,
      "learning_rate": 4.850246687184163e-05,
      "loss": 0.4703,
      "num_input_tokens_seen": 4978520,
      "step": 7620
    },
    {
      "epoch": 3.9963312368972748,
      "grad_norm": 1.1307361125946045,
      "learning_rate": 4.849856643798812e-05,
      "loss": 0.3305,
      "num_input_tokens_seen": 4981944,
      "step": 7625
    },
    {
      "epoch": 3.99895178197065,
      "grad_norm": 1.269311547279358,
      "learning_rate": 4.849466108850259e-05,
      "loss": 0.3453,
      "num_input_tokens_seen": 4984632,
      "step": 7630
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.47423693537712097,
      "eval_runtime": 11.8983,
      "eval_samples_per_second": 71.27,
      "eval_steps_per_second": 17.818,
      "num_input_tokens_seen": 4985200,
      "step": 7632
    },
    {
      "epoch": 4.001572327044025,
      "grad_norm": 0.7580965161323547,
      "learning_rate": 4.8490750824202e-05,
      "loss": 0.3711,
      "num_input_tokens_seen": 4987024,
      "step": 7635
    },
    {
      "epoch": 4.0041928721174,
      "grad_norm": 0.8015506863594055,
      "learning_rate": 4.8486835645904324e-05,
      "loss": 0.3108,
      "num_input_tokens_seen": 4989552,
      "step": 7640
    },
    {
      "epoch": 4.006813417190775,
      "grad_norm": 1.1684069633483887,
      "learning_rate": 4.84829155544286e-05,
      "loss": 0.5508,
      "num_input_tokens_seen": 4992080,
      "step": 7645
    },
    {
      "epoch": 4.009433962264151,
      "grad_norm": 1.0745774507522583,
      "learning_rate": 4.847899055059484e-05,
      "loss": 0.512,
      "num_input_tokens_seen": 4994512,
      "step": 7650
    },
    {
      "epoch": 4.012054507337526,
      "grad_norm": 0.8534103631973267,
      "learning_rate": 4.847506063522412e-05,
      "loss": 0.2413,
      "num_input_tokens_seen": 4998000,
      "step": 7655
    },
    {
      "epoch": 4.014675052410902,
      "grad_norm": 1.0996434688568115,
      "learning_rate": 4.847112580913854e-05,
      "loss": 0.3379,
      "num_input_tokens_seen": 5002000,
      "step": 7660
    },
    {
      "epoch": 4.017295597484277,
      "grad_norm": 0.8011983036994934,
      "learning_rate": 4.846718607316121e-05,
      "loss": 0.3241,
      "num_input_tokens_seen": 5005648,
      "step": 7665
    },
    {
      "epoch": 4.019916142557652,
      "grad_norm": 1.497835397720337,
      "learning_rate": 4.846324142811628e-05,
      "loss": 0.4635,
      "num_input_tokens_seen": 5009104,
      "step": 7670
    },
    {
      "epoch": 4.022536687631027,
      "grad_norm": 0.6097611784934998,
      "learning_rate": 4.8459291874828924e-05,
      "loss": 0.2832,
      "num_input_tokens_seen": 5012304,
      "step": 7675
    },
    {
      "epoch": 4.0251572327044025,
      "grad_norm": 3.829503059387207,
      "learning_rate": 4.845533741412534e-05,
      "loss": 0.4143,
      "num_input_tokens_seen": 5014608,
      "step": 7680
    },
    {
      "epoch": 4.027777777777778,
      "grad_norm": 1.5499905347824097,
      "learning_rate": 4.845137804683277e-05,
      "loss": 0.3892,
      "num_input_tokens_seen": 5017776,
      "step": 7685
    },
    {
      "epoch": 4.030398322851153,
      "grad_norm": 1.1924973726272583,
      "learning_rate": 4.844741377377946e-05,
      "loss": 0.4519,
      "num_input_tokens_seen": 5020432,
      "step": 7690
    },
    {
      "epoch": 4.033018867924528,
      "grad_norm": 0.9926273226737976,
      "learning_rate": 4.844344459579469e-05,
      "loss": 0.242,
      "num_input_tokens_seen": 5023824,
      "step": 7695
    },
    {
      "epoch": 4.035639412997903,
      "grad_norm": 1.010385274887085,
      "learning_rate": 4.8439470513708766e-05,
      "loss": 0.3673,
      "num_input_tokens_seen": 5027344,
      "step": 7700
    },
    {
      "epoch": 4.038259958071279,
      "grad_norm": 1.6789219379425049,
      "learning_rate": 4.8435491528353026e-05,
      "loss": 0.3733,
      "num_input_tokens_seen": 5030928,
      "step": 7705
    },
    {
      "epoch": 4.040880503144654,
      "grad_norm": 0.789652943611145,
      "learning_rate": 4.843150764055982e-05,
      "loss": 0.3052,
      "num_input_tokens_seen": 5034320,
      "step": 7710
    },
    {
      "epoch": 4.04350104821803,
      "grad_norm": 0.9053434133529663,
      "learning_rate": 4.842751885116254e-05,
      "loss": 0.3427,
      "num_input_tokens_seen": 5037456,
      "step": 7715
    },
    {
      "epoch": 4.046121593291405,
      "grad_norm": 0.8207834959030151,
      "learning_rate": 4.842352516099559e-05,
      "loss": 0.269,
      "num_input_tokens_seen": 5040784,
      "step": 7720
    },
    {
      "epoch": 4.04874213836478,
      "grad_norm": 1.1426401138305664,
      "learning_rate": 4.84195265708944e-05,
      "loss": 0.4806,
      "num_input_tokens_seen": 5044176,
      "step": 7725
    },
    {
      "epoch": 4.051362683438155,
      "grad_norm": 1.136000394821167,
      "learning_rate": 4.841552308169545e-05,
      "loss": 0.4261,
      "num_input_tokens_seen": 5047376,
      "step": 7730
    },
    {
      "epoch": 4.05398322851153,
      "grad_norm": 1.0895394086837769,
      "learning_rate": 4.841151469423619e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 5053712,
      "step": 7735
    },
    {
      "epoch": 4.056603773584905,
      "grad_norm": 0.7191569805145264,
      "learning_rate": 4.840750140935516e-05,
      "loss": 0.3722,
      "num_input_tokens_seen": 5057968,
      "step": 7740
    },
    {
      "epoch": 4.059224318658281,
      "grad_norm": 2.441256523132324,
      "learning_rate": 4.840348322789189e-05,
      "loss": 0.4813,
      "num_input_tokens_seen": 5061328,
      "step": 7745
    },
    {
      "epoch": 4.061844863731656,
      "grad_norm": 1.0979472398757935,
      "learning_rate": 4.839946015068692e-05,
      "loss": 0.4019,
      "num_input_tokens_seen": 5064464,
      "step": 7750
    },
    {
      "epoch": 4.064465408805032,
      "grad_norm": 1.1533421277999878,
      "learning_rate": 4.839543217858184e-05,
      "loss": 0.2879,
      "num_input_tokens_seen": 5067344,
      "step": 7755
    },
    {
      "epoch": 4.067085953878407,
      "grad_norm": 0.9242939949035645,
      "learning_rate": 4.839139931241926e-05,
      "loss": 0.2937,
      "num_input_tokens_seen": 5070512,
      "step": 7760
    },
    {
      "epoch": 4.069706498951782,
      "grad_norm": 1.7721658945083618,
      "learning_rate": 4.838736155304281e-05,
      "loss": 0.2993,
      "num_input_tokens_seen": 5073488,
      "step": 7765
    },
    {
      "epoch": 4.072327044025157,
      "grad_norm": 1.5273630619049072,
      "learning_rate": 4.8383318901297135e-05,
      "loss": 0.3447,
      "num_input_tokens_seen": 5079984,
      "step": 7770
    },
    {
      "epoch": 4.0749475890985325,
      "grad_norm": 1.0182654857635498,
      "learning_rate": 4.837927135802792e-05,
      "loss": 0.37,
      "num_input_tokens_seen": 5083056,
      "step": 7775
    },
    {
      "epoch": 4.077568134171908,
      "grad_norm": 1.4329997301101685,
      "learning_rate": 4.8375218924081856e-05,
      "loss": 0.3693,
      "num_input_tokens_seen": 5085904,
      "step": 7780
    },
    {
      "epoch": 4.080188679245283,
      "grad_norm": 1.1641106605529785,
      "learning_rate": 4.837116160030668e-05,
      "loss": 0.3297,
      "num_input_tokens_seen": 5088368,
      "step": 7785
    },
    {
      "epoch": 4.082809224318658,
      "grad_norm": 0.8915539383888245,
      "learning_rate": 4.8367099387551126e-05,
      "loss": 0.3393,
      "num_input_tokens_seen": 5091088,
      "step": 7790
    },
    {
      "epoch": 4.085429769392033,
      "grad_norm": 0.9661099314689636,
      "learning_rate": 4.8363032286664964e-05,
      "loss": 0.366,
      "num_input_tokens_seen": 5094960,
      "step": 7795
    },
    {
      "epoch": 4.088050314465409,
      "grad_norm": 0.9638537764549255,
      "learning_rate": 4.835896029849899e-05,
      "loss": 0.3557,
      "num_input_tokens_seen": 5098352,
      "step": 7800
    },
    {
      "epoch": 4.090670859538784,
      "grad_norm": 1.4538124799728394,
      "learning_rate": 4.835488342390502e-05,
      "loss": 0.362,
      "num_input_tokens_seen": 5102416,
      "step": 7805
    },
    {
      "epoch": 4.09329140461216,
      "grad_norm": 1.1247313022613525,
      "learning_rate": 4.835080166373587e-05,
      "loss": 0.4739,
      "num_input_tokens_seen": 5105072,
      "step": 7810
    },
    {
      "epoch": 4.095911949685535,
      "grad_norm": 1.4985315799713135,
      "learning_rate": 4.834671501884542e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 5108528,
      "step": 7815
    },
    {
      "epoch": 4.09853249475891,
      "grad_norm": 1.1057209968566895,
      "learning_rate": 4.8342623490088543e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 5112176,
      "step": 7820
    },
    {
      "epoch": 4.101153039832285,
      "grad_norm": 0.9175267815589905,
      "learning_rate": 4.833852707832114e-05,
      "loss": 0.3747,
      "num_input_tokens_seen": 5116848,
      "step": 7825
    },
    {
      "epoch": 4.10377358490566,
      "grad_norm": 0.7338249683380127,
      "learning_rate": 4.833442578440013e-05,
      "loss": 0.3673,
      "num_input_tokens_seen": 5120464,
      "step": 7830
    },
    {
      "epoch": 4.106394129979035,
      "grad_norm": 0.6419327259063721,
      "learning_rate": 4.833031960918346e-05,
      "loss": 0.312,
      "num_input_tokens_seen": 5123312,
      "step": 7835
    },
    {
      "epoch": 4.109014675052411,
      "grad_norm": 0.9026173949241638,
      "learning_rate": 4.8326208553530095e-05,
      "loss": 0.3564,
      "num_input_tokens_seen": 5127088,
      "step": 7840
    },
    {
      "epoch": 4.111635220125786,
      "grad_norm": 0.7726770639419556,
      "learning_rate": 4.832209261830002e-05,
      "loss": 0.3596,
      "num_input_tokens_seen": 5130384,
      "step": 7845
    },
    {
      "epoch": 4.114255765199162,
      "grad_norm": 0.8391430377960205,
      "learning_rate": 4.831797180435423e-05,
      "loss": 0.3184,
      "num_input_tokens_seen": 5134288,
      "step": 7850
    },
    {
      "epoch": 4.116876310272537,
      "grad_norm": 1.724987268447876,
      "learning_rate": 4.831384611255478e-05,
      "loss": 0.4354,
      "num_input_tokens_seen": 5137392,
      "step": 7855
    },
    {
      "epoch": 4.119496855345912,
      "grad_norm": 1.264626383781433,
      "learning_rate": 4.8309715543764697e-05,
      "loss": 0.3599,
      "num_input_tokens_seen": 5139824,
      "step": 7860
    },
    {
      "epoch": 4.122117400419287,
      "grad_norm": 0.9819135665893555,
      "learning_rate": 4.8305580098848046e-05,
      "loss": 0.4276,
      "num_input_tokens_seen": 5143280,
      "step": 7865
    },
    {
      "epoch": 4.1247379454926625,
      "grad_norm": 1.3192486763000488,
      "learning_rate": 4.830143977866993e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 5145584,
      "step": 7870
    },
    {
      "epoch": 4.127358490566038,
      "grad_norm": 0.9422605037689209,
      "learning_rate": 4.829729458409645e-05,
      "loss": 0.3316,
      "num_input_tokens_seen": 5148240,
      "step": 7875
    },
    {
      "epoch": 4.129979035639413,
      "grad_norm": 1.231140375137329,
      "learning_rate": 4.8293144515994725e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 5151632,
      "step": 7880
    },
    {
      "epoch": 4.132599580712788,
      "grad_norm": 1.9961808919906616,
      "learning_rate": 4.828898957523292e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 5154576,
      "step": 7885
    },
    {
      "epoch": 4.135220125786163,
      "grad_norm": 1.2486892938613892,
      "learning_rate": 4.8284829762680183e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 5158224,
      "step": 7890
    },
    {
      "epoch": 4.137840670859539,
      "grad_norm": 0.9739739894866943,
      "learning_rate": 4.828066507920671e-05,
      "loss": 0.3147,
      "num_input_tokens_seen": 5161360,
      "step": 7895
    },
    {
      "epoch": 4.140461215932914,
      "grad_norm": 1.3493579626083374,
      "learning_rate": 4.82764955256837e-05,
      "loss": 0.2949,
      "num_input_tokens_seen": 5164656,
      "step": 7900
    },
    {
      "epoch": 4.1430817610062896,
      "grad_norm": 0.8230133056640625,
      "learning_rate": 4.827232110298338e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 5168336,
      "step": 7905
    },
    {
      "epoch": 4.145702306079665,
      "grad_norm": 1.2425345182418823,
      "learning_rate": 4.8268141811978995e-05,
      "loss": 0.3497,
      "num_input_tokens_seen": 5170736,
      "step": 7910
    },
    {
      "epoch": 4.14832285115304,
      "grad_norm": 1.4594351053237915,
      "learning_rate": 4.826395765354479e-05,
      "loss": 0.2175,
      "num_input_tokens_seen": 5173424,
      "step": 7915
    },
    {
      "epoch": 4.150943396226415,
      "grad_norm": 0.94520103931427,
      "learning_rate": 4.825976862855606e-05,
      "loss": 0.3003,
      "num_input_tokens_seen": 5175856,
      "step": 7920
    },
    {
      "epoch": 4.15356394129979,
      "grad_norm": 0.7999460697174072,
      "learning_rate": 4.8255574737889094e-05,
      "loss": 0.3216,
      "num_input_tokens_seen": 5178672,
      "step": 7925
    },
    {
      "epoch": 4.156184486373165,
      "grad_norm": 1.100630283355713,
      "learning_rate": 4.82513759824212e-05,
      "loss": 0.3153,
      "num_input_tokens_seen": 5184496,
      "step": 7930
    },
    {
      "epoch": 4.158805031446541,
      "grad_norm": 0.8390920758247375,
      "learning_rate": 4.824717236303072e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 5188144,
      "step": 7935
    },
    {
      "epoch": 4.161425576519916,
      "grad_norm": 1.9720306396484375,
      "learning_rate": 4.8242963880597005e-05,
      "loss": 0.3471,
      "num_input_tokens_seen": 5192112,
      "step": 7940
    },
    {
      "epoch": 4.164046121593292,
      "grad_norm": 1.4659905433654785,
      "learning_rate": 4.823875053600041e-05,
      "loss": 0.3807,
      "num_input_tokens_seen": 5194320,
      "step": 7945
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 2.41985821723938,
      "learning_rate": 4.823453233012233e-05,
      "loss": 0.3488,
      "num_input_tokens_seen": 5197456,
      "step": 7950
    },
    {
      "epoch": 4.169287211740042,
      "grad_norm": 0.8616049289703369,
      "learning_rate": 4.8230309263845155e-05,
      "loss": 0.3813,
      "num_input_tokens_seen": 5199792,
      "step": 7955
    },
    {
      "epoch": 4.171907756813417,
      "grad_norm": 1.4199012517929077,
      "learning_rate": 4.822608133805231e-05,
      "loss": 0.4268,
      "num_input_tokens_seen": 5202672,
      "step": 7960
    },
    {
      "epoch": 4.1745283018867925,
      "grad_norm": 1.2058526277542114,
      "learning_rate": 4.8221848553628226e-05,
      "loss": 0.4139,
      "num_input_tokens_seen": 5205808,
      "step": 7965
    },
    {
      "epoch": 4.177148846960168,
      "grad_norm": 1.1680599451065063,
      "learning_rate": 4.8217610911458346e-05,
      "loss": 0.2531,
      "num_input_tokens_seen": 5208176,
      "step": 7970
    },
    {
      "epoch": 4.179769392033543,
      "grad_norm": 0.6050263047218323,
      "learning_rate": 4.821336841242915e-05,
      "loss": 0.3658,
      "num_input_tokens_seen": 5215376,
      "step": 7975
    },
    {
      "epoch": 4.182389937106918,
      "grad_norm": 1.1451953649520874,
      "learning_rate": 4.820912105742811e-05,
      "loss": 0.3933,
      "num_input_tokens_seen": 5218064,
      "step": 7980
    },
    {
      "epoch": 4.185010482180293,
      "grad_norm": 1.0573296546936035,
      "learning_rate": 4.8204868847343736e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 5221328,
      "step": 7985
    },
    {
      "epoch": 4.187631027253669,
      "grad_norm": 0.8095621466636658,
      "learning_rate": 4.820061178306553e-05,
      "loss": 0.3734,
      "num_input_tokens_seen": 5224432,
      "step": 7990
    },
    {
      "epoch": 4.190251572327044,
      "grad_norm": 1.3864423036575317,
      "learning_rate": 4.819634986548403e-05,
      "loss": 0.3728,
      "num_input_tokens_seen": 5226896,
      "step": 7995
    },
    {
      "epoch": 4.1928721174004195,
      "grad_norm": 1.0889339447021484,
      "learning_rate": 4.819208309549077e-05,
      "loss": 0.3081,
      "num_input_tokens_seen": 5231056,
      "step": 8000
    },
    {
      "epoch": 4.195492662473795,
      "grad_norm": 1.3575325012207031,
      "learning_rate": 4.8187811473978316e-05,
      "loss": 0.2997,
      "num_input_tokens_seen": 5234480,
      "step": 8005
    },
    {
      "epoch": 4.19811320754717,
      "grad_norm": 3.4542486667633057,
      "learning_rate": 4.818353500184025e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 5236816,
      "step": 8010
    },
    {
      "epoch": 4.200733752620545,
      "grad_norm": 3.5715010166168213,
      "learning_rate": 4.817925367997115e-05,
      "loss": 0.3766,
      "num_input_tokens_seen": 5239600,
      "step": 8015
    },
    {
      "epoch": 4.20335429769392,
      "grad_norm": 1.0285781621932983,
      "learning_rate": 4.8174967509266615e-05,
      "loss": 0.4151,
      "num_input_tokens_seen": 5242896,
      "step": 8020
    },
    {
      "epoch": 4.205974842767295,
      "grad_norm": 1.503268837928772,
      "learning_rate": 4.817067649062328e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 5245904,
      "step": 8025
    },
    {
      "epoch": 4.2085953878406706,
      "grad_norm": 0.9905422329902649,
      "learning_rate": 4.816638062493876e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 5249968,
      "step": 8030
    },
    {
      "epoch": 4.211215932914046,
      "grad_norm": 1.3887603282928467,
      "learning_rate": 4.8162079913111716e-05,
      "loss": 0.3264,
      "num_input_tokens_seen": 5253680,
      "step": 8035
    },
    {
      "epoch": 4.213836477987422,
      "grad_norm": 1.7203937768936157,
      "learning_rate": 4.81577743560418e-05,
      "loss": 0.5064,
      "num_input_tokens_seen": 5256144,
      "step": 8040
    },
    {
      "epoch": 4.216457023060797,
      "grad_norm": 1.14395272731781,
      "learning_rate": 4.8153463954629684e-05,
      "loss": 0.2762,
      "num_input_tokens_seen": 5261456,
      "step": 8045
    },
    {
      "epoch": 4.219077568134172,
      "grad_norm": 1.0460158586502075,
      "learning_rate": 4.814914870977706e-05,
      "loss": 0.4007,
      "num_input_tokens_seen": 5264688,
      "step": 8050
    },
    {
      "epoch": 4.221698113207547,
      "grad_norm": 1.0820523500442505,
      "learning_rate": 4.8144828622386627e-05,
      "loss": 0.4569,
      "num_input_tokens_seen": 5267184,
      "step": 8055
    },
    {
      "epoch": 4.2243186582809225,
      "grad_norm": 1.670304298400879,
      "learning_rate": 4.8140503693362095e-05,
      "loss": 0.3461,
      "num_input_tokens_seen": 5269776,
      "step": 8060
    },
    {
      "epoch": 4.226939203354298,
      "grad_norm": 1.1915403604507446,
      "learning_rate": 4.813617392360818e-05,
      "loss": 0.3675,
      "num_input_tokens_seen": 5273296,
      "step": 8065
    },
    {
      "epoch": 4.229559748427673,
      "grad_norm": 1.0331227779388428,
      "learning_rate": 4.813183931403065e-05,
      "loss": 0.3563,
      "num_input_tokens_seen": 5276432,
      "step": 8070
    },
    {
      "epoch": 4.232180293501048,
      "grad_norm": 1.0136306285858154,
      "learning_rate": 4.8127499865536226e-05,
      "loss": 0.3082,
      "num_input_tokens_seen": 5279696,
      "step": 8075
    },
    {
      "epoch": 4.234800838574423,
      "grad_norm": 1.5070323944091797,
      "learning_rate": 4.812315557903269e-05,
      "loss": 0.3465,
      "num_input_tokens_seen": 5283312,
      "step": 8080
    },
    {
      "epoch": 4.237421383647799,
      "grad_norm": 1.4557815790176392,
      "learning_rate": 4.8118806455428806e-05,
      "loss": 0.273,
      "num_input_tokens_seen": 5285712,
      "step": 8085
    },
    {
      "epoch": 4.240041928721174,
      "grad_norm": 1.4375983476638794,
      "learning_rate": 4.811445249563437e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 5289136,
      "step": 8090
    },
    {
      "epoch": 4.2426624737945495,
      "grad_norm": 0.7901821732521057,
      "learning_rate": 4.811009370056017e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 5292528,
      "step": 8095
    },
    {
      "epoch": 4.245283018867925,
      "grad_norm": 0.9744093418121338,
      "learning_rate": 4.8105730071118016e-05,
      "loss": 0.4117,
      "num_input_tokens_seen": 5295984,
      "step": 8100
    },
    {
      "epoch": 4.2479035639413,
      "grad_norm": 1.1715258359909058,
      "learning_rate": 4.810136160822074e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 5299120,
      "step": 8105
    },
    {
      "epoch": 4.250524109014675,
      "grad_norm": 1.2227531671524048,
      "learning_rate": 4.8096988312782174e-05,
      "loss": 0.3187,
      "num_input_tokens_seen": 5302192,
      "step": 8110
    },
    {
      "epoch": 4.25314465408805,
      "grad_norm": 1.9309266805648804,
      "learning_rate": 4.809261018571715e-05,
      "loss": 0.6206,
      "num_input_tokens_seen": 5304880,
      "step": 8115
    },
    {
      "epoch": 4.255765199161425,
      "grad_norm": 1.3268072605133057,
      "learning_rate": 4.808822722794153e-05,
      "loss": 0.4106,
      "num_input_tokens_seen": 5308304,
      "step": 8120
    },
    {
      "epoch": 4.2583857442348005,
      "grad_norm": 1.3603904247283936,
      "learning_rate": 4.8083839440372194e-05,
      "loss": 0.4103,
      "num_input_tokens_seen": 5310960,
      "step": 8125
    },
    {
      "epoch": 4.261006289308176,
      "grad_norm": 1.3297841548919678,
      "learning_rate": 4.8079446823926984e-05,
      "loss": 0.3483,
      "num_input_tokens_seen": 5313264,
      "step": 8130
    },
    {
      "epoch": 4.263626834381552,
      "grad_norm": 1.684444785118103,
      "learning_rate": 4.8075049379524805e-05,
      "loss": 0.3897,
      "num_input_tokens_seen": 5315600,
      "step": 8135
    },
    {
      "epoch": 4.266247379454927,
      "grad_norm": 0.7043099999427795,
      "learning_rate": 4.807064710808555e-05,
      "loss": 0.2927,
      "num_input_tokens_seen": 5319024,
      "step": 8140
    },
    {
      "epoch": 4.268867924528302,
      "grad_norm": 1.2633920907974243,
      "learning_rate": 4.806624001053013e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 5322160,
      "step": 8145
    },
    {
      "epoch": 4.271488469601677,
      "grad_norm": 1.2937217950820923,
      "learning_rate": 4.8061828087780455e-05,
      "loss": 0.3201,
      "num_input_tokens_seen": 5327152,
      "step": 8150
    },
    {
      "epoch": 4.274109014675052,
      "grad_norm": 2.490375280380249,
      "learning_rate": 4.805741134075944e-05,
      "loss": 0.3016,
      "num_input_tokens_seen": 5330256,
      "step": 8155
    },
    {
      "epoch": 4.276729559748428,
      "grad_norm": 1.6399507522583008,
      "learning_rate": 4.8052989770391026e-05,
      "loss": 0.3903,
      "num_input_tokens_seen": 5333392,
      "step": 8160
    },
    {
      "epoch": 4.279350104821803,
      "grad_norm": 1.2100104093551636,
      "learning_rate": 4.804856337760016e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 5336720,
      "step": 8165
    },
    {
      "epoch": 4.281970649895178,
      "grad_norm": 0.6660452485084534,
      "learning_rate": 4.804413216331278e-05,
      "loss": 0.2722,
      "num_input_tokens_seen": 5339696,
      "step": 8170
    },
    {
      "epoch": 4.284591194968553,
      "grad_norm": 1.4039101600646973,
      "learning_rate": 4.8039696128455856e-05,
      "loss": 0.4589,
      "num_input_tokens_seen": 5342896,
      "step": 8175
    },
    {
      "epoch": 4.287211740041929,
      "grad_norm": 1.467054009437561,
      "learning_rate": 4.803525527395736e-05,
      "loss": 0.3507,
      "num_input_tokens_seen": 5345392,
      "step": 8180
    },
    {
      "epoch": 4.289832285115304,
      "grad_norm": 0.9052250385284424,
      "learning_rate": 4.803080960074625e-05,
      "loss": 0.3204,
      "num_input_tokens_seen": 5349392,
      "step": 8185
    },
    {
      "epoch": 4.2924528301886795,
      "grad_norm": 0.7629581093788147,
      "learning_rate": 4.8026359109752525e-05,
      "loss": 0.3742,
      "num_input_tokens_seen": 5353008,
      "step": 8190
    },
    {
      "epoch": 4.295073375262055,
      "grad_norm": 1.7921019792556763,
      "learning_rate": 4.8021903801907175e-05,
      "loss": 0.5795,
      "num_input_tokens_seen": 5356464,
      "step": 8195
    },
    {
      "epoch": 4.29769392033543,
      "grad_norm": 0.6859711408615112,
      "learning_rate": 4.801744367814219e-05,
      "loss": 0.3045,
      "num_input_tokens_seen": 5359440,
      "step": 8200
    },
    {
      "epoch": 4.300314465408805,
      "grad_norm": 1.9742748737335205,
      "learning_rate": 4.801297873939059e-05,
      "loss": 0.4801,
      "num_input_tokens_seen": 5361968,
      "step": 8205
    },
    {
      "epoch": 4.30293501048218,
      "grad_norm": 1.0551023483276367,
      "learning_rate": 4.8008508986586384e-05,
      "loss": 0.5104,
      "num_input_tokens_seen": 5365744,
      "step": 8210
    },
    {
      "epoch": 4.305555555555555,
      "grad_norm": 1.2719646692276,
      "learning_rate": 4.800403442066459e-05,
      "loss": 0.2969,
      "num_input_tokens_seen": 5368720,
      "step": 8215
    },
    {
      "epoch": 4.3081761006289305,
      "grad_norm": 1.9782341718673706,
      "learning_rate": 4.799955504256125e-05,
      "loss": 0.336,
      "num_input_tokens_seen": 5371408,
      "step": 8220
    },
    {
      "epoch": 4.310796645702306,
      "grad_norm": 1.690124750137329,
      "learning_rate": 4.7995070853213376e-05,
      "loss": 0.411,
      "num_input_tokens_seen": 5374256,
      "step": 8225
    },
    {
      "epoch": 4.313417190775682,
      "grad_norm": 1.73470139503479,
      "learning_rate": 4.799058185355903e-05,
      "loss": 0.4132,
      "num_input_tokens_seen": 5376624,
      "step": 8230
    },
    {
      "epoch": 4.316037735849057,
      "grad_norm": 1.6366348266601562,
      "learning_rate": 4.7986088044537235e-05,
      "loss": 0.3729,
      "num_input_tokens_seen": 5380048,
      "step": 8235
    },
    {
      "epoch": 4.318658280922432,
      "grad_norm": 0.93311607837677,
      "learning_rate": 4.798158942708807e-05,
      "loss": 0.2837,
      "num_input_tokens_seen": 5382416,
      "step": 8240
    },
    {
      "epoch": 4.321278825995807,
      "grad_norm": 1.0985826253890991,
      "learning_rate": 4.797708600215258e-05,
      "loss": 0.4746,
      "num_input_tokens_seen": 5385520,
      "step": 8245
    },
    {
      "epoch": 4.323899371069182,
      "grad_norm": 1.3266310691833496,
      "learning_rate": 4.7972577770672845e-05,
      "loss": 0.5343,
      "num_input_tokens_seen": 5388368,
      "step": 8250
    },
    {
      "epoch": 4.326519916142558,
      "grad_norm": 1.067050814628601,
      "learning_rate": 4.796806473359192e-05,
      "loss": 0.3241,
      "num_input_tokens_seen": 5391920,
      "step": 8255
    },
    {
      "epoch": 4.329140461215933,
      "grad_norm": 1.1251661777496338,
      "learning_rate": 4.796354689185389e-05,
      "loss": 0.2613,
      "num_input_tokens_seen": 5394864,
      "step": 8260
    },
    {
      "epoch": 4.331761006289308,
      "grad_norm": 0.9794865846633911,
      "learning_rate": 4.795902424640383e-05,
      "loss": 0.3523,
      "num_input_tokens_seen": 5398448,
      "step": 8265
    },
    {
      "epoch": 4.334381551362683,
      "grad_norm": 0.8579138517379761,
      "learning_rate": 4.795449679818782e-05,
      "loss": 0.3222,
      "num_input_tokens_seen": 5401904,
      "step": 8270
    },
    {
      "epoch": 4.337002096436059,
      "grad_norm": 1.279823899269104,
      "learning_rate": 4.794996454815297e-05,
      "loss": 0.3074,
      "num_input_tokens_seen": 5404432,
      "step": 8275
    },
    {
      "epoch": 4.339622641509434,
      "grad_norm": 1.549853801727295,
      "learning_rate": 4.7945427497247355e-05,
      "loss": 0.3808,
      "num_input_tokens_seen": 5407376,
      "step": 8280
    },
    {
      "epoch": 4.3422431865828095,
      "grad_norm": 1.3814119100570679,
      "learning_rate": 4.7940885646420084e-05,
      "loss": 0.3589,
      "num_input_tokens_seen": 5410768,
      "step": 8285
    },
    {
      "epoch": 4.344863731656185,
      "grad_norm": 1.034523844718933,
      "learning_rate": 4.7936338996621264e-05,
      "loss": 0.3953,
      "num_input_tokens_seen": 5414256,
      "step": 8290
    },
    {
      "epoch": 4.34748427672956,
      "grad_norm": 2.1484551429748535,
      "learning_rate": 4.7931787548801996e-05,
      "loss": 0.3168,
      "num_input_tokens_seen": 5417296,
      "step": 8295
    },
    {
      "epoch": 4.350104821802935,
      "grad_norm": 1.357748031616211,
      "learning_rate": 4.7927231303914394e-05,
      "loss": 0.3634,
      "num_input_tokens_seen": 5420560,
      "step": 8300
    },
    {
      "epoch": 4.35272536687631,
      "grad_norm": 1.3510619401931763,
      "learning_rate": 4.7922670262911565e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 5424496,
      "step": 8305
    },
    {
      "epoch": 4.355345911949685,
      "grad_norm": 0.997955322265625,
      "learning_rate": 4.791810442674764e-05,
      "loss": 0.3609,
      "num_input_tokens_seen": 5428944,
      "step": 8310
    },
    {
      "epoch": 4.3579664570230605,
      "grad_norm": 0.9123691320419312,
      "learning_rate": 4.791353379637773e-05,
      "loss": 0.4191,
      "num_input_tokens_seen": 5432624,
      "step": 8315
    },
    {
      "epoch": 4.360587002096436,
      "grad_norm": 1.987991213798523,
      "learning_rate": 4.790895837275796e-05,
      "loss": 0.3895,
      "num_input_tokens_seen": 5435152,
      "step": 8320
    },
    {
      "epoch": 4.363207547169811,
      "grad_norm": 1.035747766494751,
      "learning_rate": 4.790437815684545e-05,
      "loss": 0.3536,
      "num_input_tokens_seen": 5438640,
      "step": 8325
    },
    {
      "epoch": 4.365828092243187,
      "grad_norm": 1.3910504579544067,
      "learning_rate": 4.789979314959835e-05,
      "loss": 0.3408,
      "num_input_tokens_seen": 5441264,
      "step": 8330
    },
    {
      "epoch": 4.368448637316562,
      "grad_norm": 0.9060177206993103,
      "learning_rate": 4.789520335197577e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 5444752,
      "step": 8335
    },
    {
      "epoch": 4.371069182389937,
      "grad_norm": 1.7303522825241089,
      "learning_rate": 4.789060876493785e-05,
      "loss": 0.3532,
      "num_input_tokens_seen": 5447632,
      "step": 8340
    },
    {
      "epoch": 4.373689727463312,
      "grad_norm": 1.1267722845077515,
      "learning_rate": 4.788600938944573e-05,
      "loss": 0.4608,
      "num_input_tokens_seen": 5451344,
      "step": 8345
    },
    {
      "epoch": 4.376310272536688,
      "grad_norm": 1.2808505296707153,
      "learning_rate": 4.788140522646155e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 5454736,
      "step": 8350
    },
    {
      "epoch": 4.378930817610063,
      "grad_norm": 0.825915515422821,
      "learning_rate": 4.787679627694843e-05,
      "loss": 0.392,
      "num_input_tokens_seen": 5458096,
      "step": 8355
    },
    {
      "epoch": 4.381551362683438,
      "grad_norm": 1.3312039375305176,
      "learning_rate": 4.787218254187053e-05,
      "loss": 0.3785,
      "num_input_tokens_seen": 5460560,
      "step": 8360
    },
    {
      "epoch": 4.384171907756813,
      "grad_norm": 1.8132309913635254,
      "learning_rate": 4.7867564022192965e-05,
      "loss": 0.3366,
      "num_input_tokens_seen": 5463792,
      "step": 8365
    },
    {
      "epoch": 4.386792452830189,
      "grad_norm": 0.8270902633666992,
      "learning_rate": 4.786294071888191e-05,
      "loss": 0.3248,
      "num_input_tokens_seen": 5467216,
      "step": 8370
    },
    {
      "epoch": 4.389412997903564,
      "grad_norm": 0.7529436349868774,
      "learning_rate": 4.785831263290449e-05,
      "loss": 0.341,
      "num_input_tokens_seen": 5470320,
      "step": 8375
    },
    {
      "epoch": 4.3920335429769395,
      "grad_norm": 0.7746215462684631,
      "learning_rate": 4.7853679765228845e-05,
      "loss": 0.3379,
      "num_input_tokens_seen": 5473232,
      "step": 8380
    },
    {
      "epoch": 4.394654088050315,
      "grad_norm": 1.1747363805770874,
      "learning_rate": 4.7849042116824116e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 5476432,
      "step": 8385
    },
    {
      "epoch": 4.39727463312369,
      "grad_norm": 1.691651463508606,
      "learning_rate": 4.7844399688660456e-05,
      "loss": 0.4278,
      "num_input_tokens_seen": 5478928,
      "step": 8390
    },
    {
      "epoch": 4.399895178197065,
      "grad_norm": 1.1033178567886353,
      "learning_rate": 4.783975248170901e-05,
      "loss": 0.2741,
      "num_input_tokens_seen": 5481776,
      "step": 8395
    },
    {
      "epoch": 4.40251572327044,
      "grad_norm": 0.9324385523796082,
      "learning_rate": 4.783510049694191e-05,
      "loss": 0.3825,
      "num_input_tokens_seen": 5485008,
      "step": 8400
    },
    {
      "epoch": 4.405136268343815,
      "grad_norm": 0.8030489683151245,
      "learning_rate": 4.78304437353323e-05,
      "loss": 0.374,
      "num_input_tokens_seen": 5492496,
      "step": 8405
    },
    {
      "epoch": 4.4077568134171905,
      "grad_norm": 1.4310225248336792,
      "learning_rate": 4.782578219785433e-05,
      "loss": 0.3602,
      "num_input_tokens_seen": 5495664,
      "step": 8410
    },
    {
      "epoch": 4.410377358490566,
      "grad_norm": 1.3777287006378174,
      "learning_rate": 4.782111588548312e-05,
      "loss": 0.3456,
      "num_input_tokens_seen": 5499792,
      "step": 8415
    },
    {
      "epoch": 4.412997903563941,
      "grad_norm": 0.9960631132125854,
      "learning_rate": 4.781644479919484e-05,
      "loss": 0.3161,
      "num_input_tokens_seen": 5502736,
      "step": 8420
    },
    {
      "epoch": 4.415618448637317,
      "grad_norm": 1.3105499744415283,
      "learning_rate": 4.7811768939966605e-05,
      "loss": 0.3342,
      "num_input_tokens_seen": 5505328,
      "step": 8425
    },
    {
      "epoch": 4.418238993710692,
      "grad_norm": 1.4136700630187988,
      "learning_rate": 4.780708830877656e-05,
      "loss": 0.3836,
      "num_input_tokens_seen": 5507952,
      "step": 8430
    },
    {
      "epoch": 4.420859538784067,
      "grad_norm": 2.4346442222595215,
      "learning_rate": 4.780240290660384e-05,
      "loss": 0.3056,
      "num_input_tokens_seen": 5510544,
      "step": 8435
    },
    {
      "epoch": 4.423480083857442,
      "grad_norm": 1.1008329391479492,
      "learning_rate": 4.7797712734428575e-05,
      "loss": 0.4228,
      "num_input_tokens_seen": 5512976,
      "step": 8440
    },
    {
      "epoch": 4.426100628930818,
      "grad_norm": 1.1273630857467651,
      "learning_rate": 4.779301779323189e-05,
      "loss": 0.4265,
      "num_input_tokens_seen": 5516496,
      "step": 8445
    },
    {
      "epoch": 4.428721174004193,
      "grad_norm": 0.9292343854904175,
      "learning_rate": 4.778831808399593e-05,
      "loss": 0.2724,
      "num_input_tokens_seen": 5520720,
      "step": 8450
    },
    {
      "epoch": 4.431341719077568,
      "grad_norm": 0.5874778628349304,
      "learning_rate": 4.77836136077038e-05,
      "loss": 0.3771,
      "num_input_tokens_seen": 5524336,
      "step": 8455
    },
    {
      "epoch": 4.433962264150943,
      "grad_norm": 1.027931809425354,
      "learning_rate": 4.777890436533963e-05,
      "loss": 0.3322,
      "num_input_tokens_seen": 5528112,
      "step": 8460
    },
    {
      "epoch": 4.436582809224318,
      "grad_norm": 1.2436425685882568,
      "learning_rate": 4.777419035788855e-05,
      "loss": 0.3922,
      "num_input_tokens_seen": 5531664,
      "step": 8465
    },
    {
      "epoch": 4.439203354297694,
      "grad_norm": 2.160872220993042,
      "learning_rate": 4.776947158633667e-05,
      "loss": 0.3035,
      "num_input_tokens_seen": 5536368,
      "step": 8470
    },
    {
      "epoch": 4.4418238993710695,
      "grad_norm": 1.154362678527832,
      "learning_rate": 4.7764748051671095e-05,
      "loss": 0.3371,
      "num_input_tokens_seen": 5539600,
      "step": 8475
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.8054919838905334,
      "learning_rate": 4.7760019754879946e-05,
      "loss": 0.2937,
      "num_input_tokens_seen": 5542544,
      "step": 8480
    },
    {
      "epoch": 4.44706498951782,
      "grad_norm": 0.9067243337631226,
      "learning_rate": 4.7755286696952324e-05,
      "loss": 0.3326,
      "num_input_tokens_seen": 5546704,
      "step": 8485
    },
    {
      "epoch": 4.449685534591195,
      "grad_norm": 1.2349488735198975,
      "learning_rate": 4.775054887887833e-05,
      "loss": 0.2956,
      "num_input_tokens_seen": 5549520,
      "step": 8490
    },
    {
      "epoch": 4.45230607966457,
      "grad_norm": 1.2586334943771362,
      "learning_rate": 4.774580630164906e-05,
      "loss": 0.3676,
      "num_input_tokens_seen": 5553040,
      "step": 8495
    },
    {
      "epoch": 4.454926624737945,
      "grad_norm": 1.1682864427566528,
      "learning_rate": 4.774105896625661e-05,
      "loss": 0.3963,
      "num_input_tokens_seen": 5556880,
      "step": 8500
    },
    {
      "epoch": 4.4575471698113205,
      "grad_norm": 1.2204103469848633,
      "learning_rate": 4.773630687369407e-05,
      "loss": 0.3225,
      "num_input_tokens_seen": 5559760,
      "step": 8505
    },
    {
      "epoch": 4.460167714884696,
      "grad_norm": 1.2595806121826172,
      "learning_rate": 4.773155002495552e-05,
      "loss": 0.3468,
      "num_input_tokens_seen": 5564688,
      "step": 8510
    },
    {
      "epoch": 4.462788259958071,
      "grad_norm": 1.510389804840088,
      "learning_rate": 4.772678842103605e-05,
      "loss": 0.4322,
      "num_input_tokens_seen": 5568112,
      "step": 8515
    },
    {
      "epoch": 4.465408805031447,
      "grad_norm": 1.1296550035476685,
      "learning_rate": 4.772202206293172e-05,
      "loss": 0.4696,
      "num_input_tokens_seen": 5570928,
      "step": 8520
    },
    {
      "epoch": 4.468029350104822,
      "grad_norm": 1.406389594078064,
      "learning_rate": 4.7717250951639586e-05,
      "loss": 0.3574,
      "num_input_tokens_seen": 5574480,
      "step": 8525
    },
    {
      "epoch": 4.470649895178197,
      "grad_norm": 2.0556647777557373,
      "learning_rate": 4.771247508815774e-05,
      "loss": 0.3725,
      "num_input_tokens_seen": 5577296,
      "step": 8530
    },
    {
      "epoch": 4.473270440251572,
      "grad_norm": 0.6910753846168518,
      "learning_rate": 4.770769447348522e-05,
      "loss": 0.2617,
      "num_input_tokens_seen": 5580784,
      "step": 8535
    },
    {
      "epoch": 4.475890985324948,
      "grad_norm": 0.950068473815918,
      "learning_rate": 4.7702909108622074e-05,
      "loss": 0.3372,
      "num_input_tokens_seen": 5583856,
      "step": 8540
    },
    {
      "epoch": 4.478511530398323,
      "grad_norm": 1.2922494411468506,
      "learning_rate": 4.7698118994569355e-05,
      "loss": 0.3364,
      "num_input_tokens_seen": 5586256,
      "step": 8545
    },
    {
      "epoch": 4.481132075471698,
      "grad_norm": 1.0639622211456299,
      "learning_rate": 4.769332413232909e-05,
      "loss": 0.3479,
      "num_input_tokens_seen": 5589200,
      "step": 8550
    },
    {
      "epoch": 4.483752620545073,
      "grad_norm": 0.8757538795471191,
      "learning_rate": 4.768852452290432e-05,
      "loss": 0.4229,
      "num_input_tokens_seen": 5592976,
      "step": 8555
    },
    {
      "epoch": 4.486373165618448,
      "grad_norm": 1.1065958738327026,
      "learning_rate": 4.768372016729907e-05,
      "loss": 0.3812,
      "num_input_tokens_seen": 5595696,
      "step": 8560
    },
    {
      "epoch": 4.488993710691824,
      "grad_norm": 1.3029744625091553,
      "learning_rate": 4.767891106651834e-05,
      "loss": 0.4033,
      "num_input_tokens_seen": 5598288,
      "step": 8565
    },
    {
      "epoch": 4.4916142557651995,
      "grad_norm": 1.3836170434951782,
      "learning_rate": 4.767409722156816e-05,
      "loss": 0.4444,
      "num_input_tokens_seen": 5601296,
      "step": 8570
    },
    {
      "epoch": 4.494234800838575,
      "grad_norm": 1.377317190170288,
      "learning_rate": 4.7669278633455515e-05,
      "loss": 0.3087,
      "num_input_tokens_seen": 5604752,
      "step": 8575
    },
    {
      "epoch": 4.49685534591195,
      "grad_norm": 0.9923830032348633,
      "learning_rate": 4.7664455303188404e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 5607824,
      "step": 8580
    },
    {
      "epoch": 4.499475890985325,
      "grad_norm": 1.198838710784912,
      "learning_rate": 4.7659627231775815e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 5611248,
      "step": 8585
    },
    {
      "epoch": 4.5020964360587,
      "grad_norm": 0.59546959400177,
      "learning_rate": 4.765479442022773e-05,
      "loss": 0.3121,
      "num_input_tokens_seen": 5614192,
      "step": 8590
    },
    {
      "epoch": 4.504716981132075,
      "grad_norm": 1.591599702835083,
      "learning_rate": 4.76499568695551e-05,
      "loss": 0.3894,
      "num_input_tokens_seen": 5616944,
      "step": 8595
    },
    {
      "epoch": 4.5073375262054505,
      "grad_norm": 1.4099866151809692,
      "learning_rate": 4.764511458076991e-05,
      "loss": 0.3693,
      "num_input_tokens_seen": 5619344,
      "step": 8600
    },
    {
      "epoch": 4.509958071278826,
      "grad_norm": 1.091892957687378,
      "learning_rate": 4.7640267554885085e-05,
      "loss": 0.2901,
      "num_input_tokens_seen": 5622608,
      "step": 8605
    },
    {
      "epoch": 4.512578616352201,
      "grad_norm": 1.42496919631958,
      "learning_rate": 4.763541579291459e-05,
      "loss": 0.3918,
      "num_input_tokens_seen": 5625296,
      "step": 8610
    },
    {
      "epoch": 4.515199161425577,
      "grad_norm": 0.8977536559104919,
      "learning_rate": 4.763055929587336e-05,
      "loss": 0.2757,
      "num_input_tokens_seen": 5628272,
      "step": 8615
    },
    {
      "epoch": 4.517819706498952,
      "grad_norm": 0.9496619701385498,
      "learning_rate": 4.7625698064777294e-05,
      "loss": 0.3261,
      "num_input_tokens_seen": 5632944,
      "step": 8620
    },
    {
      "epoch": 4.520440251572327,
      "grad_norm": 1.6606007814407349,
      "learning_rate": 4.762083210064332e-05,
      "loss": 0.333,
      "num_input_tokens_seen": 5636016,
      "step": 8625
    },
    {
      "epoch": 4.523060796645702,
      "grad_norm": 1.6074339151382446,
      "learning_rate": 4.761596140448934e-05,
      "loss": 0.2938,
      "num_input_tokens_seen": 5642064,
      "step": 8630
    },
    {
      "epoch": 4.5256813417190775,
      "grad_norm": 1.1448179483413696,
      "learning_rate": 4.7611085977334256e-05,
      "loss": 0.4475,
      "num_input_tokens_seen": 5645072,
      "step": 8635
    },
    {
      "epoch": 4.528301886792453,
      "grad_norm": 0.8962420225143433,
      "learning_rate": 4.7606205820197946e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 5648304,
      "step": 8640
    },
    {
      "epoch": 4.530922431865828,
      "grad_norm": 0.9848628640174866,
      "learning_rate": 4.760132093410129e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 5651120,
      "step": 8645
    },
    {
      "epoch": 4.533542976939203,
      "grad_norm": 1.1194193363189697,
      "learning_rate": 4.759643132006614e-05,
      "loss": 0.2774,
      "num_input_tokens_seen": 5653712,
      "step": 8650
    },
    {
      "epoch": 4.536163522012579,
      "grad_norm": 1.1351158618927002,
      "learning_rate": 4.759153697911535e-05,
      "loss": 0.3248,
      "num_input_tokens_seen": 5656624,
      "step": 8655
    },
    {
      "epoch": 4.538784067085954,
      "grad_norm": 1.5532195568084717,
      "learning_rate": 4.758663791227276e-05,
      "loss": 0.3754,
      "num_input_tokens_seen": 5658960,
      "step": 8660
    },
    {
      "epoch": 4.5414046121593294,
      "grad_norm": 1.5126252174377441,
      "learning_rate": 4.7581734120563206e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 5661872,
      "step": 8665
    },
    {
      "epoch": 4.544025157232705,
      "grad_norm": 1.1415789127349854,
      "learning_rate": 4.757682560501249e-05,
      "loss": 0.4336,
      "num_input_tokens_seen": 5664560,
      "step": 8670
    },
    {
      "epoch": 4.54664570230608,
      "grad_norm": 1.3059381246566772,
      "learning_rate": 4.757191236664744e-05,
      "loss": 0.3309,
      "num_input_tokens_seen": 5667664,
      "step": 8675
    },
    {
      "epoch": 4.549266247379455,
      "grad_norm": 1.5161789655685425,
      "learning_rate": 4.7566994406495825e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 5670128,
      "step": 8680
    },
    {
      "epoch": 4.55188679245283,
      "grad_norm": 1.400935411453247,
      "learning_rate": 4.756207172558645e-05,
      "loss": 0.4222,
      "num_input_tokens_seen": 5673296,
      "step": 8685
    },
    {
      "epoch": 4.554507337526205,
      "grad_norm": 1.5185573101043701,
      "learning_rate": 4.7557144324949065e-05,
      "loss": 0.4472,
      "num_input_tokens_seen": 5675920,
      "step": 8690
    },
    {
      "epoch": 4.5571278825995805,
      "grad_norm": 2.4118502140045166,
      "learning_rate": 4.755221220561443e-05,
      "loss": 0.3139,
      "num_input_tokens_seen": 5678896,
      "step": 8695
    },
    {
      "epoch": 4.559748427672956,
      "grad_norm": 2.2883689403533936,
      "learning_rate": 4.7547275368614295e-05,
      "loss": 0.3659,
      "num_input_tokens_seen": 5681488,
      "step": 8700
    },
    {
      "epoch": 4.562368972746331,
      "grad_norm": 2.0868959426879883,
      "learning_rate": 4.754233381498139e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 5684432,
      "step": 8705
    },
    {
      "epoch": 4.564989517819707,
      "grad_norm": 0.7729544639587402,
      "learning_rate": 4.753738754574942e-05,
      "loss": 0.3558,
      "num_input_tokens_seen": 5688112,
      "step": 8710
    },
    {
      "epoch": 4.567610062893082,
      "grad_norm": 0.7623236179351807,
      "learning_rate": 4.75324365619531e-05,
      "loss": 0.4023,
      "num_input_tokens_seen": 5690416,
      "step": 8715
    },
    {
      "epoch": 4.570230607966457,
      "grad_norm": 1.6285557746887207,
      "learning_rate": 4.752748086462811e-05,
      "loss": 0.4179,
      "num_input_tokens_seen": 5693744,
      "step": 8720
    },
    {
      "epoch": 4.572851153039832,
      "grad_norm": 0.8048228621482849,
      "learning_rate": 4.752252045481113e-05,
      "loss": 0.3855,
      "num_input_tokens_seen": 5697776,
      "step": 8725
    },
    {
      "epoch": 4.5754716981132075,
      "grad_norm": 1.7365210056304932,
      "learning_rate": 4.7517555333539834e-05,
      "loss": 0.5682,
      "num_input_tokens_seen": 5700240,
      "step": 8730
    },
    {
      "epoch": 4.578092243186583,
      "grad_norm": 1.0664663314819336,
      "learning_rate": 4.751258550185284e-05,
      "loss": 0.3339,
      "num_input_tokens_seen": 5702992,
      "step": 8735
    },
    {
      "epoch": 4.580712788259958,
      "grad_norm": 2.607496976852417,
      "learning_rate": 4.7507610960789804e-05,
      "loss": 0.3337,
      "num_input_tokens_seen": 5706544,
      "step": 8740
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 1.4268417358398438,
      "learning_rate": 4.7502631711391334e-05,
      "loss": 0.3701,
      "num_input_tokens_seen": 5709328,
      "step": 8745
    },
    {
      "epoch": 4.585953878406709,
      "grad_norm": 0.8560550808906555,
      "learning_rate": 4.749764775469904e-05,
      "loss": 0.3228,
      "num_input_tokens_seen": 5712720,
      "step": 8750
    },
    {
      "epoch": 4.588574423480084,
      "grad_norm": 1.2560005187988281,
      "learning_rate": 4.749265909175549e-05,
      "loss": 0.4431,
      "num_input_tokens_seen": 5715536,
      "step": 8755
    },
    {
      "epoch": 4.591194968553459,
      "grad_norm": 1.4145058393478394,
      "learning_rate": 4.7487665723604275e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 5718352,
      "step": 8760
    },
    {
      "epoch": 4.593815513626835,
      "grad_norm": 0.7449318766593933,
      "learning_rate": 4.7482667651289944e-05,
      "loss": 0.3594,
      "num_input_tokens_seen": 5722096,
      "step": 8765
    },
    {
      "epoch": 4.59643605870021,
      "grad_norm": 1.35439932346344,
      "learning_rate": 4.747766487585803e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 5724816,
      "step": 8770
    },
    {
      "epoch": 4.599056603773585,
      "grad_norm": 1.3077634572982788,
      "learning_rate": 4.747265739835507e-05,
      "loss": 0.3598,
      "num_input_tokens_seen": 5729040,
      "step": 8775
    },
    {
      "epoch": 4.60167714884696,
      "grad_norm": 1.3257979154586792,
      "learning_rate": 4.746764521982856e-05,
      "loss": 0.3614,
      "num_input_tokens_seen": 5732464,
      "step": 8780
    },
    {
      "epoch": 4.604297693920335,
      "grad_norm": 1.0334008932113647,
      "learning_rate": 4.7462628341326995e-05,
      "loss": 0.2482,
      "num_input_tokens_seen": 5735472,
      "step": 8785
    },
    {
      "epoch": 4.6069182389937104,
      "grad_norm": 1.205169677734375,
      "learning_rate": 4.7457606763899856e-05,
      "loss": 0.3582,
      "num_input_tokens_seen": 5738896,
      "step": 8790
    },
    {
      "epoch": 4.609538784067086,
      "grad_norm": 0.722272515296936,
      "learning_rate": 4.745258048859758e-05,
      "loss": 0.2598,
      "num_input_tokens_seen": 5742032,
      "step": 8795
    },
    {
      "epoch": 4.612159329140461,
      "grad_norm": 0.9347507953643799,
      "learning_rate": 4.744754951647162e-05,
      "loss": 0.3728,
      "num_input_tokens_seen": 5745168,
      "step": 8800
    },
    {
      "epoch": 4.614779874213837,
      "grad_norm": 1.5650969743728638,
      "learning_rate": 4.74425138485744e-05,
      "loss": 0.446,
      "num_input_tokens_seen": 5747920,
      "step": 8805
    },
    {
      "epoch": 4.617400419287212,
      "grad_norm": 1.4631344079971313,
      "learning_rate": 4.743747348595932e-05,
      "loss": 0.367,
      "num_input_tokens_seen": 5751120,
      "step": 8810
    },
    {
      "epoch": 4.620020964360587,
      "grad_norm": 1.2989332675933838,
      "learning_rate": 4.743242842968077e-05,
      "loss": 0.3672,
      "num_input_tokens_seen": 5754352,
      "step": 8815
    },
    {
      "epoch": 4.622641509433962,
      "grad_norm": 3.213423013687134,
      "learning_rate": 4.742737868079411e-05,
      "loss": 0.3404,
      "num_input_tokens_seen": 5757200,
      "step": 8820
    },
    {
      "epoch": 4.6252620545073375,
      "grad_norm": 1.3104385137557983,
      "learning_rate": 4.7422324240355696e-05,
      "loss": 0.2625,
      "num_input_tokens_seen": 5760944,
      "step": 8825
    },
    {
      "epoch": 4.627882599580713,
      "grad_norm": 1.2159061431884766,
      "learning_rate": 4.741726510942286e-05,
      "loss": 0.404,
      "num_input_tokens_seen": 5764848,
      "step": 8830
    },
    {
      "epoch": 4.630503144654088,
      "grad_norm": 1.032213568687439,
      "learning_rate": 4.741220128905391e-05,
      "loss": 0.3113,
      "num_input_tokens_seen": 5767984,
      "step": 8835
    },
    {
      "epoch": 4.633123689727463,
      "grad_norm": 0.9644655585289001,
      "learning_rate": 4.740713278030815e-05,
      "loss": 0.3048,
      "num_input_tokens_seen": 5771472,
      "step": 8840
    },
    {
      "epoch": 4.635744234800838,
      "grad_norm": 1.8224502801895142,
      "learning_rate": 4.740205958424584e-05,
      "loss": 0.3674,
      "num_input_tokens_seen": 5775536,
      "step": 8845
    },
    {
      "epoch": 4.638364779874214,
      "grad_norm": 1.2894327640533447,
      "learning_rate": 4.739698170192824e-05,
      "loss": 0.3667,
      "num_input_tokens_seen": 5778704,
      "step": 8850
    },
    {
      "epoch": 4.640985324947589,
      "grad_norm": 1.7509524822235107,
      "learning_rate": 4.7391899134417596e-05,
      "loss": 0.3682,
      "num_input_tokens_seen": 5781104,
      "step": 8855
    },
    {
      "epoch": 4.643605870020965,
      "grad_norm": 1.1441689729690552,
      "learning_rate": 4.73868118827771e-05,
      "loss": 0.4055,
      "num_input_tokens_seen": 5784528,
      "step": 8860
    },
    {
      "epoch": 4.64622641509434,
      "grad_norm": 1.2502371072769165,
      "learning_rate": 4.738171994807097e-05,
      "loss": 0.5805,
      "num_input_tokens_seen": 5787184,
      "step": 8865
    },
    {
      "epoch": 4.648846960167715,
      "grad_norm": 1.8047382831573486,
      "learning_rate": 4.7376623331364364e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 5790032,
      "step": 8870
    },
    {
      "epoch": 4.65146750524109,
      "grad_norm": 1.3174197673797607,
      "learning_rate": 4.737152203372345e-05,
      "loss": 0.3737,
      "num_input_tokens_seen": 5793104,
      "step": 8875
    },
    {
      "epoch": 4.654088050314465,
      "grad_norm": 2.1660382747650146,
      "learning_rate": 4.7366416056215345e-05,
      "loss": 0.4095,
      "num_input_tokens_seen": 5796368,
      "step": 8880
    },
    {
      "epoch": 4.65670859538784,
      "grad_norm": 1.9955555200576782,
      "learning_rate": 4.7361305399908185e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 5799248,
      "step": 8885
    },
    {
      "epoch": 4.659329140461216,
      "grad_norm": 0.9044361114501953,
      "learning_rate": 4.735619006587103e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 5802448,
      "step": 8890
    },
    {
      "epoch": 4.661949685534591,
      "grad_norm": 1.2488735914230347,
      "learning_rate": 4.735107005517398e-05,
      "loss": 0.4161,
      "num_input_tokens_seen": 5805936,
      "step": 8895
    },
    {
      "epoch": 4.664570230607967,
      "grad_norm": 1.5738182067871094,
      "learning_rate": 4.734594536888806e-05,
      "loss": 0.3092,
      "num_input_tokens_seen": 5809744,
      "step": 8900
    },
    {
      "epoch": 4.667190775681342,
      "grad_norm": 1.170397162437439,
      "learning_rate": 4.734081600808531e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 5812656,
      "step": 8905
    },
    {
      "epoch": 4.669811320754717,
      "grad_norm": 2.305769443511963,
      "learning_rate": 4.733568197383872e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 5816048,
      "step": 8910
    },
    {
      "epoch": 4.672431865828092,
      "grad_norm": 0.9217071533203125,
      "learning_rate": 4.7330543267222294e-05,
      "loss": 0.4796,
      "num_input_tokens_seen": 5820080,
      "step": 8915
    },
    {
      "epoch": 4.6750524109014675,
      "grad_norm": 0.7888553738594055,
      "learning_rate": 4.732539988931097e-05,
      "loss": 0.4074,
      "num_input_tokens_seen": 5823344,
      "step": 8920
    },
    {
      "epoch": 4.677672955974843,
      "grad_norm": 1.0872032642364502,
      "learning_rate": 4.732025184118068e-05,
      "loss": 0.3591,
      "num_input_tokens_seen": 5826448,
      "step": 8925
    },
    {
      "epoch": 4.680293501048218,
      "grad_norm": 2.7942965030670166,
      "learning_rate": 4.731509912390837e-05,
      "loss": 0.3282,
      "num_input_tokens_seen": 5830000,
      "step": 8930
    },
    {
      "epoch": 4.682914046121593,
      "grad_norm": 1.1643545627593994,
      "learning_rate": 4.730994173857189e-05,
      "loss": 0.2842,
      "num_input_tokens_seen": 5832464,
      "step": 8935
    },
    {
      "epoch": 4.685534591194968,
      "grad_norm": 1.0662139654159546,
      "learning_rate": 4.7304779686250126e-05,
      "loss": 0.3502,
      "num_input_tokens_seen": 5835568,
      "step": 8940
    },
    {
      "epoch": 4.688155136268344,
      "grad_norm": 1.2425007820129395,
      "learning_rate": 4.729961296802292e-05,
      "loss": 0.35,
      "num_input_tokens_seen": 5841360,
      "step": 8945
    },
    {
      "epoch": 4.690775681341719,
      "grad_norm": 1.340151071548462,
      "learning_rate": 4.7294441584971095e-05,
      "loss": 0.2893,
      "num_input_tokens_seen": 5844240,
      "step": 8950
    },
    {
      "epoch": 4.693396226415095,
      "grad_norm": 1.3502776622772217,
      "learning_rate": 4.728926553817643e-05,
      "loss": 0.2843,
      "num_input_tokens_seen": 5848144,
      "step": 8955
    },
    {
      "epoch": 4.69601677148847,
      "grad_norm": 1.3451123237609863,
      "learning_rate": 4.728408482872171e-05,
      "loss": 0.433,
      "num_input_tokens_seen": 5850800,
      "step": 8960
    },
    {
      "epoch": 4.698637316561845,
      "grad_norm": 1.0204932689666748,
      "learning_rate": 4.727889945769067e-05,
      "loss": 0.3634,
      "num_input_tokens_seen": 5853712,
      "step": 8965
    },
    {
      "epoch": 4.70125786163522,
      "grad_norm": 1.1589381694793701,
      "learning_rate": 4.727370942616803e-05,
      "loss": 0.3336,
      "num_input_tokens_seen": 5856400,
      "step": 8970
    },
    {
      "epoch": 4.703878406708595,
      "grad_norm": 1.1005322933197021,
      "learning_rate": 4.72685147352395e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 5859760,
      "step": 8975
    },
    {
      "epoch": 4.70649895178197,
      "grad_norm": 0.8283928036689758,
      "learning_rate": 4.7263315385991735e-05,
      "loss": 0.2979,
      "num_input_tokens_seen": 5862672,
      "step": 8980
    },
    {
      "epoch": 4.709119496855346,
      "grad_norm": 1.595271348953247,
      "learning_rate": 4.7258111379512384e-05,
      "loss": 0.3644,
      "num_input_tokens_seen": 5866288,
      "step": 8985
    },
    {
      "epoch": 4.711740041928721,
      "grad_norm": 1.0952712297439575,
      "learning_rate": 4.725290271689007e-05,
      "loss": 0.4432,
      "num_input_tokens_seen": 5869424,
      "step": 8990
    },
    {
      "epoch": 4.714360587002097,
      "grad_norm": 0.7360647320747375,
      "learning_rate": 4.724768939921438e-05,
      "loss": 0.2653,
      "num_input_tokens_seen": 5872240,
      "step": 8995
    },
    {
      "epoch": 4.716981132075472,
      "grad_norm": 1.2779178619384766,
      "learning_rate": 4.7242471427575884e-05,
      "loss": 0.2932,
      "num_input_tokens_seen": 5874672,
      "step": 9000
    },
    {
      "epoch": 4.719601677148847,
      "grad_norm": 0.8277409076690674,
      "learning_rate": 4.723724880306612e-05,
      "loss": 0.3062,
      "num_input_tokens_seen": 5878544,
      "step": 9005
    },
    {
      "epoch": 4.722222222222222,
      "grad_norm": 1.9168635606765747,
      "learning_rate": 4.72320215267776e-05,
      "loss": 0.421,
      "num_input_tokens_seen": 5882096,
      "step": 9010
    },
    {
      "epoch": 4.7248427672955975,
      "grad_norm": 1.2063502073287964,
      "learning_rate": 4.722678959980381e-05,
      "loss": 0.3799,
      "num_input_tokens_seen": 5886512,
      "step": 9015
    },
    {
      "epoch": 4.727463312368973,
      "grad_norm": 1.1176519393920898,
      "learning_rate": 4.7221553023239216e-05,
      "loss": 0.2996,
      "num_input_tokens_seen": 5889648,
      "step": 9020
    },
    {
      "epoch": 4.730083857442348,
      "grad_norm": 1.5791270732879639,
      "learning_rate": 4.721631179817925e-05,
      "loss": 0.3455,
      "num_input_tokens_seen": 5892752,
      "step": 9025
    },
    {
      "epoch": 4.732704402515723,
      "grad_norm": 0.9601627588272095,
      "learning_rate": 4.721106592572031e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 5896112,
      "step": 9030
    },
    {
      "epoch": 4.735324947589098,
      "grad_norm": 1.1542524099349976,
      "learning_rate": 4.7205815406959774e-05,
      "loss": 0.3456,
      "num_input_tokens_seen": 5899440,
      "step": 9035
    },
    {
      "epoch": 4.737945492662474,
      "grad_norm": 1.412641167640686,
      "learning_rate": 4.7200560242996e-05,
      "loss": 0.4645,
      "num_input_tokens_seen": 5902096,
      "step": 9040
    },
    {
      "epoch": 4.740566037735849,
      "grad_norm": 1.0936100482940674,
      "learning_rate": 4.719530043492829e-05,
      "loss": 0.3478,
      "num_input_tokens_seen": 5904880,
      "step": 9045
    },
    {
      "epoch": 4.743186582809225,
      "grad_norm": 1.109804630279541,
      "learning_rate": 4.7190035983856954e-05,
      "loss": 0.4116,
      "num_input_tokens_seen": 5907632,
      "step": 9050
    },
    {
      "epoch": 4.7458071278826,
      "grad_norm": 1.155685305595398,
      "learning_rate": 4.718476689088325e-05,
      "loss": 0.3145,
      "num_input_tokens_seen": 5911120,
      "step": 9055
    },
    {
      "epoch": 4.748427672955975,
      "grad_norm": 0.7908098697662354,
      "learning_rate": 4.71794931571094e-05,
      "loss": 0.3193,
      "num_input_tokens_seen": 5914928,
      "step": 9060
    },
    {
      "epoch": 4.75104821802935,
      "grad_norm": 0.9476117491722107,
      "learning_rate": 4.717421478363862e-05,
      "loss": 0.3295,
      "num_input_tokens_seen": 5918192,
      "step": 9065
    },
    {
      "epoch": 4.753668763102725,
      "grad_norm": 1.0521379709243774,
      "learning_rate": 4.7168931771575085e-05,
      "loss": 0.2664,
      "num_input_tokens_seen": 5921232,
      "step": 9070
    },
    {
      "epoch": 4.7562893081761,
      "grad_norm": 1.0785818099975586,
      "learning_rate": 4.716364412202394e-05,
      "loss": 0.3054,
      "num_input_tokens_seen": 5925040,
      "step": 9075
    },
    {
      "epoch": 4.758909853249476,
      "grad_norm": 1.4604980945587158,
      "learning_rate": 4.7158351836091306e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 5927376,
      "step": 9080
    },
    {
      "epoch": 4.761530398322851,
      "grad_norm": 0.7918881773948669,
      "learning_rate": 4.7153054914884266e-05,
      "loss": 0.3451,
      "num_input_tokens_seen": 5930576,
      "step": 9085
    },
    {
      "epoch": 4.764150943396227,
      "grad_norm": 2.179694414138794,
      "learning_rate": 4.714775335951086e-05,
      "loss": 0.3036,
      "num_input_tokens_seen": 5933328,
      "step": 9090
    },
    {
      "epoch": 4.766771488469602,
      "grad_norm": 0.7443260550498962,
      "learning_rate": 4.7142447171080137e-05,
      "loss": 0.3708,
      "num_input_tokens_seen": 5936208,
      "step": 9095
    },
    {
      "epoch": 4.769392033542977,
      "grad_norm": 1.0681755542755127,
      "learning_rate": 4.713713635070208e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 5938832,
      "step": 9100
    },
    {
      "epoch": 4.772012578616352,
      "grad_norm": 1.1711331605911255,
      "learning_rate": 4.713182089948765e-05,
      "loss": 0.2813,
      "num_input_tokens_seen": 5942128,
      "step": 9105
    },
    {
      "epoch": 4.7746331236897275,
      "grad_norm": 0.9026827216148376,
      "learning_rate": 4.712650081854878e-05,
      "loss": 0.2277,
      "num_input_tokens_seen": 5945648,
      "step": 9110
    },
    {
      "epoch": 4.777253668763103,
      "grad_norm": 1.0373995304107666,
      "learning_rate": 4.712117610899837e-05,
      "loss": 0.305,
      "num_input_tokens_seen": 5950352,
      "step": 9115
    },
    {
      "epoch": 4.779874213836478,
      "grad_norm": 1.501562237739563,
      "learning_rate": 4.7115846771950286e-05,
      "loss": 0.4747,
      "num_input_tokens_seen": 5953488,
      "step": 9120
    },
    {
      "epoch": 4.782494758909853,
      "grad_norm": 1.0891956090927124,
      "learning_rate": 4.711051280851938e-05,
      "loss": 0.3468,
      "num_input_tokens_seen": 5956784,
      "step": 9125
    },
    {
      "epoch": 4.785115303983228,
      "grad_norm": 1.4967347383499146,
      "learning_rate": 4.7105174219821435e-05,
      "loss": 0.3956,
      "num_input_tokens_seen": 5960080,
      "step": 9130
    },
    {
      "epoch": 4.787735849056604,
      "grad_norm": 1.0458524227142334,
      "learning_rate": 4.709983100697323e-05,
      "loss": 0.2839,
      "num_input_tokens_seen": 5962864,
      "step": 9135
    },
    {
      "epoch": 4.790356394129979,
      "grad_norm": 0.8619617223739624,
      "learning_rate": 4.709448317109252e-05,
      "loss": 0.3351,
      "num_input_tokens_seen": 5966000,
      "step": 9140
    },
    {
      "epoch": 4.7929769392033545,
      "grad_norm": 0.9923753142356873,
      "learning_rate": 4.708913071329798e-05,
      "loss": 0.3849,
      "num_input_tokens_seen": 5969168,
      "step": 9145
    },
    {
      "epoch": 4.79559748427673,
      "grad_norm": 0.8414306044578552,
      "learning_rate": 4.708377363470932e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 5972720,
      "step": 9150
    },
    {
      "epoch": 4.798218029350105,
      "grad_norm": 0.8288646340370178,
      "learning_rate": 4.7078411936447144e-05,
      "loss": 0.253,
      "num_input_tokens_seen": 5975760,
      "step": 9155
    },
    {
      "epoch": 4.80083857442348,
      "grad_norm": 1.343782663345337,
      "learning_rate": 4.707304561963308e-05,
      "loss": 0.2767,
      "num_input_tokens_seen": 5979120,
      "step": 9160
    },
    {
      "epoch": 4.803459119496855,
      "grad_norm": 0.8468167185783386,
      "learning_rate": 4.70676746853897e-05,
      "loss": 0.4002,
      "num_input_tokens_seen": 5983376,
      "step": 9165
    },
    {
      "epoch": 4.80607966457023,
      "grad_norm": 1.6172020435333252,
      "learning_rate": 4.706229913484054e-05,
      "loss": 0.4218,
      "num_input_tokens_seen": 5986416,
      "step": 9170
    },
    {
      "epoch": 4.808700209643606,
      "grad_norm": 2.1257083415985107,
      "learning_rate": 4.705691896911009e-05,
      "loss": 0.2802,
      "num_input_tokens_seen": 5989136,
      "step": 9175
    },
    {
      "epoch": 4.811320754716981,
      "grad_norm": 1.4203699827194214,
      "learning_rate": 4.705153418932383e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 5992112,
      "step": 9180
    },
    {
      "epoch": 4.813941299790356,
      "grad_norm": 1.4251188039779663,
      "learning_rate": 4.7046144796608206e-05,
      "loss": 0.2862,
      "num_input_tokens_seen": 5995472,
      "step": 9185
    },
    {
      "epoch": 4.816561844863732,
      "grad_norm": 0.8401001691818237,
      "learning_rate": 4.70407507920906e-05,
      "loss": 0.3115,
      "num_input_tokens_seen": 5998736,
      "step": 9190
    },
    {
      "epoch": 4.819182389937107,
      "grad_norm": 1.5310262441635132,
      "learning_rate": 4.7035352176899384e-05,
      "loss": 0.4981,
      "num_input_tokens_seen": 6001552,
      "step": 9195
    },
    {
      "epoch": 4.821802935010482,
      "grad_norm": 1.5768941640853882,
      "learning_rate": 4.7029948952163885e-05,
      "loss": 0.3869,
      "num_input_tokens_seen": 6005520,
      "step": 9200
    },
    {
      "epoch": 4.8244234800838575,
      "grad_norm": 1.4058291912078857,
      "learning_rate": 4.7024541119014386e-05,
      "loss": 0.3469,
      "num_input_tokens_seen": 6009200,
      "step": 9205
    },
    {
      "epoch": 4.827044025157233,
      "grad_norm": 1.4050214290618896,
      "learning_rate": 4.701912867858217e-05,
      "loss": 0.3385,
      "num_input_tokens_seen": 6012912,
      "step": 9210
    },
    {
      "epoch": 4.829664570230608,
      "grad_norm": 1.1012938022613525,
      "learning_rate": 4.7013711631999434e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 6017744,
      "step": 9215
    },
    {
      "epoch": 4.832285115303983,
      "grad_norm": 1.707855224609375,
      "learning_rate": 4.700828998039937e-05,
      "loss": 0.345,
      "num_input_tokens_seen": 6020624,
      "step": 9220
    },
    {
      "epoch": 4.834905660377358,
      "grad_norm": 0.9267650842666626,
      "learning_rate": 4.700286372491612e-05,
      "loss": 0.3595,
      "num_input_tokens_seen": 6023312,
      "step": 9225
    },
    {
      "epoch": 4.837526205450734,
      "grad_norm": 1.8047680854797363,
      "learning_rate": 4.6997432866684816e-05,
      "loss": 0.3568,
      "num_input_tokens_seen": 6026032,
      "step": 9230
    },
    {
      "epoch": 4.840146750524109,
      "grad_norm": 1.2279425859451294,
      "learning_rate": 4.6991997406841504e-05,
      "loss": 0.3219,
      "num_input_tokens_seen": 6029872,
      "step": 9235
    },
    {
      "epoch": 4.8427672955974845,
      "grad_norm": 0.8013585209846497,
      "learning_rate": 4.698655734652324e-05,
      "loss": 0.3666,
      "num_input_tokens_seen": 6033232,
      "step": 9240
    },
    {
      "epoch": 4.84538784067086,
      "grad_norm": 2.0282247066497803,
      "learning_rate": 4.698111268686801e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 6035888,
      "step": 9245
    },
    {
      "epoch": 4.848008385744235,
      "grad_norm": 0.7770785093307495,
      "learning_rate": 4.6975663429014785e-05,
      "loss": 0.3631,
      "num_input_tokens_seen": 6039408,
      "step": 9250
    },
    {
      "epoch": 4.85062893081761,
      "grad_norm": 0.9023821949958801,
      "learning_rate": 4.697020957410347e-05,
      "loss": 0.4463,
      "num_input_tokens_seen": 6042736,
      "step": 9255
    },
    {
      "epoch": 4.853249475890985,
      "grad_norm": 1.0844001770019531,
      "learning_rate": 4.6964751123274974e-05,
      "loss": 0.3639,
      "num_input_tokens_seen": 6045296,
      "step": 9260
    },
    {
      "epoch": 4.85587002096436,
      "grad_norm": 1.9146453142166138,
      "learning_rate": 4.695928807767113e-05,
      "loss": 0.357,
      "num_input_tokens_seen": 6047920,
      "step": 9265
    },
    {
      "epoch": 4.8584905660377355,
      "grad_norm": 0.810806393623352,
      "learning_rate": 4.695382043843474e-05,
      "loss": 0.2745,
      "num_input_tokens_seen": 6052432,
      "step": 9270
    },
    {
      "epoch": 4.861111111111111,
      "grad_norm": 2.27162504196167,
      "learning_rate": 4.694834820670958e-05,
      "loss": 0.3252,
      "num_input_tokens_seen": 6055472,
      "step": 9275
    },
    {
      "epoch": 4.863731656184486,
      "grad_norm": 1.9364477396011353,
      "learning_rate": 4.694287138364037e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 6059184,
      "step": 9280
    },
    {
      "epoch": 4.866352201257862,
      "grad_norm": 1.3506932258605957,
      "learning_rate": 4.6937389970372815e-05,
      "loss": 0.4664,
      "num_input_tokens_seen": 6062192,
      "step": 9285
    },
    {
      "epoch": 4.868972746331237,
      "grad_norm": 1.3627891540527344,
      "learning_rate": 4.693190396805354e-05,
      "loss": 0.3279,
      "num_input_tokens_seen": 6065264,
      "step": 9290
    },
    {
      "epoch": 4.871593291404612,
      "grad_norm": 1.0170775651931763,
      "learning_rate": 4.692641337783017e-05,
      "loss": 0.4494,
      "num_input_tokens_seen": 6069008,
      "step": 9295
    },
    {
      "epoch": 4.8742138364779874,
      "grad_norm": 1.043174386024475,
      "learning_rate": 4.6920918200851274e-05,
      "loss": 0.3033,
      "num_input_tokens_seen": 6072624,
      "step": 9300
    },
    {
      "epoch": 4.876834381551363,
      "grad_norm": 0.9293798208236694,
      "learning_rate": 4.691541843826638e-05,
      "loss": 0.4537,
      "num_input_tokens_seen": 6075280,
      "step": 9305
    },
    {
      "epoch": 4.879454926624738,
      "grad_norm": 1.4659154415130615,
      "learning_rate": 4.690991409122597e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 6078448,
      "step": 9310
    },
    {
      "epoch": 4.882075471698113,
      "grad_norm": 1.3988903760910034,
      "learning_rate": 4.690440516088149e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 6081008,
      "step": 9315
    },
    {
      "epoch": 4.884696016771488,
      "grad_norm": 0.7942651510238647,
      "learning_rate": 4.689889164838536e-05,
      "loss": 0.3887,
      "num_input_tokens_seen": 6085296,
      "step": 9320
    },
    {
      "epoch": 4.887316561844864,
      "grad_norm": 0.9705697298049927,
      "learning_rate": 4.6893373554890925e-05,
      "loss": 0.403,
      "num_input_tokens_seen": 6087664,
      "step": 9325
    },
    {
      "epoch": 4.889937106918239,
      "grad_norm": 1.6465189456939697,
      "learning_rate": 4.688785088155251e-05,
      "loss": 0.3801,
      "num_input_tokens_seen": 6090832,
      "step": 9330
    },
    {
      "epoch": 4.8925576519916145,
      "grad_norm": 1.1254940032958984,
      "learning_rate": 4.688232362952541e-05,
      "loss": 0.4131,
      "num_input_tokens_seen": 6093616,
      "step": 9335
    },
    {
      "epoch": 4.89517819706499,
      "grad_norm": 1.430930495262146,
      "learning_rate": 4.687679179996585e-05,
      "loss": 0.2815,
      "num_input_tokens_seen": 6096592,
      "step": 9340
    },
    {
      "epoch": 4.897798742138365,
      "grad_norm": 0.7674494981765747,
      "learning_rate": 4.687125539403103e-05,
      "loss": 0.2589,
      "num_input_tokens_seen": 6100144,
      "step": 9345
    },
    {
      "epoch": 4.90041928721174,
      "grad_norm": 1.0551748275756836,
      "learning_rate": 4.68657144128791e-05,
      "loss": 0.4031,
      "num_input_tokens_seen": 6103536,
      "step": 9350
    },
    {
      "epoch": 4.903039832285115,
      "grad_norm": 0.9784752130508423,
      "learning_rate": 4.6860168857669175e-05,
      "loss": 0.3618,
      "num_input_tokens_seen": 6107408,
      "step": 9355
    },
    {
      "epoch": 4.90566037735849,
      "grad_norm": 1.7104347944259644,
      "learning_rate": 4.6854618729561315e-05,
      "loss": 0.3789,
      "num_input_tokens_seen": 6110480,
      "step": 9360
    },
    {
      "epoch": 4.9082809224318655,
      "grad_norm": 1.6398917436599731,
      "learning_rate": 4.684906402971655e-05,
      "loss": 0.2984,
      "num_input_tokens_seen": 6113488,
      "step": 9365
    },
    {
      "epoch": 4.910901467505241,
      "grad_norm": 1.4294259548187256,
      "learning_rate": 4.684350475929685e-05,
      "loss": 0.3795,
      "num_input_tokens_seen": 6117648,
      "step": 9370
    },
    {
      "epoch": 4.913522012578616,
      "grad_norm": 1.6461282968521118,
      "learning_rate": 4.683794091946517e-05,
      "loss": 0.372,
      "num_input_tokens_seen": 6120592,
      "step": 9375
    },
    {
      "epoch": 4.916142557651992,
      "grad_norm": 1.253264307975769,
      "learning_rate": 4.683237251138538e-05,
      "loss": 0.4053,
      "num_input_tokens_seen": 6124400,
      "step": 9380
    },
    {
      "epoch": 4.918763102725367,
      "grad_norm": 1.1984034776687622,
      "learning_rate": 4.6826799536222346e-05,
      "loss": 0.3844,
      "num_input_tokens_seen": 6128400,
      "step": 9385
    },
    {
      "epoch": 4.921383647798742,
      "grad_norm": 1.0357370376586914,
      "learning_rate": 4.6821221995141865e-05,
      "loss": 0.3179,
      "num_input_tokens_seen": 6131536,
      "step": 9390
    },
    {
      "epoch": 4.924004192872117,
      "grad_norm": 0.9693331718444824,
      "learning_rate": 4.681563988931068e-05,
      "loss": 0.2556,
      "num_input_tokens_seen": 6135280,
      "step": 9395
    },
    {
      "epoch": 4.926624737945493,
      "grad_norm": 1.9604767560958862,
      "learning_rate": 4.681005321989652e-05,
      "loss": 0.3626,
      "num_input_tokens_seen": 6137680,
      "step": 9400
    },
    {
      "epoch": 4.929245283018868,
      "grad_norm": 1.224919319152832,
      "learning_rate": 4.6804461988068055e-05,
      "loss": 0.5751,
      "num_input_tokens_seen": 6141264,
      "step": 9405
    },
    {
      "epoch": 4.931865828092243,
      "grad_norm": 1.116469383239746,
      "learning_rate": 4.6798866194994896e-05,
      "loss": 0.323,
      "num_input_tokens_seen": 6145104,
      "step": 9410
    },
    {
      "epoch": 4.934486373165618,
      "grad_norm": 0.9145447015762329,
      "learning_rate": 4.6793265841847624e-05,
      "loss": 0.3408,
      "num_input_tokens_seen": 6148976,
      "step": 9415
    },
    {
      "epoch": 4.937106918238994,
      "grad_norm": 1.651356816291809,
      "learning_rate": 4.6787660929797766e-05,
      "loss": 0.4685,
      "num_input_tokens_seen": 6152560,
      "step": 9420
    },
    {
      "epoch": 4.939727463312369,
      "grad_norm": 1.5456360578536987,
      "learning_rate": 4.678205146001781e-05,
      "loss": 0.3893,
      "num_input_tokens_seen": 6155024,
      "step": 9425
    },
    {
      "epoch": 4.9423480083857445,
      "grad_norm": 0.8498872518539429,
      "learning_rate": 4.67764374336812e-05,
      "loss": 0.3421,
      "num_input_tokens_seen": 6157584,
      "step": 9430
    },
    {
      "epoch": 4.94496855345912,
      "grad_norm": 1.332139492034912,
      "learning_rate": 4.6770818851962305e-05,
      "loss": 0.387,
      "num_input_tokens_seen": 6160240,
      "step": 9435
    },
    {
      "epoch": 4.947589098532495,
      "grad_norm": 0.9820585250854492,
      "learning_rate": 4.676519571603648e-05,
      "loss": 0.4112,
      "num_input_tokens_seen": 6163696,
      "step": 9440
    },
    {
      "epoch": 4.95020964360587,
      "grad_norm": 1.0771465301513672,
      "learning_rate": 4.6759568027080034e-05,
      "loss": 0.2873,
      "num_input_tokens_seen": 6168880,
      "step": 9445
    },
    {
      "epoch": 4.952830188679245,
      "grad_norm": 1.5757278203964233,
      "learning_rate": 4.67539357862702e-05,
      "loss": 0.3763,
      "num_input_tokens_seen": 6172208,
      "step": 9450
    },
    {
      "epoch": 4.95545073375262,
      "grad_norm": 1.047003149986267,
      "learning_rate": 4.6748298994785175e-05,
      "loss": 0.2955,
      "num_input_tokens_seen": 6174928,
      "step": 9455
    },
    {
      "epoch": 4.9580712788259955,
      "grad_norm": 1.1035757064819336,
      "learning_rate": 4.6742657653804116e-05,
      "loss": 0.388,
      "num_input_tokens_seen": 6178992,
      "step": 9460
    },
    {
      "epoch": 4.960691823899371,
      "grad_norm": 1.1731783151626587,
      "learning_rate": 4.673701176450713e-05,
      "loss": 0.3534,
      "num_input_tokens_seen": 6182128,
      "step": 9465
    },
    {
      "epoch": 4.963312368972746,
      "grad_norm": 1.629440426826477,
      "learning_rate": 4.6731361328075286e-05,
      "loss": 0.3122,
      "num_input_tokens_seen": 6185328,
      "step": 9470
    },
    {
      "epoch": 4.965932914046122,
      "grad_norm": 0.9141956567764282,
      "learning_rate": 4.672570634569056e-05,
      "loss": 0.3856,
      "num_input_tokens_seen": 6188560,
      "step": 9475
    },
    {
      "epoch": 4.968553459119497,
      "grad_norm": 0.7984622716903687,
      "learning_rate": 4.672004681853593e-05,
      "loss": 0.4483,
      "num_input_tokens_seen": 6193136,
      "step": 9480
    },
    {
      "epoch": 4.971174004192872,
      "grad_norm": 1.3357529640197754,
      "learning_rate": 4.6714382747795304e-05,
      "loss": 0.5353,
      "num_input_tokens_seen": 6196816,
      "step": 9485
    },
    {
      "epoch": 4.973794549266247,
      "grad_norm": 1.5262409448623657,
      "learning_rate": 4.670871413465354e-05,
      "loss": 0.3617,
      "num_input_tokens_seen": 6200272,
      "step": 9490
    },
    {
      "epoch": 4.976415094339623,
      "grad_norm": 2.3251006603240967,
      "learning_rate": 4.670304098029645e-05,
      "loss": 0.2916,
      "num_input_tokens_seen": 6203216,
      "step": 9495
    },
    {
      "epoch": 4.979035639412998,
      "grad_norm": 0.7728943228721619,
      "learning_rate": 4.669736328591078e-05,
      "loss": 0.2961,
      "num_input_tokens_seen": 6206736,
      "step": 9500
    },
    {
      "epoch": 4.981656184486373,
      "grad_norm": 0.9305855631828308,
      "learning_rate": 4.669168105268424e-05,
      "loss": 0.3203,
      "num_input_tokens_seen": 6209936,
      "step": 9505
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.9628187417984009,
      "learning_rate": 4.6685994281805514e-05,
      "loss": 0.3416,
      "num_input_tokens_seen": 6213424,
      "step": 9510
    },
    {
      "epoch": 4.986897274633124,
      "grad_norm": 1.5937732458114624,
      "learning_rate": 4.6680302974464175e-05,
      "loss": 0.3071,
      "num_input_tokens_seen": 6217360,
      "step": 9515
    },
    {
      "epoch": 4.989517819706499,
      "grad_norm": 2.0076582431793213,
      "learning_rate": 4.66746071318508e-05,
      "loss": 0.4092,
      "num_input_tokens_seen": 6219984,
      "step": 9520
    },
    {
      "epoch": 4.9921383647798745,
      "grad_norm": 0.8899292945861816,
      "learning_rate": 4.6668906755156896e-05,
      "loss": 0.2863,
      "num_input_tokens_seen": 6223664,
      "step": 9525
    },
    {
      "epoch": 4.99475890985325,
      "grad_norm": 0.9406707882881165,
      "learning_rate": 4.6663201845574906e-05,
      "loss": 0.3646,
      "num_input_tokens_seen": 6227440,
      "step": 9530
    },
    {
      "epoch": 4.997379454926625,
      "grad_norm": 1.2350692749023438,
      "learning_rate": 4.6657492404298234e-05,
      "loss": 0.2882,
      "num_input_tokens_seen": 6230608,
      "step": 9535
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.7570518255233765,
      "learning_rate": 4.665177843252124e-05,
      "loss": 0.406,
      "num_input_tokens_seen": 6233920,
      "step": 9540
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4919157326221466,
      "eval_runtime": 11.7031,
      "eval_samples_per_second": 72.459,
      "eval_steps_per_second": 18.115,
      "num_input_tokens_seen": 6233920,
      "step": 9540
    },
    {
      "epoch": 5.002620545073375,
      "grad_norm": 1.385807752609253,
      "learning_rate": 4.664605993143921e-05,
      "loss": 0.27,
      "num_input_tokens_seen": 6237312,
      "step": 9545
    },
    {
      "epoch": 5.00524109014675,
      "grad_norm": 1.0586283206939697,
      "learning_rate": 4.6640336902248396e-05,
      "loss": 0.2943,
      "num_input_tokens_seen": 6241344,
      "step": 9550
    },
    {
      "epoch": 5.0078616352201255,
      "grad_norm": 1.018068552017212,
      "learning_rate": 4.6634609346145984e-05,
      "loss": 0.372,
      "num_input_tokens_seen": 6244064,
      "step": 9555
    },
    {
      "epoch": 5.010482180293501,
      "grad_norm": 0.9581907987594604,
      "learning_rate": 4.6628877264330116e-05,
      "loss": 0.4397,
      "num_input_tokens_seen": 6247008,
      "step": 9560
    },
    {
      "epoch": 5.013102725366877,
      "grad_norm": 0.8538011908531189,
      "learning_rate": 4.6623140657999886e-05,
      "loss": 0.3941,
      "num_input_tokens_seen": 6250016,
      "step": 9565
    },
    {
      "epoch": 5.015723270440252,
      "grad_norm": 0.8962982296943665,
      "learning_rate": 4.661739952835532e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 6253504,
      "step": 9570
    },
    {
      "epoch": 5.018343815513627,
      "grad_norm": 0.7643075585365295,
      "learning_rate": 4.66116538765974e-05,
      "loss": 0.3131,
      "num_input_tokens_seen": 6256256,
      "step": 9575
    },
    {
      "epoch": 5.020964360587002,
      "grad_norm": 1.0345934629440308,
      "learning_rate": 4.6605903703928034e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 6260160,
      "step": 9580
    },
    {
      "epoch": 5.023584905660377,
      "grad_norm": 0.9320261478424072,
      "learning_rate": 4.6600149011550116e-05,
      "loss": 0.324,
      "num_input_tokens_seen": 6264032,
      "step": 9585
    },
    {
      "epoch": 5.026205450733753,
      "grad_norm": 0.7138766050338745,
      "learning_rate": 4.659438980066745e-05,
      "loss": 0.3344,
      "num_input_tokens_seen": 6267552,
      "step": 9590
    },
    {
      "epoch": 5.028825995807128,
      "grad_norm": 0.8427238464355469,
      "learning_rate": 4.65886260724848e-05,
      "loss": 0.3248,
      "num_input_tokens_seen": 6270272,
      "step": 9595
    },
    {
      "epoch": 5.031446540880503,
      "grad_norm": 1.0358812808990479,
      "learning_rate": 4.658285782820787e-05,
      "loss": 0.164,
      "num_input_tokens_seen": 6274624,
      "step": 9600
    },
    {
      "epoch": 5.034067085953878,
      "grad_norm": 1.11892831325531,
      "learning_rate": 4.657708506904331e-05,
      "loss": 0.3192,
      "num_input_tokens_seen": 6277728,
      "step": 9605
    },
    {
      "epoch": 5.036687631027253,
      "grad_norm": 1.998375654220581,
      "learning_rate": 4.6571307796198724e-05,
      "loss": 0.2789,
      "num_input_tokens_seen": 6281024,
      "step": 9610
    },
    {
      "epoch": 5.039308176100629,
      "grad_norm": 1.5428208112716675,
      "learning_rate": 4.656552601088265e-05,
      "loss": 0.3877,
      "num_input_tokens_seen": 6284576,
      "step": 9615
    },
    {
      "epoch": 5.0419287211740045,
      "grad_norm": 1.3602931499481201,
      "learning_rate": 4.655973971430456e-05,
      "loss": 0.3723,
      "num_input_tokens_seen": 6288096,
      "step": 9620
    },
    {
      "epoch": 5.04454926624738,
      "grad_norm": 1.5716016292572021,
      "learning_rate": 4.6553948907674895e-05,
      "loss": 0.3451,
      "num_input_tokens_seen": 6292128,
      "step": 9625
    },
    {
      "epoch": 5.047169811320755,
      "grad_norm": 1.4839264154434204,
      "learning_rate": 4.654815359220501e-05,
      "loss": 0.307,
      "num_input_tokens_seen": 6298912,
      "step": 9630
    },
    {
      "epoch": 5.04979035639413,
      "grad_norm": 1.5905890464782715,
      "learning_rate": 4.6542353769107226e-05,
      "loss": 0.3188,
      "num_input_tokens_seen": 6301536,
      "step": 9635
    },
    {
      "epoch": 5.052410901467505,
      "grad_norm": 1.1167922019958496,
      "learning_rate": 4.653654943959481e-05,
      "loss": 0.2761,
      "num_input_tokens_seen": 6303936,
      "step": 9640
    },
    {
      "epoch": 5.05503144654088,
      "grad_norm": 1.5528984069824219,
      "learning_rate": 4.653074060488195e-05,
      "loss": 0.2972,
      "num_input_tokens_seen": 6306880,
      "step": 9645
    },
    {
      "epoch": 5.0576519916142555,
      "grad_norm": 1.0916048288345337,
      "learning_rate": 4.6524927266183796e-05,
      "loss": 0.3,
      "num_input_tokens_seen": 6309920,
      "step": 9650
    },
    {
      "epoch": 5.060272536687631,
      "grad_norm": 1.0002533197402954,
      "learning_rate": 4.6519109424716426e-05,
      "loss": 0.2766,
      "num_input_tokens_seen": 6312480,
      "step": 9655
    },
    {
      "epoch": 5.062893081761007,
      "grad_norm": 0.9864354133605957,
      "learning_rate": 4.6513287081696855e-05,
      "loss": 0.2831,
      "num_input_tokens_seen": 6316256,
      "step": 9660
    },
    {
      "epoch": 5.065513626834382,
      "grad_norm": 1.7890654802322388,
      "learning_rate": 4.6507460238343076e-05,
      "loss": 0.3217,
      "num_input_tokens_seen": 6319520,
      "step": 9665
    },
    {
      "epoch": 5.068134171907757,
      "grad_norm": 1.0623159408569336,
      "learning_rate": 4.6501628895873984e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 6323328,
      "step": 9670
    },
    {
      "epoch": 5.070754716981132,
      "grad_norm": 1.0313074588775635,
      "learning_rate": 4.649579305550942e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 6326720,
      "step": 9675
    },
    {
      "epoch": 5.073375262054507,
      "grad_norm": 1.8335700035095215,
      "learning_rate": 4.64899527184702e-05,
      "loss": 0.334,
      "num_input_tokens_seen": 6329280,
      "step": 9680
    },
    {
      "epoch": 5.075995807127883,
      "grad_norm": 0.7582823634147644,
      "learning_rate": 4.6484107885978035e-05,
      "loss": 0.283,
      "num_input_tokens_seen": 6332384,
      "step": 9685
    },
    {
      "epoch": 5.078616352201258,
      "grad_norm": 1.1272215843200684,
      "learning_rate": 4.64782585592556e-05,
      "loss": 0.2464,
      "num_input_tokens_seen": 6336096,
      "step": 9690
    },
    {
      "epoch": 5.081236897274633,
      "grad_norm": 1.3872877359390259,
      "learning_rate": 4.647240473952652e-05,
      "loss": 0.3347,
      "num_input_tokens_seen": 6338432,
      "step": 9695
    },
    {
      "epoch": 5.083857442348008,
      "grad_norm": 0.9859622716903687,
      "learning_rate": 4.6466546428015336e-05,
      "loss": 0.2565,
      "num_input_tokens_seen": 6342016,
      "step": 9700
    },
    {
      "epoch": 5.086477987421383,
      "grad_norm": 1.6021440029144287,
      "learning_rate": 4.646068362594753e-05,
      "loss": 0.3348,
      "num_input_tokens_seen": 6344896,
      "step": 9705
    },
    {
      "epoch": 5.089098532494759,
      "grad_norm": 1.56919527053833,
      "learning_rate": 4.645481633454956e-05,
      "loss": 0.3056,
      "num_input_tokens_seen": 6348096,
      "step": 9710
    },
    {
      "epoch": 5.0917190775681345,
      "grad_norm": 1.5828124284744263,
      "learning_rate": 4.644894455504878e-05,
      "loss": 0.3949,
      "num_input_tokens_seen": 6351488,
      "step": 9715
    },
    {
      "epoch": 5.09433962264151,
      "grad_norm": 1.4286863803863525,
      "learning_rate": 4.64430682886735e-05,
      "loss": 0.3559,
      "num_input_tokens_seen": 6354464,
      "step": 9720
    },
    {
      "epoch": 5.096960167714885,
      "grad_norm": 0.996644139289856,
      "learning_rate": 4.643718753665297e-05,
      "loss": 0.3004,
      "num_input_tokens_seen": 6358912,
      "step": 9725
    },
    {
      "epoch": 5.09958071278826,
      "grad_norm": 2.4576470851898193,
      "learning_rate": 4.643130230021737e-05,
      "loss": 0.2573,
      "num_input_tokens_seen": 6362112,
      "step": 9730
    },
    {
      "epoch": 5.102201257861635,
      "grad_norm": 1.3871402740478516,
      "learning_rate": 4.6425412580597836e-05,
      "loss": 0.3051,
      "num_input_tokens_seen": 6364896,
      "step": 9735
    },
    {
      "epoch": 5.10482180293501,
      "grad_norm": 1.531683325767517,
      "learning_rate": 4.641951837902643e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 6368160,
      "step": 9740
    },
    {
      "epoch": 5.1074423480083855,
      "grad_norm": 1.3664476871490479,
      "learning_rate": 4.641361969673614e-05,
      "loss": 0.2429,
      "num_input_tokens_seen": 6371264,
      "step": 9745
    },
    {
      "epoch": 5.110062893081761,
      "grad_norm": 1.938157081604004,
      "learning_rate": 4.640771653496091e-05,
      "loss": 0.3834,
      "num_input_tokens_seen": 6374144,
      "step": 9750
    },
    {
      "epoch": 5.112683438155136,
      "grad_norm": 1.9458590745925903,
      "learning_rate": 4.64018088949356e-05,
      "loss": 0.322,
      "num_input_tokens_seen": 6377600,
      "step": 9755
    },
    {
      "epoch": 5.115303983228512,
      "grad_norm": 1.6032116413116455,
      "learning_rate": 4.6395896777896055e-05,
      "loss": 0.2916,
      "num_input_tokens_seen": 6380608,
      "step": 9760
    },
    {
      "epoch": 5.117924528301887,
      "grad_norm": 1.4853368997573853,
      "learning_rate": 4.638998018507899e-05,
      "loss": 0.4509,
      "num_input_tokens_seen": 6384640,
      "step": 9765
    },
    {
      "epoch": 5.120545073375262,
      "grad_norm": 4.421380043029785,
      "learning_rate": 4.63840591177221e-05,
      "loss": 0.2593,
      "num_input_tokens_seen": 6387456,
      "step": 9770
    },
    {
      "epoch": 5.123165618448637,
      "grad_norm": 1.401327133178711,
      "learning_rate": 4.637813357706401e-05,
      "loss": 0.2854,
      "num_input_tokens_seen": 6390880,
      "step": 9775
    },
    {
      "epoch": 5.1257861635220126,
      "grad_norm": 1.4158689975738525,
      "learning_rate": 4.6372203564344264e-05,
      "loss": 0.2481,
      "num_input_tokens_seen": 6393696,
      "step": 9780
    },
    {
      "epoch": 5.128406708595388,
      "grad_norm": 0.8331968188285828,
      "learning_rate": 4.6366269080803364e-05,
      "loss": 0.3013,
      "num_input_tokens_seen": 6397472,
      "step": 9785
    },
    {
      "epoch": 5.131027253668763,
      "grad_norm": 1.1530077457427979,
      "learning_rate": 4.636033012768273e-05,
      "loss": 0.3188,
      "num_input_tokens_seen": 6400576,
      "step": 9790
    },
    {
      "epoch": 5.133647798742138,
      "grad_norm": 1.4429740905761719,
      "learning_rate": 4.635438670622473e-05,
      "loss": 0.3943,
      "num_input_tokens_seen": 6403168,
      "step": 9795
    },
    {
      "epoch": 5.136268343815513,
      "grad_norm": 1.3301833868026733,
      "learning_rate": 4.6348438817672646e-05,
      "loss": 0.3139,
      "num_input_tokens_seen": 6406144,
      "step": 9800
    },
    {
      "epoch": 5.138888888888889,
      "grad_norm": 1.7313990592956543,
      "learning_rate": 4.634248646327072e-05,
      "loss": 0.327,
      "num_input_tokens_seen": 6409344,
      "step": 9805
    },
    {
      "epoch": 5.1415094339622645,
      "grad_norm": 1.7000540494918823,
      "learning_rate": 4.6336529644264114e-05,
      "loss": 0.3205,
      "num_input_tokens_seen": 6411776,
      "step": 9810
    },
    {
      "epoch": 5.14412997903564,
      "grad_norm": 1.0425705909729004,
      "learning_rate": 4.633056836189893e-05,
      "loss": 0.2738,
      "num_input_tokens_seen": 6414784,
      "step": 9815
    },
    {
      "epoch": 5.146750524109015,
      "grad_norm": 2.4268414974212646,
      "learning_rate": 4.632460261742219e-05,
      "loss": 0.2902,
      "num_input_tokens_seen": 6417248,
      "step": 9820
    },
    {
      "epoch": 5.14937106918239,
      "grad_norm": 1.3097305297851562,
      "learning_rate": 4.631863241208187e-05,
      "loss": 0.3909,
      "num_input_tokens_seen": 6420416,
      "step": 9825
    },
    {
      "epoch": 5.151991614255765,
      "grad_norm": 1.641663908958435,
      "learning_rate": 4.6312657747126865e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 6424288,
      "step": 9830
    },
    {
      "epoch": 5.15461215932914,
      "grad_norm": 0.9938166737556458,
      "learning_rate": 4.630667862380701e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 6426784,
      "step": 9835
    },
    {
      "epoch": 5.1572327044025155,
      "grad_norm": 1.3534471988677979,
      "learning_rate": 4.630069504337306e-05,
      "loss": 0.3193,
      "num_input_tokens_seen": 6429760,
      "step": 9840
    },
    {
      "epoch": 5.159853249475891,
      "grad_norm": 1.5023486614227295,
      "learning_rate": 4.629470700707672e-05,
      "loss": 0.2744,
      "num_input_tokens_seen": 6432480,
      "step": 9845
    },
    {
      "epoch": 5.162473794549266,
      "grad_norm": 1.2144652605056763,
      "learning_rate": 4.628871451617062e-05,
      "loss": 0.3321,
      "num_input_tokens_seen": 6437312,
      "step": 9850
    },
    {
      "epoch": 5.165094339622642,
      "grad_norm": 1.2672055959701538,
      "learning_rate": 4.628271757190832e-05,
      "loss": 0.384,
      "num_input_tokens_seen": 6441664,
      "step": 9855
    },
    {
      "epoch": 5.167714884696017,
      "grad_norm": 0.6554237604141235,
      "learning_rate": 4.62767161755443e-05,
      "loss": 0.2653,
      "num_input_tokens_seen": 6445120,
      "step": 9860
    },
    {
      "epoch": 5.170335429769392,
      "grad_norm": 1.6566194295883179,
      "learning_rate": 4.6270710328334004e-05,
      "loss": 0.3262,
      "num_input_tokens_seen": 6447968,
      "step": 9865
    },
    {
      "epoch": 5.172955974842767,
      "grad_norm": 1.2746278047561646,
      "learning_rate": 4.626470003153377e-05,
      "loss": 0.2896,
      "num_input_tokens_seen": 6450592,
      "step": 9870
    },
    {
      "epoch": 5.1755765199161425,
      "grad_norm": 2.0372774600982666,
      "learning_rate": 4.625868528640089e-05,
      "loss": 0.296,
      "num_input_tokens_seen": 6454496,
      "step": 9875
    },
    {
      "epoch": 5.178197064989518,
      "grad_norm": 1.4946799278259277,
      "learning_rate": 4.6252666094193586e-05,
      "loss": 0.2297,
      "num_input_tokens_seen": 6457088,
      "step": 9880
    },
    {
      "epoch": 5.180817610062893,
      "grad_norm": 1.4579551219940186,
      "learning_rate": 4.6246642456171e-05,
      "loss": 0.3669,
      "num_input_tokens_seen": 6461088,
      "step": 9885
    },
    {
      "epoch": 5.183438155136268,
      "grad_norm": 1.0127812623977661,
      "learning_rate": 4.6240614373593196e-05,
      "loss": 0.3565,
      "num_input_tokens_seen": 6463872,
      "step": 9890
    },
    {
      "epoch": 5.186058700209643,
      "grad_norm": 1.2796547412872314,
      "learning_rate": 4.623458184772119e-05,
      "loss": 0.3355,
      "num_input_tokens_seen": 6467424,
      "step": 9895
    },
    {
      "epoch": 5.188679245283019,
      "grad_norm": 0.8867960572242737,
      "learning_rate": 4.622854487981693e-05,
      "loss": 0.2888,
      "num_input_tokens_seen": 6470688,
      "step": 9900
    },
    {
      "epoch": 5.191299790356394,
      "grad_norm": 0.8327339291572571,
      "learning_rate": 4.622250347114326e-05,
      "loss": 0.3018,
      "num_input_tokens_seen": 6473440,
      "step": 9905
    },
    {
      "epoch": 5.19392033542977,
      "grad_norm": 2.15572190284729,
      "learning_rate": 4.621645762296399e-05,
      "loss": 0.3653,
      "num_input_tokens_seen": 6476352,
      "step": 9910
    },
    {
      "epoch": 5.196540880503145,
      "grad_norm": 1.0485261678695679,
      "learning_rate": 4.6210407336543835e-05,
      "loss": 0.2625,
      "num_input_tokens_seen": 6480064,
      "step": 9915
    },
    {
      "epoch": 5.19916142557652,
      "grad_norm": 1.5806711912155151,
      "learning_rate": 4.620435261314844e-05,
      "loss": 0.2812,
      "num_input_tokens_seen": 6483840,
      "step": 9920
    },
    {
      "epoch": 5.201781970649895,
      "grad_norm": 2.1232848167419434,
      "learning_rate": 4.61982934540444e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 6486368,
      "step": 9925
    },
    {
      "epoch": 5.20440251572327,
      "grad_norm": 1.0768465995788574,
      "learning_rate": 4.6192229860499206e-05,
      "loss": 0.2906,
      "num_input_tokens_seen": 6489152,
      "step": 9930
    },
    {
      "epoch": 5.2070230607966455,
      "grad_norm": 1.0690789222717285,
      "learning_rate": 4.6186161833781295e-05,
      "loss": 0.2863,
      "num_input_tokens_seen": 6493088,
      "step": 9935
    },
    {
      "epoch": 5.209643605870021,
      "grad_norm": 0.7495095729827881,
      "learning_rate": 4.618008937516003e-05,
      "loss": 0.3073,
      "num_input_tokens_seen": 6496256,
      "step": 9940
    },
    {
      "epoch": 5.212264150943396,
      "grad_norm": 1.4127461910247803,
      "learning_rate": 4.617401248590572e-05,
      "loss": 0.2673,
      "num_input_tokens_seen": 6499296,
      "step": 9945
    },
    {
      "epoch": 5.214884696016772,
      "grad_norm": 1.2880098819732666,
      "learning_rate": 4.616793116728955e-05,
      "loss": 0.2805,
      "num_input_tokens_seen": 6502624,
      "step": 9950
    },
    {
      "epoch": 5.217505241090147,
      "grad_norm": 0.9584534168243408,
      "learning_rate": 4.6161845420583685e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 6506272,
      "step": 9955
    },
    {
      "epoch": 5.220125786163522,
      "grad_norm": 1.731711506843567,
      "learning_rate": 4.6155755247061174e-05,
      "loss": 0.312,
      "num_input_tokens_seen": 6509504,
      "step": 9960
    },
    {
      "epoch": 5.222746331236897,
      "grad_norm": 2.4987285137176514,
      "learning_rate": 4.614966064799603e-05,
      "loss": 0.3527,
      "num_input_tokens_seen": 6512512,
      "step": 9965
    },
    {
      "epoch": 5.2253668763102725,
      "grad_norm": 0.6754045486450195,
      "learning_rate": 4.6143561624663166e-05,
      "loss": 0.3546,
      "num_input_tokens_seen": 6515456,
      "step": 9970
    },
    {
      "epoch": 5.227987421383648,
      "grad_norm": 1.4157376289367676,
      "learning_rate": 4.6137458178338425e-05,
      "loss": 0.2624,
      "num_input_tokens_seen": 6518880,
      "step": 9975
    },
    {
      "epoch": 5.230607966457023,
      "grad_norm": 1.206865668296814,
      "learning_rate": 4.613135031029859e-05,
      "loss": 0.2955,
      "num_input_tokens_seen": 6521664,
      "step": 9980
    },
    {
      "epoch": 5.233228511530398,
      "grad_norm": 2.6114070415496826,
      "learning_rate": 4.612523802182134e-05,
      "loss": 0.299,
      "num_input_tokens_seen": 6524160,
      "step": 9985
    },
    {
      "epoch": 5.235849056603773,
      "grad_norm": 1.5182744264602661,
      "learning_rate": 4.611912131418531e-05,
      "loss": 0.2255,
      "num_input_tokens_seen": 6527904,
      "step": 9990
    },
    {
      "epoch": 5.238469601677149,
      "grad_norm": 3.278249740600586,
      "learning_rate": 4.611300018867004e-05,
      "loss": 0.4939,
      "num_input_tokens_seen": 6530560,
      "step": 9995
    },
    {
      "epoch": 5.241090146750524,
      "grad_norm": 1.5374763011932373,
      "learning_rate": 4.6106874646556e-05,
      "loss": 0.3324,
      "num_input_tokens_seen": 6534048,
      "step": 10000
    },
    {
      "epoch": 5.2437106918239,
      "grad_norm": 1.2771625518798828,
      "learning_rate": 4.6100744689124584e-05,
      "loss": 0.286,
      "num_input_tokens_seen": 6536864,
      "step": 10005
    },
    {
      "epoch": 5.246331236897275,
      "grad_norm": 1.8161380290985107,
      "learning_rate": 4.6094610317658116e-05,
      "loss": 0.3708,
      "num_input_tokens_seen": 6540608,
      "step": 10010
    },
    {
      "epoch": 5.24895178197065,
      "grad_norm": 1.72498619556427,
      "learning_rate": 4.608847153343982e-05,
      "loss": 0.3579,
      "num_input_tokens_seen": 6543744,
      "step": 10015
    },
    {
      "epoch": 5.251572327044025,
      "grad_norm": 1.3513038158416748,
      "learning_rate": 4.6082328337753874e-05,
      "loss": 0.2994,
      "num_input_tokens_seen": 6546656,
      "step": 10020
    },
    {
      "epoch": 5.2541928721174,
      "grad_norm": 1.7368892431259155,
      "learning_rate": 4.6076180731885353e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 6549152,
      "step": 10025
    },
    {
      "epoch": 5.256813417190775,
      "grad_norm": 1.6126761436462402,
      "learning_rate": 4.6070028717120286e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 6552096,
      "step": 10030
    },
    {
      "epoch": 5.259433962264151,
      "grad_norm": 1.3238505125045776,
      "learning_rate": 4.606387229474558e-05,
      "loss": 0.3604,
      "num_input_tokens_seen": 6555232,
      "step": 10035
    },
    {
      "epoch": 5.262054507337526,
      "grad_norm": 2.9061455726623535,
      "learning_rate": 4.60577114660491e-05,
      "loss": 0.2832,
      "num_input_tokens_seen": 6557920,
      "step": 10040
    },
    {
      "epoch": 5.264675052410902,
      "grad_norm": 1.9545260667800903,
      "learning_rate": 4.605154623231962e-05,
      "loss": 0.2785,
      "num_input_tokens_seen": 6560928,
      "step": 10045
    },
    {
      "epoch": 5.267295597484277,
      "grad_norm": 1.5929979085922241,
      "learning_rate": 4.604537659484684e-05,
      "loss": 0.2771,
      "num_input_tokens_seen": 6563488,
      "step": 10050
    },
    {
      "epoch": 5.269916142557652,
      "grad_norm": 1.0995901823043823,
      "learning_rate": 4.603920255492138e-05,
      "loss": 0.3064,
      "num_input_tokens_seen": 6567520,
      "step": 10055
    },
    {
      "epoch": 5.272536687631027,
      "grad_norm": 0.9670811891555786,
      "learning_rate": 4.603302411383477e-05,
      "loss": 0.3483,
      "num_input_tokens_seen": 6570528,
      "step": 10060
    },
    {
      "epoch": 5.2751572327044025,
      "grad_norm": 1.14784574508667,
      "learning_rate": 4.602684127287946e-05,
      "loss": 0.2611,
      "num_input_tokens_seen": 6573824,
      "step": 10065
    },
    {
      "epoch": 5.277777777777778,
      "grad_norm": 1.3626830577850342,
      "learning_rate": 4.602065403334885e-05,
      "loss": 0.3496,
      "num_input_tokens_seen": 6576896,
      "step": 10070
    },
    {
      "epoch": 5.280398322851153,
      "grad_norm": 3.1574761867523193,
      "learning_rate": 4.601446239653724e-05,
      "loss": 0.3514,
      "num_input_tokens_seen": 6579136,
      "step": 10075
    },
    {
      "epoch": 5.283018867924528,
      "grad_norm": 1.2674216032028198,
      "learning_rate": 4.600826636373984e-05,
      "loss": 0.2666,
      "num_input_tokens_seen": 6582080,
      "step": 10080
    },
    {
      "epoch": 5.285639412997903,
      "grad_norm": 1.4100027084350586,
      "learning_rate": 4.600206593625279e-05,
      "loss": 0.2866,
      "num_input_tokens_seen": 6584896,
      "step": 10085
    },
    {
      "epoch": 5.288259958071279,
      "grad_norm": 2.649024724960327,
      "learning_rate": 4.599586111537315e-05,
      "loss": 0.3253,
      "num_input_tokens_seen": 6588448,
      "step": 10090
    },
    {
      "epoch": 5.290880503144654,
      "grad_norm": 1.3653794527053833,
      "learning_rate": 4.59896519023989e-05,
      "loss": 0.2979,
      "num_input_tokens_seen": 6592704,
      "step": 10095
    },
    {
      "epoch": 5.29350104821803,
      "grad_norm": 2.2967264652252197,
      "learning_rate": 4.5983438298628935e-05,
      "loss": 0.3082,
      "num_input_tokens_seen": 6596256,
      "step": 10100
    },
    {
      "epoch": 5.296121593291405,
      "grad_norm": 1.2812864780426025,
      "learning_rate": 4.597722030536306e-05,
      "loss": 0.3502,
      "num_input_tokens_seen": 6598656,
      "step": 10105
    },
    {
      "epoch": 5.29874213836478,
      "grad_norm": 1.267958164215088,
      "learning_rate": 4.597099792390202e-05,
      "loss": 0.3322,
      "num_input_tokens_seen": 6601504,
      "step": 10110
    },
    {
      "epoch": 5.301362683438155,
      "grad_norm": 2.4023923873901367,
      "learning_rate": 4.596477115554748e-05,
      "loss": 0.3126,
      "num_input_tokens_seen": 6604640,
      "step": 10115
    },
    {
      "epoch": 5.30398322851153,
      "grad_norm": 1.050186276435852,
      "learning_rate": 4.5958540001601976e-05,
      "loss": 0.3387,
      "num_input_tokens_seen": 6607520,
      "step": 10120
    },
    {
      "epoch": 5.306603773584905,
      "grad_norm": 0.7423079609870911,
      "learning_rate": 4.595230446336902e-05,
      "loss": 0.2452,
      "num_input_tokens_seen": 6611360,
      "step": 10125
    },
    {
      "epoch": 5.309224318658281,
      "grad_norm": 1.0955028533935547,
      "learning_rate": 4.5946064542153005e-05,
      "loss": 0.2351,
      "num_input_tokens_seen": 6616672,
      "step": 10130
    },
    {
      "epoch": 5.311844863731656,
      "grad_norm": 1.158868432044983,
      "learning_rate": 4.593982023925926e-05,
      "loss": 0.4065,
      "num_input_tokens_seen": 6619424,
      "step": 10135
    },
    {
      "epoch": 5.314465408805032,
      "grad_norm": 1.3217010498046875,
      "learning_rate": 4.5933571555994e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 6622656,
      "step": 10140
    },
    {
      "epoch": 5.317085953878407,
      "grad_norm": 1.5377986431121826,
      "learning_rate": 4.59273184936644e-05,
      "loss": 0.2537,
      "num_input_tokens_seen": 6625568,
      "step": 10145
    },
    {
      "epoch": 5.319706498951782,
      "grad_norm": 1.9239174127578735,
      "learning_rate": 4.5921061053578523e-05,
      "loss": 0.4434,
      "num_input_tokens_seen": 6628896,
      "step": 10150
    },
    {
      "epoch": 5.322327044025157,
      "grad_norm": 1.5552417039871216,
      "learning_rate": 4.5914799237045355e-05,
      "loss": 0.2702,
      "num_input_tokens_seen": 6632672,
      "step": 10155
    },
    {
      "epoch": 5.3249475890985325,
      "grad_norm": 1.9592419862747192,
      "learning_rate": 4.5908533045374796e-05,
      "loss": 0.2793,
      "num_input_tokens_seen": 6637344,
      "step": 10160
    },
    {
      "epoch": 5.327568134171908,
      "grad_norm": 1.524820327758789,
      "learning_rate": 4.590226247987766e-05,
      "loss": 0.3946,
      "num_input_tokens_seen": 6640416,
      "step": 10165
    },
    {
      "epoch": 5.330188679245283,
      "grad_norm": 1.4068013429641724,
      "learning_rate": 4.589598754186568e-05,
      "loss": 0.2387,
      "num_input_tokens_seen": 6642688,
      "step": 10170
    },
    {
      "epoch": 5.332809224318658,
      "grad_norm": 2.1330063343048096,
      "learning_rate": 4.5889708232651494e-05,
      "loss": 0.3072,
      "num_input_tokens_seen": 6645952,
      "step": 10175
    },
    {
      "epoch": 5.335429769392033,
      "grad_norm": 1.317192792892456,
      "learning_rate": 4.588342455354867e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 6648576,
      "step": 10180
    },
    {
      "epoch": 5.338050314465409,
      "grad_norm": 2.044236660003662,
      "learning_rate": 4.5877136505871685e-05,
      "loss": 0.3545,
      "num_input_tokens_seen": 6651840,
      "step": 10185
    },
    {
      "epoch": 5.340670859538784,
      "grad_norm": 2.0225069522857666,
      "learning_rate": 4.5870844090935914e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 6654336,
      "step": 10190
    },
    {
      "epoch": 5.34329140461216,
      "grad_norm": 2.347599983215332,
      "learning_rate": 4.586454731005768e-05,
      "loss": 0.3602,
      "num_input_tokens_seen": 6657600,
      "step": 10195
    },
    {
      "epoch": 5.345911949685535,
      "grad_norm": 1.1173677444458008,
      "learning_rate": 4.5858246164554175e-05,
      "loss": 0.2688,
      "num_input_tokens_seen": 6660768,
      "step": 10200
    },
    {
      "epoch": 5.34853249475891,
      "grad_norm": 2.242076873779297,
      "learning_rate": 4.5851940655743536e-05,
      "loss": 0.3306,
      "num_input_tokens_seen": 6663808,
      "step": 10205
    },
    {
      "epoch": 5.351153039832285,
      "grad_norm": 0.7590295672416687,
      "learning_rate": 4.5845630784944804e-05,
      "loss": 0.3507,
      "num_input_tokens_seen": 6666048,
      "step": 10210
    },
    {
      "epoch": 5.35377358490566,
      "grad_norm": 1.1367696523666382,
      "learning_rate": 4.583931655347793e-05,
      "loss": 0.2777,
      "num_input_tokens_seen": 6668672,
      "step": 10215
    },
    {
      "epoch": 5.356394129979035,
      "grad_norm": 1.5663371086120605,
      "learning_rate": 4.5832997962663785e-05,
      "loss": 0.4198,
      "num_input_tokens_seen": 6671680,
      "step": 10220
    },
    {
      "epoch": 5.359014675052411,
      "grad_norm": 1.0192056894302368,
      "learning_rate": 4.582667501382414e-05,
      "loss": 0.2697,
      "num_input_tokens_seen": 6674592,
      "step": 10225
    },
    {
      "epoch": 5.361635220125786,
      "grad_norm": 1.479956030845642,
      "learning_rate": 4.582034770828169e-05,
      "loss": 0.2574,
      "num_input_tokens_seen": 6677632,
      "step": 10230
    },
    {
      "epoch": 5.364255765199162,
      "grad_norm": 1.071624994277954,
      "learning_rate": 4.581401604736002e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 6680768,
      "step": 10235
    },
    {
      "epoch": 5.366876310272537,
      "grad_norm": 1.708269715309143,
      "learning_rate": 4.580768003238365e-05,
      "loss": 0.329,
      "num_input_tokens_seen": 6683712,
      "step": 10240
    },
    {
      "epoch": 5.369496855345912,
      "grad_norm": 1.349091649055481,
      "learning_rate": 4.5801339664678014e-05,
      "loss": 0.3536,
      "num_input_tokens_seen": 6688512,
      "step": 10245
    },
    {
      "epoch": 5.372117400419287,
      "grad_norm": 1.2149250507354736,
      "learning_rate": 4.579499494556943e-05,
      "loss": 0.2744,
      "num_input_tokens_seen": 6691520,
      "step": 10250
    },
    {
      "epoch": 5.3747379454926625,
      "grad_norm": 0.8915312886238098,
      "learning_rate": 4.5788645876385147e-05,
      "loss": 0.3418,
      "num_input_tokens_seen": 6695360,
      "step": 10255
    },
    {
      "epoch": 5.377358490566038,
      "grad_norm": 1.0855391025543213,
      "learning_rate": 4.578229245845331e-05,
      "loss": 0.2932,
      "num_input_tokens_seen": 6698688,
      "step": 10260
    },
    {
      "epoch": 5.379979035639413,
      "grad_norm": 1.480883240699768,
      "learning_rate": 4.5775934693102994e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 6701312,
      "step": 10265
    },
    {
      "epoch": 5.382599580712788,
      "grad_norm": 1.6831969022750854,
      "learning_rate": 4.5769572581664165e-05,
      "loss": 0.326,
      "num_input_tokens_seen": 6703840,
      "step": 10270
    },
    {
      "epoch": 5.385220125786163,
      "grad_norm": 1.9395692348480225,
      "learning_rate": 4.57632061254677e-05,
      "loss": 0.2874,
      "num_input_tokens_seen": 6706720,
      "step": 10275
    },
    {
      "epoch": 5.387840670859539,
      "grad_norm": 1.4942272901535034,
      "learning_rate": 4.5756835325845394e-05,
      "loss": 0.2894,
      "num_input_tokens_seen": 6710240,
      "step": 10280
    },
    {
      "epoch": 5.390461215932914,
      "grad_norm": 1.3892897367477417,
      "learning_rate": 4.575046018412995e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 6713632,
      "step": 10285
    },
    {
      "epoch": 5.3930817610062896,
      "grad_norm": 1.3189016580581665,
      "learning_rate": 4.5744080701654955e-05,
      "loss": 0.2871,
      "num_input_tokens_seen": 6716032,
      "step": 10290
    },
    {
      "epoch": 5.395702306079665,
      "grad_norm": 1.5492868423461914,
      "learning_rate": 4.573769687975495e-05,
      "loss": 0.2989,
      "num_input_tokens_seen": 6719136,
      "step": 10295
    },
    {
      "epoch": 5.39832285115304,
      "grad_norm": 2.701425552368164,
      "learning_rate": 4.573130871976534e-05,
      "loss": 0.3077,
      "num_input_tokens_seen": 6721824,
      "step": 10300
    },
    {
      "epoch": 5.400943396226415,
      "grad_norm": 1.0957727432250977,
      "learning_rate": 4.572491622302247e-05,
      "loss": 0.2882,
      "num_input_tokens_seen": 6724864,
      "step": 10305
    },
    {
      "epoch": 5.40356394129979,
      "grad_norm": 1.723165512084961,
      "learning_rate": 4.571851939086356e-05,
      "loss": 0.237,
      "num_input_tokens_seen": 6728064,
      "step": 10310
    },
    {
      "epoch": 5.406184486373165,
      "grad_norm": 1.4828764200210571,
      "learning_rate": 4.571211822462676e-05,
      "loss": 0.4129,
      "num_input_tokens_seen": 6730784,
      "step": 10315
    },
    {
      "epoch": 5.408805031446541,
      "grad_norm": 1.2235827445983887,
      "learning_rate": 4.570571272565114e-05,
      "loss": 0.383,
      "num_input_tokens_seen": 6733408,
      "step": 10320
    },
    {
      "epoch": 5.411425576519916,
      "grad_norm": 0.20223717391490936,
      "learning_rate": 4.569930289527663e-05,
      "loss": 0.329,
      "num_input_tokens_seen": 6738912,
      "step": 10325
    },
    {
      "epoch": 5.414046121593292,
      "grad_norm": 1.399361252784729,
      "learning_rate": 4.56928887348441e-05,
      "loss": 0.342,
      "num_input_tokens_seen": 6742784,
      "step": 10330
    },
    {
      "epoch": 5.416666666666667,
      "grad_norm": 0.8684194684028625,
      "learning_rate": 4.568647024569533e-05,
      "loss": 0.2658,
      "num_input_tokens_seen": 6746080,
      "step": 10335
    },
    {
      "epoch": 5.419287211740042,
      "grad_norm": 1.1429988145828247,
      "learning_rate": 4.568004742917299e-05,
      "loss": 0.3821,
      "num_input_tokens_seen": 6750368,
      "step": 10340
    },
    {
      "epoch": 5.421907756813417,
      "grad_norm": 1.2250961065292358,
      "learning_rate": 4.567362028662065e-05,
      "loss": 0.3279,
      "num_input_tokens_seen": 6753984,
      "step": 10345
    },
    {
      "epoch": 5.4245283018867925,
      "grad_norm": 1.4050978422164917,
      "learning_rate": 4.566718881938281e-05,
      "loss": 0.3987,
      "num_input_tokens_seen": 6756576,
      "step": 10350
    },
    {
      "epoch": 5.427148846960168,
      "grad_norm": 1.4956833124160767,
      "learning_rate": 4.5660753028804846e-05,
      "loss": 0.2989,
      "num_input_tokens_seen": 6759904,
      "step": 10355
    },
    {
      "epoch": 5.429769392033543,
      "grad_norm": 1.4609862565994263,
      "learning_rate": 4.5654312916233066e-05,
      "loss": 0.2891,
      "num_input_tokens_seen": 6762880,
      "step": 10360
    },
    {
      "epoch": 5.432389937106918,
      "grad_norm": 1.8507559299468994,
      "learning_rate": 4.564786848301466e-05,
      "loss": 0.511,
      "num_input_tokens_seen": 6765728,
      "step": 10365
    },
    {
      "epoch": 5.435010482180293,
      "grad_norm": 3.3323512077331543,
      "learning_rate": 4.564141973049773e-05,
      "loss": 0.2766,
      "num_input_tokens_seen": 6768704,
      "step": 10370
    },
    {
      "epoch": 5.437631027253669,
      "grad_norm": 0.9558278918266296,
      "learning_rate": 4.563496666003128e-05,
      "loss": 0.2345,
      "num_input_tokens_seen": 6772256,
      "step": 10375
    },
    {
      "epoch": 5.440251572327044,
      "grad_norm": 1.3414092063903809,
      "learning_rate": 4.562850927296523e-05,
      "loss": 0.3491,
      "num_input_tokens_seen": 6775552,
      "step": 10380
    },
    {
      "epoch": 5.4428721174004195,
      "grad_norm": 1.3864198923110962,
      "learning_rate": 4.5622047570650365e-05,
      "loss": 0.2695,
      "num_input_tokens_seen": 6778176,
      "step": 10385
    },
    {
      "epoch": 5.445492662473795,
      "grad_norm": 1.696243166923523,
      "learning_rate": 4.561558155443843e-05,
      "loss": 0.3208,
      "num_input_tokens_seen": 6781888,
      "step": 10390
    },
    {
      "epoch": 5.44811320754717,
      "grad_norm": 1.3520996570587158,
      "learning_rate": 4.560911122568202e-05,
      "loss": 0.3267,
      "num_input_tokens_seen": 6785056,
      "step": 10395
    },
    {
      "epoch": 5.450733752620545,
      "grad_norm": 1.3399146795272827,
      "learning_rate": 4.560263658573467e-05,
      "loss": 0.2994,
      "num_input_tokens_seen": 6788832,
      "step": 10400
    },
    {
      "epoch": 5.45335429769392,
      "grad_norm": 1.387075424194336,
      "learning_rate": 4.559615763595079e-05,
      "loss": 0.3306,
      "num_input_tokens_seen": 6792320,
      "step": 10405
    },
    {
      "epoch": 5.455974842767295,
      "grad_norm": 0.6902546882629395,
      "learning_rate": 4.558967437768571e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 6796000,
      "step": 10410
    },
    {
      "epoch": 5.4585953878406706,
      "grad_norm": 1.7374396324157715,
      "learning_rate": 4.558318681229565e-05,
      "loss": 0.2899,
      "num_input_tokens_seen": 6798400,
      "step": 10415
    },
    {
      "epoch": 5.461215932914046,
      "grad_norm": 2.6051948070526123,
      "learning_rate": 4.557669494113773e-05,
      "loss": 0.3531,
      "num_input_tokens_seen": 6801440,
      "step": 10420
    },
    {
      "epoch": 5.463836477987422,
      "grad_norm": 1.5368295907974243,
      "learning_rate": 4.557019876556998e-05,
      "loss": 0.2964,
      "num_input_tokens_seen": 6807072,
      "step": 10425
    },
    {
      "epoch": 5.466457023060797,
      "grad_norm": 1.2497780323028564,
      "learning_rate": 4.556369828695133e-05,
      "loss": 0.3032,
      "num_input_tokens_seen": 6810272,
      "step": 10430
    },
    {
      "epoch": 5.469077568134172,
      "grad_norm": 1.2452223300933838,
      "learning_rate": 4.5557193506641596e-05,
      "loss": 0.2909,
      "num_input_tokens_seen": 6813184,
      "step": 10435
    },
    {
      "epoch": 5.471698113207547,
      "grad_norm": 2.009981393814087,
      "learning_rate": 4.555068442600151e-05,
      "loss": 0.3324,
      "num_input_tokens_seen": 6817824,
      "step": 10440
    },
    {
      "epoch": 5.4743186582809225,
      "grad_norm": 1.2866415977478027,
      "learning_rate": 4.5544171046392703e-05,
      "loss": 0.2839,
      "num_input_tokens_seen": 6820480,
      "step": 10445
    },
    {
      "epoch": 5.476939203354298,
      "grad_norm": 1.5993231534957886,
      "learning_rate": 4.5537653369177685e-05,
      "loss": 0.3665,
      "num_input_tokens_seen": 6823360,
      "step": 10450
    },
    {
      "epoch": 5.479559748427673,
      "grad_norm": 1.0126702785491943,
      "learning_rate": 4.5531131395719895e-05,
      "loss": 0.3656,
      "num_input_tokens_seen": 6825856,
      "step": 10455
    },
    {
      "epoch": 5.482180293501048,
      "grad_norm": 1.9300236701965332,
      "learning_rate": 4.552460512738364e-05,
      "loss": 0.3488,
      "num_input_tokens_seen": 6828896,
      "step": 10460
    },
    {
      "epoch": 5.484800838574423,
      "grad_norm": 1.0170472860336304,
      "learning_rate": 4.551807456553416e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 6832032,
      "step": 10465
    },
    {
      "epoch": 5.487421383647799,
      "grad_norm": 3.8572750091552734,
      "learning_rate": 4.551153971153755e-05,
      "loss": 0.3436,
      "num_input_tokens_seen": 6834624,
      "step": 10470
    },
    {
      "epoch": 5.490041928721174,
      "grad_norm": 1.2319518327713013,
      "learning_rate": 4.550500056676085e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 6837504,
      "step": 10475
    },
    {
      "epoch": 5.4926624737945495,
      "grad_norm": 2.899796962738037,
      "learning_rate": 4.5498457132571966e-05,
      "loss": 0.3191,
      "num_input_tokens_seen": 6840224,
      "step": 10480
    },
    {
      "epoch": 5.495283018867925,
      "grad_norm": 1.2912105321884155,
      "learning_rate": 4.5491909410339706e-05,
      "loss": 0.2182,
      "num_input_tokens_seen": 6843520,
      "step": 10485
    },
    {
      "epoch": 5.4979035639413,
      "grad_norm": 0.9285416007041931,
      "learning_rate": 4.548535740143378e-05,
      "loss": 0.3206,
      "num_input_tokens_seen": 6846464,
      "step": 10490
    },
    {
      "epoch": 5.500524109014675,
      "grad_norm": 1.4833143949508667,
      "learning_rate": 4.54788011072248e-05,
      "loss": 0.3657,
      "num_input_tokens_seen": 6849696,
      "step": 10495
    },
    {
      "epoch": 5.50314465408805,
      "grad_norm": 1.6149802207946777,
      "learning_rate": 4.5472240529084254e-05,
      "loss": 0.4195,
      "num_input_tokens_seen": 6853216,
      "step": 10500
    },
    {
      "epoch": 5.505765199161425,
      "grad_norm": 1.7987664937973022,
      "learning_rate": 4.5465675668384557e-05,
      "loss": 0.3271,
      "num_input_tokens_seen": 6855872,
      "step": 10505
    },
    {
      "epoch": 5.5083857442348005,
      "grad_norm": 1.6823583841323853,
      "learning_rate": 4.545910652649899e-05,
      "loss": 0.2631,
      "num_input_tokens_seen": 6858688,
      "step": 10510
    },
    {
      "epoch": 5.511006289308176,
      "grad_norm": 1.8087385892868042,
      "learning_rate": 4.5452533104801756e-05,
      "loss": 0.2985,
      "num_input_tokens_seen": 6861632,
      "step": 10515
    },
    {
      "epoch": 5.513626834381551,
      "grad_norm": 1.5068367719650269,
      "learning_rate": 4.544595540466792e-05,
      "loss": 0.453,
      "num_input_tokens_seen": 6865696,
      "step": 10520
    },
    {
      "epoch": 5.516247379454927,
      "grad_norm": 1.4842867851257324,
      "learning_rate": 4.5439373427473476e-05,
      "loss": 0.3037,
      "num_input_tokens_seen": 6869056,
      "step": 10525
    },
    {
      "epoch": 5.518867924528302,
      "grad_norm": 3.3756630420684814,
      "learning_rate": 4.54327871745953e-05,
      "loss": 0.3027,
      "num_input_tokens_seen": 6871488,
      "step": 10530
    },
    {
      "epoch": 5.521488469601677,
      "grad_norm": 0.9504592418670654,
      "learning_rate": 4.542619664741116e-05,
      "loss": 0.3541,
      "num_input_tokens_seen": 6875008,
      "step": 10535
    },
    {
      "epoch": 5.524109014675052,
      "grad_norm": 2.845592975616455,
      "learning_rate": 4.5419601847299706e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 6878336,
      "step": 10540
    },
    {
      "epoch": 5.526729559748428,
      "grad_norm": 1.1685763597488403,
      "learning_rate": 4.5413002775640503e-05,
      "loss": 0.306,
      "num_input_tokens_seen": 6882080,
      "step": 10545
    },
    {
      "epoch": 5.529350104821803,
      "grad_norm": 1.0016063451766968,
      "learning_rate": 4.540639943381401e-05,
      "loss": 0.2589,
      "num_input_tokens_seen": 6885504,
      "step": 10550
    },
    {
      "epoch": 5.531970649895178,
      "grad_norm": 1.0826001167297363,
      "learning_rate": 4.539979182320157e-05,
      "loss": 0.3609,
      "num_input_tokens_seen": 6888704,
      "step": 10555
    },
    {
      "epoch": 5.534591194968553,
      "grad_norm": 1.3686978816986084,
      "learning_rate": 4.5393179945185404e-05,
      "loss": 0.3703,
      "num_input_tokens_seen": 6891936,
      "step": 10560
    },
    {
      "epoch": 5.537211740041929,
      "grad_norm": 1.7084565162658691,
      "learning_rate": 4.538656380114865e-05,
      "loss": 0.3742,
      "num_input_tokens_seen": 6894848,
      "step": 10565
    },
    {
      "epoch": 5.539832285115304,
      "grad_norm": 1.509387731552124,
      "learning_rate": 4.537994339247533e-05,
      "loss": 0.3052,
      "num_input_tokens_seen": 6897728,
      "step": 10570
    },
    {
      "epoch": 5.5424528301886795,
      "grad_norm": 1.4295278787612915,
      "learning_rate": 4.5373318720550366e-05,
      "loss": 0.2512,
      "num_input_tokens_seen": 6900672,
      "step": 10575
    },
    {
      "epoch": 5.545073375262055,
      "grad_norm": 1.3994814157485962,
      "learning_rate": 4.536668978675956e-05,
      "loss": 0.2936,
      "num_input_tokens_seen": 6905920,
      "step": 10580
    },
    {
      "epoch": 5.54769392033543,
      "grad_norm": 0.7818044424057007,
      "learning_rate": 4.53600565924896e-05,
      "loss": 0.329,
      "num_input_tokens_seen": 6909408,
      "step": 10585
    },
    {
      "epoch": 5.550314465408805,
      "grad_norm": 1.3802067041397095,
      "learning_rate": 4.535341913912808e-05,
      "loss": 0.2839,
      "num_input_tokens_seen": 6912128,
      "step": 10590
    },
    {
      "epoch": 5.55293501048218,
      "grad_norm": 1.6312711238861084,
      "learning_rate": 4.534677742806348e-05,
      "loss": 0.3664,
      "num_input_tokens_seen": 6915136,
      "step": 10595
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 1.5072205066680908,
      "learning_rate": 4.5340131460685175e-05,
      "loss": 0.3314,
      "num_input_tokens_seen": 6917856,
      "step": 10600
    },
    {
      "epoch": 5.5581761006289305,
      "grad_norm": 1.3351144790649414,
      "learning_rate": 4.533348123838341e-05,
      "loss": 0.3054,
      "num_input_tokens_seen": 6920864,
      "step": 10605
    },
    {
      "epoch": 5.560796645702306,
      "grad_norm": 1.311333179473877,
      "learning_rate": 4.532682676254936e-05,
      "loss": 0.4662,
      "num_input_tokens_seen": 6924384,
      "step": 10610
    },
    {
      "epoch": 5.563417190775681,
      "grad_norm": 1.7555687427520752,
      "learning_rate": 4.5320168034575044e-05,
      "loss": 0.3365,
      "num_input_tokens_seen": 6926848,
      "step": 10615
    },
    {
      "epoch": 5.566037735849057,
      "grad_norm": 1.0530282258987427,
      "learning_rate": 4.53135050558534e-05,
      "loss": 0.2447,
      "num_input_tokens_seen": 6929536,
      "step": 10620
    },
    {
      "epoch": 5.568658280922432,
      "grad_norm": 1.9933173656463623,
      "learning_rate": 4.530683782777825e-05,
      "loss": 0.3797,
      "num_input_tokens_seen": 6932064,
      "step": 10625
    },
    {
      "epoch": 5.571278825995807,
      "grad_norm": 1.4775598049163818,
      "learning_rate": 4.53001663517443e-05,
      "loss": 0.3196,
      "num_input_tokens_seen": 6935296,
      "step": 10630
    },
    {
      "epoch": 5.573899371069182,
      "grad_norm": 1.1688227653503418,
      "learning_rate": 4.529349062914714e-05,
      "loss": 0.2733,
      "num_input_tokens_seen": 6937824,
      "step": 10635
    },
    {
      "epoch": 5.576519916142558,
      "grad_norm": 1.568637490272522,
      "learning_rate": 4.528681066138326e-05,
      "loss": 0.3403,
      "num_input_tokens_seen": 6941600,
      "step": 10640
    },
    {
      "epoch": 5.579140461215933,
      "grad_norm": 1.1909693479537964,
      "learning_rate": 4.528012644985005e-05,
      "loss": 0.3273,
      "num_input_tokens_seen": 6944608,
      "step": 10645
    },
    {
      "epoch": 5.581761006289308,
      "grad_norm": 1.6370781660079956,
      "learning_rate": 4.527343799594574e-05,
      "loss": 0.3793,
      "num_input_tokens_seen": 6948000,
      "step": 10650
    },
    {
      "epoch": 5.584381551362683,
      "grad_norm": 1.7064728736877441,
      "learning_rate": 4.5266745301069494e-05,
      "loss": 0.248,
      "num_input_tokens_seen": 6951424,
      "step": 10655
    },
    {
      "epoch": 5.587002096436059,
      "grad_norm": 1.2501580715179443,
      "learning_rate": 4.526004836662135e-05,
      "loss": 0.2945,
      "num_input_tokens_seen": 6956000,
      "step": 10660
    },
    {
      "epoch": 5.589622641509434,
      "grad_norm": 1.3524982929229736,
      "learning_rate": 4.525334719400223e-05,
      "loss": 0.3296,
      "num_input_tokens_seen": 6958944,
      "step": 10665
    },
    {
      "epoch": 5.5922431865828095,
      "grad_norm": 1.1408910751342773,
      "learning_rate": 4.524664178461393e-05,
      "loss": 0.2856,
      "num_input_tokens_seen": 6962304,
      "step": 10670
    },
    {
      "epoch": 5.594863731656185,
      "grad_norm": 1.7092353105545044,
      "learning_rate": 4.5239932139859154e-05,
      "loss": 0.3645,
      "num_input_tokens_seen": 6965632,
      "step": 10675
    },
    {
      "epoch": 5.59748427672956,
      "grad_norm": 0.9598587155342102,
      "learning_rate": 4.523321826114149e-05,
      "loss": 0.3735,
      "num_input_tokens_seen": 6969312,
      "step": 10680
    },
    {
      "epoch": 5.600104821802935,
      "grad_norm": 0.9561189413070679,
      "learning_rate": 4.5226500149865395e-05,
      "loss": 0.2484,
      "num_input_tokens_seen": 6972576,
      "step": 10685
    },
    {
      "epoch": 5.60272536687631,
      "grad_norm": 1.1104116439819336,
      "learning_rate": 4.5219777807436224e-05,
      "loss": 0.2227,
      "num_input_tokens_seen": 6975872,
      "step": 10690
    },
    {
      "epoch": 5.605345911949685,
      "grad_norm": 0.9892953634262085,
      "learning_rate": 4.5213051235260225e-05,
      "loss": 0.2777,
      "num_input_tokens_seen": 6978784,
      "step": 10695
    },
    {
      "epoch": 5.6079664570230605,
      "grad_norm": 1.4463675022125244,
      "learning_rate": 4.52063204347445e-05,
      "loss": 0.467,
      "num_input_tokens_seen": 6981408,
      "step": 10700
    },
    {
      "epoch": 5.610587002096436,
      "grad_norm": 1.5179917812347412,
      "learning_rate": 4.519958540729707e-05,
      "loss": 0.3016,
      "num_input_tokens_seen": 6984800,
      "step": 10705
    },
    {
      "epoch": 5.613207547169811,
      "grad_norm": 1.2990341186523438,
      "learning_rate": 4.519284615432682e-05,
      "loss": 0.3303,
      "num_input_tokens_seen": 6988192,
      "step": 10710
    },
    {
      "epoch": 5.615828092243187,
      "grad_norm": 1.455154299736023,
      "learning_rate": 4.518610267724352e-05,
      "loss": 0.246,
      "num_input_tokens_seen": 6992416,
      "step": 10715
    },
    {
      "epoch": 5.618448637316562,
      "grad_norm": 1.6669937372207642,
      "learning_rate": 4.517935497745786e-05,
      "loss": 0.2863,
      "num_input_tokens_seen": 6995520,
      "step": 10720
    },
    {
      "epoch": 5.621069182389937,
      "grad_norm": 1.000957727432251,
      "learning_rate": 4.517260305638133e-05,
      "loss": 0.2472,
      "num_input_tokens_seen": 6998464,
      "step": 10725
    },
    {
      "epoch": 5.623689727463312,
      "grad_norm": 1.3473416566848755,
      "learning_rate": 4.516584691542639e-05,
      "loss": 0.3122,
      "num_input_tokens_seen": 7001600,
      "step": 10730
    },
    {
      "epoch": 5.626310272536688,
      "grad_norm": 1.6902351379394531,
      "learning_rate": 4.515908655600634e-05,
      "loss": 0.3108,
      "num_input_tokens_seen": 7004704,
      "step": 10735
    },
    {
      "epoch": 5.628930817610063,
      "grad_norm": 1.5110994577407837,
      "learning_rate": 4.515232197953537e-05,
      "loss": 0.4083,
      "num_input_tokens_seen": 7007424,
      "step": 10740
    },
    {
      "epoch": 5.631551362683438,
      "grad_norm": 0.7231755256652832,
      "learning_rate": 4.514555318742856e-05,
      "loss": 0.2514,
      "num_input_tokens_seen": 7010688,
      "step": 10745
    },
    {
      "epoch": 5.634171907756813,
      "grad_norm": 1.4764353036880493,
      "learning_rate": 4.513878018110184e-05,
      "loss": 0.229,
      "num_input_tokens_seen": 7014048,
      "step": 10750
    },
    {
      "epoch": 5.636792452830189,
      "grad_norm": 1.2911092042922974,
      "learning_rate": 4.513200296197207e-05,
      "loss": 0.2816,
      "num_input_tokens_seen": 7017856,
      "step": 10755
    },
    {
      "epoch": 5.639412997903564,
      "grad_norm": 1.2339463233947754,
      "learning_rate": 4.512522153145696e-05,
      "loss": 0.2412,
      "num_input_tokens_seen": 7021248,
      "step": 10760
    },
    {
      "epoch": 5.6420335429769395,
      "grad_norm": 1.3583383560180664,
      "learning_rate": 4.5118435890975106e-05,
      "loss": 0.3291,
      "num_input_tokens_seen": 7024320,
      "step": 10765
    },
    {
      "epoch": 5.644654088050315,
      "grad_norm": 2.670746088027954,
      "learning_rate": 4.511164604194598e-05,
      "loss": 0.3411,
      "num_input_tokens_seen": 7027296,
      "step": 10770
    },
    {
      "epoch": 5.64727463312369,
      "grad_norm": 0.8700955510139465,
      "learning_rate": 4.5104851985789956e-05,
      "loss": 0.3025,
      "num_input_tokens_seen": 7030784,
      "step": 10775
    },
    {
      "epoch": 5.649895178197065,
      "grad_norm": 1.005408763885498,
      "learning_rate": 4.5098053723928254e-05,
      "loss": 0.5014,
      "num_input_tokens_seen": 7034112,
      "step": 10780
    },
    {
      "epoch": 5.65251572327044,
      "grad_norm": 1.8125096559524536,
      "learning_rate": 4.509125125778302e-05,
      "loss": 0.2582,
      "num_input_tokens_seen": 7039072,
      "step": 10785
    },
    {
      "epoch": 5.655136268343815,
      "grad_norm": 1.033643364906311,
      "learning_rate": 4.508444458877722e-05,
      "loss": 0.2979,
      "num_input_tokens_seen": 7042304,
      "step": 10790
    },
    {
      "epoch": 5.6577568134171905,
      "grad_norm": 1.3249632120132446,
      "learning_rate": 4.5077633718334755e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 7046400,
      "step": 10795
    },
    {
      "epoch": 5.660377358490566,
      "grad_norm": 1.2735317945480347,
      "learning_rate": 4.5070818647880375e-05,
      "loss": 0.3113,
      "num_input_tokens_seen": 7050656,
      "step": 10800
    },
    {
      "epoch": 5.662997903563941,
      "grad_norm": 1.8864376544952393,
      "learning_rate": 4.506399937883972e-05,
      "loss": 0.3019,
      "num_input_tokens_seen": 7053824,
      "step": 10805
    },
    {
      "epoch": 5.665618448637317,
      "grad_norm": 1.5942356586456299,
      "learning_rate": 4.5057175912639286e-05,
      "loss": 0.3289,
      "num_input_tokens_seen": 7057760,
      "step": 10810
    },
    {
      "epoch": 5.668238993710692,
      "grad_norm": 1.2036046981811523,
      "learning_rate": 4.5050348250706475e-05,
      "loss": 0.2798,
      "num_input_tokens_seen": 7061888,
      "step": 10815
    },
    {
      "epoch": 5.670859538784067,
      "grad_norm": 0.828471302986145,
      "learning_rate": 4.5043516394469566e-05,
      "loss": 0.2484,
      "num_input_tokens_seen": 7064960,
      "step": 10820
    },
    {
      "epoch": 5.673480083857442,
      "grad_norm": 1.256422758102417,
      "learning_rate": 4.503668034535768e-05,
      "loss": 0.2948,
      "num_input_tokens_seen": 7067968,
      "step": 10825
    },
    {
      "epoch": 5.676100628930818,
      "grad_norm": 0.9076723456382751,
      "learning_rate": 4.5029840104800876e-05,
      "loss": 0.3577,
      "num_input_tokens_seen": 7071424,
      "step": 10830
    },
    {
      "epoch": 5.678721174004193,
      "grad_norm": 1.6966363191604614,
      "learning_rate": 4.5022995674230025e-05,
      "loss": 0.3678,
      "num_input_tokens_seen": 7074656,
      "step": 10835
    },
    {
      "epoch": 5.681341719077568,
      "grad_norm": 1.7022022008895874,
      "learning_rate": 4.5016147055076914e-05,
      "loss": 0.3533,
      "num_input_tokens_seen": 7077952,
      "step": 10840
    },
    {
      "epoch": 5.683962264150943,
      "grad_norm": 1.9799598455429077,
      "learning_rate": 4.500929424877419e-05,
      "loss": 0.2403,
      "num_input_tokens_seen": 7080512,
      "step": 10845
    },
    {
      "epoch": 5.686582809224319,
      "grad_norm": 2.319260597229004,
      "learning_rate": 4.5002437256755394e-05,
      "loss": 0.309,
      "num_input_tokens_seen": 7083904,
      "step": 10850
    },
    {
      "epoch": 5.689203354297694,
      "grad_norm": 1.7240008115768433,
      "learning_rate": 4.4995576080454924e-05,
      "loss": 0.2994,
      "num_input_tokens_seen": 7086144,
      "step": 10855
    },
    {
      "epoch": 5.6918238993710695,
      "grad_norm": 1.349819540977478,
      "learning_rate": 4.4988710721308056e-05,
      "loss": 0.3283,
      "num_input_tokens_seen": 7088608,
      "step": 10860
    },
    {
      "epoch": 5.694444444444445,
      "grad_norm": 0.9618187546730042,
      "learning_rate": 4.498184118075095e-05,
      "loss": 0.2564,
      "num_input_tokens_seen": 7091296,
      "step": 10865
    },
    {
      "epoch": 5.69706498951782,
      "grad_norm": 1.8769710063934326,
      "learning_rate": 4.497496746022063e-05,
      "loss": 0.2748,
      "num_input_tokens_seen": 7095008,
      "step": 10870
    },
    {
      "epoch": 5.699685534591195,
      "grad_norm": 2.127739191055298,
      "learning_rate": 4.4968089561155015e-05,
      "loss": 0.3163,
      "num_input_tokens_seen": 7098016,
      "step": 10875
    },
    {
      "epoch": 5.70230607966457,
      "grad_norm": 0.8810366988182068,
      "learning_rate": 4.496120748499286e-05,
      "loss": 0.2523,
      "num_input_tokens_seen": 7101056,
      "step": 10880
    },
    {
      "epoch": 5.704926624737945,
      "grad_norm": 1.501263976097107,
      "learning_rate": 4.4954321233173833e-05,
      "loss": 0.2434,
      "num_input_tokens_seen": 7103488,
      "step": 10885
    },
    {
      "epoch": 5.7075471698113205,
      "grad_norm": 1.2046176195144653,
      "learning_rate": 4.494743080713846e-05,
      "loss": 0.274,
      "num_input_tokens_seen": 7107072,
      "step": 10890
    },
    {
      "epoch": 5.710167714884696,
      "grad_norm": 1.290231704711914,
      "learning_rate": 4.494053620832812e-05,
      "loss": 0.3316,
      "num_input_tokens_seen": 7110336,
      "step": 10895
    },
    {
      "epoch": 5.712788259958071,
      "grad_norm": 1.3049407005310059,
      "learning_rate": 4.493363743818511e-05,
      "loss": 0.3698,
      "num_input_tokens_seen": 7114240,
      "step": 10900
    },
    {
      "epoch": 5.715408805031447,
      "grad_norm": 1.0704704523086548,
      "learning_rate": 4.492673449815257e-05,
      "loss": 0.3589,
      "num_input_tokens_seen": 7121216,
      "step": 10905
    },
    {
      "epoch": 5.718029350104822,
      "grad_norm": 1.789264440536499,
      "learning_rate": 4.4919827389674496e-05,
      "loss": 0.2935,
      "num_input_tokens_seen": 7124800,
      "step": 10910
    },
    {
      "epoch": 5.720649895178197,
      "grad_norm": 1.3858706951141357,
      "learning_rate": 4.491291611419579e-05,
      "loss": 0.394,
      "num_input_tokens_seen": 7128224,
      "step": 10915
    },
    {
      "epoch": 5.723270440251572,
      "grad_norm": 1.5043528079986572,
      "learning_rate": 4.4906000673162216e-05,
      "loss": 0.2532,
      "num_input_tokens_seen": 7131456,
      "step": 10920
    },
    {
      "epoch": 5.725890985324948,
      "grad_norm": 1.5285065174102783,
      "learning_rate": 4.48990810680204e-05,
      "loss": 0.3263,
      "num_input_tokens_seen": 7134944,
      "step": 10925
    },
    {
      "epoch": 5.728511530398323,
      "grad_norm": 0.9201670289039612,
      "learning_rate": 4.4892157300217844e-05,
      "loss": 0.4392,
      "num_input_tokens_seen": 7139200,
      "step": 10930
    },
    {
      "epoch": 5.731132075471698,
      "grad_norm": 1.6029655933380127,
      "learning_rate": 4.488522937120293e-05,
      "loss": 0.2547,
      "num_input_tokens_seen": 7141792,
      "step": 10935
    },
    {
      "epoch": 5.733752620545073,
      "grad_norm": 0.6383281350135803,
      "learning_rate": 4.4878297282424886e-05,
      "loss": 0.2893,
      "num_input_tokens_seen": 7145152,
      "step": 10940
    },
    {
      "epoch": 5.736373165618449,
      "grad_norm": 1.8511571884155273,
      "learning_rate": 4.4871361035333836e-05,
      "loss": 0.4461,
      "num_input_tokens_seen": 7148192,
      "step": 10945
    },
    {
      "epoch": 5.738993710691824,
      "grad_norm": 1.4222509860992432,
      "learning_rate": 4.486442063138076e-05,
      "loss": 0.2745,
      "num_input_tokens_seen": 7151520,
      "step": 10950
    },
    {
      "epoch": 5.7416142557651995,
      "grad_norm": 2.301616668701172,
      "learning_rate": 4.4857476072017516e-05,
      "loss": 0.2743,
      "num_input_tokens_seen": 7154496,
      "step": 10955
    },
    {
      "epoch": 5.744234800838575,
      "grad_norm": 1.075271725654602,
      "learning_rate": 4.485052735869682e-05,
      "loss": 0.3487,
      "num_input_tokens_seen": 7157600,
      "step": 10960
    },
    {
      "epoch": 5.74685534591195,
      "grad_norm": 1.7491475343704224,
      "learning_rate": 4.484357449287228e-05,
      "loss": 0.3314,
      "num_input_tokens_seen": 7160224,
      "step": 10965
    },
    {
      "epoch": 5.749475890985325,
      "grad_norm": 0.9699289202690125,
      "learning_rate": 4.483661747599833e-05,
      "loss": 0.241,
      "num_input_tokens_seen": 7162720,
      "step": 10970
    },
    {
      "epoch": 5.7520964360587,
      "grad_norm": 0.9987399578094482,
      "learning_rate": 4.482965630953031e-05,
      "loss": 0.4171,
      "num_input_tokens_seen": 7166240,
      "step": 10975
    },
    {
      "epoch": 5.754716981132075,
      "grad_norm": 1.1940622329711914,
      "learning_rate": 4.4822690994924425e-05,
      "loss": 0.3717,
      "num_input_tokens_seen": 7172480,
      "step": 10980
    },
    {
      "epoch": 5.7573375262054505,
      "grad_norm": 1.4611202478408813,
      "learning_rate": 4.481572153363773e-05,
      "loss": 0.3721,
      "num_input_tokens_seen": 7176416,
      "step": 10985
    },
    {
      "epoch": 5.759958071278826,
      "grad_norm": 3.1561508178710938,
      "learning_rate": 4.480874792712817e-05,
      "loss": 0.4179,
      "num_input_tokens_seen": 7180832,
      "step": 10990
    },
    {
      "epoch": 5.762578616352201,
      "grad_norm": 2.524629831314087,
      "learning_rate": 4.480177017685452e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 7184416,
      "step": 10995
    },
    {
      "epoch": 5.765199161425577,
      "grad_norm": 1.5218448638916016,
      "learning_rate": 4.479478828427647e-05,
      "loss": 0.2885,
      "num_input_tokens_seen": 7186816,
      "step": 11000
    },
    {
      "epoch": 5.767819706498952,
      "grad_norm": 1.5243637561798096,
      "learning_rate": 4.478780225085454e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 7189824,
      "step": 11005
    },
    {
      "epoch": 5.770440251572327,
      "grad_norm": 2.17729115486145,
      "learning_rate": 4.4780812078050125e-05,
      "loss": 0.3819,
      "num_input_tokens_seen": 7192640,
      "step": 11010
    },
    {
      "epoch": 5.773060796645702,
      "grad_norm": 1.9334243535995483,
      "learning_rate": 4.47738177673255e-05,
      "loss": 0.3513,
      "num_input_tokens_seen": 7195328,
      "step": 11015
    },
    {
      "epoch": 5.7756813417190775,
      "grad_norm": 2.921928644180298,
      "learning_rate": 4.4766819320143785e-05,
      "loss": 0.4853,
      "num_input_tokens_seen": 7198464,
      "step": 11020
    },
    {
      "epoch": 5.778301886792453,
      "grad_norm": 1.2404422760009766,
      "learning_rate": 4.475981673796899e-05,
      "loss": 0.286,
      "num_input_tokens_seen": 7202080,
      "step": 11025
    },
    {
      "epoch": 5.780922431865828,
      "grad_norm": 1.7783629894256592,
      "learning_rate": 4.475281002226596e-05,
      "loss": 0.321,
      "num_input_tokens_seen": 7204832,
      "step": 11030
    },
    {
      "epoch": 5.783542976939203,
      "grad_norm": 1.374849796295166,
      "learning_rate": 4.4745799174500414e-05,
      "loss": 0.2806,
      "num_input_tokens_seen": 7208128,
      "step": 11035
    },
    {
      "epoch": 5.786163522012579,
      "grad_norm": 1.6811445951461792,
      "learning_rate": 4.473878419613896e-05,
      "loss": 0.2864,
      "num_input_tokens_seen": 7211232,
      "step": 11040
    },
    {
      "epoch": 5.788784067085954,
      "grad_norm": 1.4545032978057861,
      "learning_rate": 4.473176508864905e-05,
      "loss": 0.292,
      "num_input_tokens_seen": 7214720,
      "step": 11045
    },
    {
      "epoch": 5.7914046121593294,
      "grad_norm": 1.537474274635315,
      "learning_rate": 4.4724741853498995e-05,
      "loss": 0.3966,
      "num_input_tokens_seen": 7219200,
      "step": 11050
    },
    {
      "epoch": 5.794025157232705,
      "grad_norm": 1.1136126518249512,
      "learning_rate": 4.471771449215797e-05,
      "loss": 0.318,
      "num_input_tokens_seen": 7223808,
      "step": 11055
    },
    {
      "epoch": 5.79664570230608,
      "grad_norm": 2.19179630279541,
      "learning_rate": 4.471068300609603e-05,
      "loss": 0.3468,
      "num_input_tokens_seen": 7227264,
      "step": 11060
    },
    {
      "epoch": 5.799266247379455,
      "grad_norm": 1.277186393737793,
      "learning_rate": 4.4703647396784065e-05,
      "loss": 0.3682,
      "num_input_tokens_seen": 7232672,
      "step": 11065
    },
    {
      "epoch": 5.80188679245283,
      "grad_norm": 1.638135313987732,
      "learning_rate": 4.469660766569387e-05,
      "loss": 0.284,
      "num_input_tokens_seen": 7235584,
      "step": 11070
    },
    {
      "epoch": 5.804507337526205,
      "grad_norm": 1.681447982788086,
      "learning_rate": 4.4689563814298057e-05,
      "loss": 0.3608,
      "num_input_tokens_seen": 7238592,
      "step": 11075
    },
    {
      "epoch": 5.8071278825995805,
      "grad_norm": 1.4223651885986328,
      "learning_rate": 4.468251584407012e-05,
      "loss": 0.3333,
      "num_input_tokens_seen": 7241952,
      "step": 11080
    },
    {
      "epoch": 5.809748427672956,
      "grad_norm": 1.128317952156067,
      "learning_rate": 4.467546375648443e-05,
      "loss": 0.3138,
      "num_input_tokens_seen": 7245056,
      "step": 11085
    },
    {
      "epoch": 5.812368972746331,
      "grad_norm": 1.0918641090393066,
      "learning_rate": 4.4668407553016176e-05,
      "loss": 0.3458,
      "num_input_tokens_seen": 7248896,
      "step": 11090
    },
    {
      "epoch": 5.814989517819707,
      "grad_norm": 1.5493465662002563,
      "learning_rate": 4.466134723514146e-05,
      "loss": 0.4015,
      "num_input_tokens_seen": 7251680,
      "step": 11095
    },
    {
      "epoch": 5.817610062893082,
      "grad_norm": 1.4326845407485962,
      "learning_rate": 4.465428280433721e-05,
      "loss": 0.3965,
      "num_input_tokens_seen": 7254432,
      "step": 11100
    },
    {
      "epoch": 5.820230607966457,
      "grad_norm": 4.047443866729736,
      "learning_rate": 4.464721426208123e-05,
      "loss": 0.3124,
      "num_input_tokens_seen": 7257280,
      "step": 11105
    },
    {
      "epoch": 5.822851153039832,
      "grad_norm": 1.016156792640686,
      "learning_rate": 4.464014160985217e-05,
      "loss": 0.2752,
      "num_input_tokens_seen": 7259840,
      "step": 11110
    },
    {
      "epoch": 5.8254716981132075,
      "grad_norm": 1.3173089027404785,
      "learning_rate": 4.463306484912955e-05,
      "loss": 0.3036,
      "num_input_tokens_seen": 7262624,
      "step": 11115
    },
    {
      "epoch": 5.828092243186583,
      "grad_norm": 0.9862499833106995,
      "learning_rate": 4.462598398139375e-05,
      "loss": 0.3028,
      "num_input_tokens_seen": 7265568,
      "step": 11120
    },
    {
      "epoch": 5.830712788259958,
      "grad_norm": 1.9172276258468628,
      "learning_rate": 4.4618899008126005e-05,
      "loss": 0.2722,
      "num_input_tokens_seen": 7268896,
      "step": 11125
    },
    {
      "epoch": 5.833333333333333,
      "grad_norm": 1.758162260055542,
      "learning_rate": 4.461180993080841e-05,
      "loss": 0.2738,
      "num_input_tokens_seen": 7271328,
      "step": 11130
    },
    {
      "epoch": 5.835953878406709,
      "grad_norm": 1.5454579591751099,
      "learning_rate": 4.460471675092393e-05,
      "loss": 0.378,
      "num_input_tokens_seen": 7274752,
      "step": 11135
    },
    {
      "epoch": 5.838574423480084,
      "grad_norm": 1.1476938724517822,
      "learning_rate": 4.4597619469956356e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 7279168,
      "step": 11140
    },
    {
      "epoch": 5.841194968553459,
      "grad_norm": 1.5660768747329712,
      "learning_rate": 4.459051808939038e-05,
      "loss": 0.317,
      "num_input_tokens_seen": 7282944,
      "step": 11145
    },
    {
      "epoch": 5.843815513626835,
      "grad_norm": 2.6496222019195557,
      "learning_rate": 4.458341261071151e-05,
      "loss": 0.3901,
      "num_input_tokens_seen": 7285824,
      "step": 11150
    },
    {
      "epoch": 5.84643605870021,
      "grad_norm": 2.521008253097534,
      "learning_rate": 4.457630303540615e-05,
      "loss": 0.3073,
      "num_input_tokens_seen": 7288384,
      "step": 11155
    },
    {
      "epoch": 5.849056603773585,
      "grad_norm": 1.635538935661316,
      "learning_rate": 4.456918936496154e-05,
      "loss": 0.3647,
      "num_input_tokens_seen": 7291840,
      "step": 11160
    },
    {
      "epoch": 5.85167714884696,
      "grad_norm": 1.159880518913269,
      "learning_rate": 4.456207160086575e-05,
      "loss": 0.3062,
      "num_input_tokens_seen": 7296032,
      "step": 11165
    },
    {
      "epoch": 5.854297693920335,
      "grad_norm": 1.2385594844818115,
      "learning_rate": 4.4554949744607775e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 7298560,
      "step": 11170
    },
    {
      "epoch": 5.8569182389937104,
      "grad_norm": 1.3231070041656494,
      "learning_rate": 4.4547823797677405e-05,
      "loss": 0.2893,
      "num_input_tokens_seen": 7301376,
      "step": 11175
    },
    {
      "epoch": 5.859538784067086,
      "grad_norm": 1.9939652681350708,
      "learning_rate": 4.45406937615653e-05,
      "loss": 0.3752,
      "num_input_tokens_seen": 7304896,
      "step": 11180
    },
    {
      "epoch": 5.862159329140461,
      "grad_norm": 1.4023303985595703,
      "learning_rate": 4.4533559637763006e-05,
      "loss": 0.3333,
      "num_input_tokens_seen": 7307968,
      "step": 11185
    },
    {
      "epoch": 5.864779874213837,
      "grad_norm": 1.3463027477264404,
      "learning_rate": 4.452642142776288e-05,
      "loss": 0.3422,
      "num_input_tokens_seen": 7311040,
      "step": 11190
    },
    {
      "epoch": 5.867400419287212,
      "grad_norm": 1.2330178022384644,
      "learning_rate": 4.451927913305816e-05,
      "loss": 0.2813,
      "num_input_tokens_seen": 7314848,
      "step": 11195
    },
    {
      "epoch": 5.870020964360587,
      "grad_norm": 1.3436932563781738,
      "learning_rate": 4.4512132755142935e-05,
      "loss": 0.3422,
      "num_input_tokens_seen": 7317440,
      "step": 11200
    },
    {
      "epoch": 5.872641509433962,
      "grad_norm": 1.219965934753418,
      "learning_rate": 4.450498229551214e-05,
      "loss": 0.3481,
      "num_input_tokens_seen": 7320288,
      "step": 11205
    },
    {
      "epoch": 5.8752620545073375,
      "grad_norm": 1.4213132858276367,
      "learning_rate": 4.449782775566158e-05,
      "loss": 0.3764,
      "num_input_tokens_seen": 7323168,
      "step": 11210
    },
    {
      "epoch": 5.877882599580713,
      "grad_norm": 1.0711055994033813,
      "learning_rate": 4.449066913708789e-05,
      "loss": 0.291,
      "num_input_tokens_seen": 7326048,
      "step": 11215
    },
    {
      "epoch": 5.880503144654088,
      "grad_norm": 1.4361766576766968,
      "learning_rate": 4.448350644128858e-05,
      "loss": 0.3297,
      "num_input_tokens_seen": 7328288,
      "step": 11220
    },
    {
      "epoch": 5.883123689727463,
      "grad_norm": 2.8279199600219727,
      "learning_rate": 4.4476339669762e-05,
      "loss": 0.3533,
      "num_input_tokens_seen": 7331328,
      "step": 11225
    },
    {
      "epoch": 5.885744234800838,
      "grad_norm": 1.0398497581481934,
      "learning_rate": 4.446916882400737e-05,
      "loss": 0.2884,
      "num_input_tokens_seen": 7334720,
      "step": 11230
    },
    {
      "epoch": 5.888364779874214,
      "grad_norm": 1.34121835231781,
      "learning_rate": 4.446199390552472e-05,
      "loss": 0.2676,
      "num_input_tokens_seen": 7337312,
      "step": 11235
    },
    {
      "epoch": 5.890985324947589,
      "grad_norm": 1.5882622003555298,
      "learning_rate": 4.445481491581499e-05,
      "loss": 0.3822,
      "num_input_tokens_seen": 7340384,
      "step": 11240
    },
    {
      "epoch": 5.893605870020965,
      "grad_norm": 0.8331185579299927,
      "learning_rate": 4.444763185637993e-05,
      "loss": 0.3615,
      "num_input_tokens_seen": 7343936,
      "step": 11245
    },
    {
      "epoch": 5.89622641509434,
      "grad_norm": 1.395374059677124,
      "learning_rate": 4.4440444728722166e-05,
      "loss": 0.3192,
      "num_input_tokens_seen": 7346848,
      "step": 11250
    },
    {
      "epoch": 5.898846960167715,
      "grad_norm": 1.9048618078231812,
      "learning_rate": 4.4433253534345136e-05,
      "loss": 0.3991,
      "num_input_tokens_seen": 7350112,
      "step": 11255
    },
    {
      "epoch": 5.90146750524109,
      "grad_norm": 1.0674384832382202,
      "learning_rate": 4.442605827475318e-05,
      "loss": 0.2958,
      "num_input_tokens_seen": 7353952,
      "step": 11260
    },
    {
      "epoch": 5.904088050314465,
      "grad_norm": 1.8741483688354492,
      "learning_rate": 4.4418858951451457e-05,
      "loss": 0.3979,
      "num_input_tokens_seen": 7357440,
      "step": 11265
    },
    {
      "epoch": 5.90670859538784,
      "grad_norm": 1.572999119758606,
      "learning_rate": 4.441165556594599e-05,
      "loss": 0.3836,
      "num_input_tokens_seen": 7360800,
      "step": 11270
    },
    {
      "epoch": 5.909329140461216,
      "grad_norm": 1.855232834815979,
      "learning_rate": 4.440444811974363e-05,
      "loss": 0.3847,
      "num_input_tokens_seen": 7364032,
      "step": 11275
    },
    {
      "epoch": 5.911949685534591,
      "grad_norm": 1.890014410018921,
      "learning_rate": 4.43972366143521e-05,
      "loss": 0.3654,
      "num_input_tokens_seen": 7366880,
      "step": 11280
    },
    {
      "epoch": 5.914570230607967,
      "grad_norm": 1.9710819721221924,
      "learning_rate": 4.4390021051279975e-05,
      "loss": 0.2861,
      "num_input_tokens_seen": 7369664,
      "step": 11285
    },
    {
      "epoch": 5.917190775681342,
      "grad_norm": 1.3559972047805786,
      "learning_rate": 4.438280143203665e-05,
      "loss": 0.4137,
      "num_input_tokens_seen": 7372192,
      "step": 11290
    },
    {
      "epoch": 5.919811320754717,
      "grad_norm": 2.641216278076172,
      "learning_rate": 4.43755777581324e-05,
      "loss": 0.3048,
      "num_input_tokens_seen": 7374720,
      "step": 11295
    },
    {
      "epoch": 5.922431865828092,
      "grad_norm": 1.3351495265960693,
      "learning_rate": 4.4368350031078334e-05,
      "loss": 0.302,
      "num_input_tokens_seen": 7378112,
      "step": 11300
    },
    {
      "epoch": 5.9250524109014675,
      "grad_norm": 1.8230235576629639,
      "learning_rate": 4.43611182523864e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 7381728,
      "step": 11305
    },
    {
      "epoch": 5.927672955974843,
      "grad_norm": 2.201660394668579,
      "learning_rate": 4.4353882423569416e-05,
      "loss": 0.3529,
      "num_input_tokens_seen": 7385504,
      "step": 11310
    },
    {
      "epoch": 5.930293501048218,
      "grad_norm": 0.8464011549949646,
      "learning_rate": 4.434664254614103e-05,
      "loss": 0.2765,
      "num_input_tokens_seen": 7389600,
      "step": 11315
    },
    {
      "epoch": 5.932914046121593,
      "grad_norm": 2.156007766723633,
      "learning_rate": 4.4339398621615734e-05,
      "loss": 0.3063,
      "num_input_tokens_seen": 7394240,
      "step": 11320
    },
    {
      "epoch": 5.935534591194968,
      "grad_norm": 0.911374568939209,
      "learning_rate": 4.433215065150889e-05,
      "loss": 0.3342,
      "num_input_tokens_seen": 7396704,
      "step": 11325
    },
    {
      "epoch": 5.938155136268344,
      "grad_norm": 0.9641897678375244,
      "learning_rate": 4.4324898637336674e-05,
      "loss": 0.282,
      "num_input_tokens_seen": 7400352,
      "step": 11330
    },
    {
      "epoch": 5.940775681341719,
      "grad_norm": 0.9615439176559448,
      "learning_rate": 4.431764258061615e-05,
      "loss": 0.2618,
      "num_input_tokens_seen": 7402816,
      "step": 11335
    },
    {
      "epoch": 5.943396226415095,
      "grad_norm": 1.361656904220581,
      "learning_rate": 4.431038248286517e-05,
      "loss": 0.2878,
      "num_input_tokens_seen": 7405568,
      "step": 11340
    },
    {
      "epoch": 5.94601677148847,
      "grad_norm": 1.1012935638427734,
      "learning_rate": 4.430311834560248e-05,
      "loss": 0.3416,
      "num_input_tokens_seen": 7408608,
      "step": 11345
    },
    {
      "epoch": 5.948637316561845,
      "grad_norm": 1.372615933418274,
      "learning_rate": 4.4295850170347664e-05,
      "loss": 0.3531,
      "num_input_tokens_seen": 7412448,
      "step": 11350
    },
    {
      "epoch": 5.95125786163522,
      "grad_norm": 3.305636405944824,
      "learning_rate": 4.428857795862113e-05,
      "loss": 0.26,
      "num_input_tokens_seen": 7415104,
      "step": 11355
    },
    {
      "epoch": 5.953878406708595,
      "grad_norm": 1.5520421266555786,
      "learning_rate": 4.428130171194414e-05,
      "loss": 0.2973,
      "num_input_tokens_seen": 7418944,
      "step": 11360
    },
    {
      "epoch": 5.95649895178197,
      "grad_norm": 0.9626286625862122,
      "learning_rate": 4.427402143183879e-05,
      "loss": 0.2703,
      "num_input_tokens_seen": 7422816,
      "step": 11365
    },
    {
      "epoch": 5.959119496855346,
      "grad_norm": 1.3667161464691162,
      "learning_rate": 4.426673711982806e-05,
      "loss": 0.238,
      "num_input_tokens_seen": 7425824,
      "step": 11370
    },
    {
      "epoch": 5.961740041928721,
      "grad_norm": 0.8425794243812561,
      "learning_rate": 4.425944877743573e-05,
      "loss": 0.3839,
      "num_input_tokens_seen": 7429088,
      "step": 11375
    },
    {
      "epoch": 5.964360587002097,
      "grad_norm": 1.0415303707122803,
      "learning_rate": 4.425215640618644e-05,
      "loss": 0.3912,
      "num_input_tokens_seen": 7432736,
      "step": 11380
    },
    {
      "epoch": 5.966981132075472,
      "grad_norm": 1.3298499584197998,
      "learning_rate": 4.424486000760567e-05,
      "loss": 0.3752,
      "num_input_tokens_seen": 7435840,
      "step": 11385
    },
    {
      "epoch": 5.969601677148847,
      "grad_norm": 2.1381375789642334,
      "learning_rate": 4.423755958321974e-05,
      "loss": 0.276,
      "num_input_tokens_seen": 7441056,
      "step": 11390
    },
    {
      "epoch": 5.972222222222222,
      "grad_norm": 0.9094357490539551,
      "learning_rate": 4.423025513455582e-05,
      "loss": 0.3614,
      "num_input_tokens_seen": 7444640,
      "step": 11395
    },
    {
      "epoch": 5.9748427672955975,
      "grad_norm": 1.4976441860198975,
      "learning_rate": 4.422294666314192e-05,
      "loss": 0.3314,
      "num_input_tokens_seen": 7448096,
      "step": 11400
    },
    {
      "epoch": 5.977463312368973,
      "grad_norm": 1.1663553714752197,
      "learning_rate": 4.421563417050688e-05,
      "loss": 0.2885,
      "num_input_tokens_seen": 7451520,
      "step": 11405
    },
    {
      "epoch": 5.980083857442348,
      "grad_norm": 1.9030872583389282,
      "learning_rate": 4.42083176581804e-05,
      "loss": 0.2929,
      "num_input_tokens_seen": 7454048,
      "step": 11410
    },
    {
      "epoch": 5.982704402515723,
      "grad_norm": 1.397810459136963,
      "learning_rate": 4.4200997127693e-05,
      "loss": 0.2551,
      "num_input_tokens_seen": 7457216,
      "step": 11415
    },
    {
      "epoch": 5.985324947589098,
      "grad_norm": 1.1302074193954468,
      "learning_rate": 4.419367258057606e-05,
      "loss": 0.3328,
      "num_input_tokens_seen": 7460992,
      "step": 11420
    },
    {
      "epoch": 5.987945492662474,
      "grad_norm": 1.7426843643188477,
      "learning_rate": 4.418634401836179e-05,
      "loss": 0.2707,
      "num_input_tokens_seen": 7463488,
      "step": 11425
    },
    {
      "epoch": 5.990566037735849,
      "grad_norm": 1.1650952100753784,
      "learning_rate": 4.417901144258324e-05,
      "loss": 0.338,
      "num_input_tokens_seen": 7466304,
      "step": 11430
    },
    {
      "epoch": 5.993186582809225,
      "grad_norm": 1.0702451467514038,
      "learning_rate": 4.4171674854774305e-05,
      "loss": 0.3519,
      "num_input_tokens_seen": 7470176,
      "step": 11435
    },
    {
      "epoch": 5.9958071278826,
      "grad_norm": 1.6108235120773315,
      "learning_rate": 4.416433425646971e-05,
      "loss": 0.2776,
      "num_input_tokens_seen": 7473376,
      "step": 11440
    },
    {
      "epoch": 5.998427672955975,
      "grad_norm": 1.7362595796585083,
      "learning_rate": 4.415698964920503e-05,
      "loss": 0.3266,
      "num_input_tokens_seen": 7476864,
      "step": 11445
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.5208126306533813,
      "eval_runtime": 11.8671,
      "eval_samples_per_second": 71.458,
      "eval_steps_per_second": 17.865,
      "num_input_tokens_seen": 7478504,
      "step": 11448
    },
    {
      "epoch": 6.00104821802935,
      "grad_norm": 0.972389280796051,
      "learning_rate": 4.414964103451667e-05,
      "loss": 0.2806,
      "num_input_tokens_seen": 7479688,
      "step": 11450
    },
    {
      "epoch": 6.003668763102725,
      "grad_norm": 0.8405939936637878,
      "learning_rate": 4.414228841394187e-05,
      "loss": 0.2332,
      "num_input_tokens_seen": 7483176,
      "step": 11455
    },
    {
      "epoch": 6.0062893081761,
      "grad_norm": 1.1055725812911987,
      "learning_rate": 4.413493178901872e-05,
      "loss": 0.2558,
      "num_input_tokens_seen": 7486120,
      "step": 11460
    },
    {
      "epoch": 6.008909853249476,
      "grad_norm": 1.4292423725128174,
      "learning_rate": 4.412757116128614e-05,
      "loss": 0.2623,
      "num_input_tokens_seen": 7488872,
      "step": 11465
    },
    {
      "epoch": 6.011530398322851,
      "grad_norm": 2.0132369995117188,
      "learning_rate": 4.41202065322839e-05,
      "loss": 0.2907,
      "num_input_tokens_seen": 7491880,
      "step": 11470
    },
    {
      "epoch": 6.014150943396227,
      "grad_norm": 1.6288392543792725,
      "learning_rate": 4.411283790355259e-05,
      "loss": 0.2313,
      "num_input_tokens_seen": 7494952,
      "step": 11475
    },
    {
      "epoch": 6.016771488469602,
      "grad_norm": 1.023410677909851,
      "learning_rate": 4.410546527663364e-05,
      "loss": 0.2925,
      "num_input_tokens_seen": 7498600,
      "step": 11480
    },
    {
      "epoch": 6.019392033542977,
      "grad_norm": 1.1032999753952026,
      "learning_rate": 4.409808865306931e-05,
      "loss": 0.2857,
      "num_input_tokens_seen": 7502696,
      "step": 11485
    },
    {
      "epoch": 6.022012578616352,
      "grad_norm": 1.6792126893997192,
      "learning_rate": 4.409070803440273e-05,
      "loss": 0.2533,
      "num_input_tokens_seen": 7505800,
      "step": 11490
    },
    {
      "epoch": 6.0246331236897275,
      "grad_norm": 2.3090319633483887,
      "learning_rate": 4.4083323422177824e-05,
      "loss": 0.2735,
      "num_input_tokens_seen": 7508712,
      "step": 11495
    },
    {
      "epoch": 6.027253668763103,
      "grad_norm": 1.3311057090759277,
      "learning_rate": 4.407593481793937e-05,
      "loss": 0.2131,
      "num_input_tokens_seen": 7511400,
      "step": 11500
    },
    {
      "epoch": 6.029874213836478,
      "grad_norm": 1.5048727989196777,
      "learning_rate": 4.406854222323298e-05,
      "loss": 0.2816,
      "num_input_tokens_seen": 7515208,
      "step": 11505
    },
    {
      "epoch": 6.032494758909853,
      "grad_norm": 1.68955659866333,
      "learning_rate": 4.406114563960511e-05,
      "loss": 0.1415,
      "num_input_tokens_seen": 7520712,
      "step": 11510
    },
    {
      "epoch": 6.035115303983228,
      "grad_norm": 0.9680950045585632,
      "learning_rate": 4.4053745068603026e-05,
      "loss": 0.2202,
      "num_input_tokens_seen": 7522984,
      "step": 11515
    },
    {
      "epoch": 6.037735849056604,
      "grad_norm": 1.028122067451477,
      "learning_rate": 4.404634051177485e-05,
      "loss": 0.2744,
      "num_input_tokens_seen": 7526024,
      "step": 11520
    },
    {
      "epoch": 6.040356394129979,
      "grad_norm": 2.4939990043640137,
      "learning_rate": 4.403893197066953e-05,
      "loss": 0.2362,
      "num_input_tokens_seen": 7529288,
      "step": 11525
    },
    {
      "epoch": 6.0429769392033545,
      "grad_norm": 1.523971438407898,
      "learning_rate": 4.4031519446836855e-05,
      "loss": 0.226,
      "num_input_tokens_seen": 7531752,
      "step": 11530
    },
    {
      "epoch": 6.04559748427673,
      "grad_norm": 2.0408616065979004,
      "learning_rate": 4.402410294182743e-05,
      "loss": 0.2787,
      "num_input_tokens_seen": 7535112,
      "step": 11535
    },
    {
      "epoch": 6.048218029350105,
      "grad_norm": 3.310154676437378,
      "learning_rate": 4.4016682457192714e-05,
      "loss": 0.226,
      "num_input_tokens_seen": 7538344,
      "step": 11540
    },
    {
      "epoch": 6.05083857442348,
      "grad_norm": 2.005661725997925,
      "learning_rate": 4.400925799448497e-05,
      "loss": 0.2433,
      "num_input_tokens_seen": 7541064,
      "step": 11545
    },
    {
      "epoch": 6.053459119496855,
      "grad_norm": 1.47284996509552,
      "learning_rate": 4.400182955525733e-05,
      "loss": 0.1972,
      "num_input_tokens_seen": 7544808,
      "step": 11550
    },
    {
      "epoch": 6.05607966457023,
      "grad_norm": 1.672224760055542,
      "learning_rate": 4.3994397141063734e-05,
      "loss": 0.2052,
      "num_input_tokens_seen": 7548872,
      "step": 11555
    },
    {
      "epoch": 6.058700209643606,
      "grad_norm": 1.643698811531067,
      "learning_rate": 4.398696075345896e-05,
      "loss": 0.3188,
      "num_input_tokens_seen": 7551912,
      "step": 11560
    },
    {
      "epoch": 6.061320754716981,
      "grad_norm": 1.2905969619750977,
      "learning_rate": 4.39795203939986e-05,
      "loss": 0.207,
      "num_input_tokens_seen": 7554856,
      "step": 11565
    },
    {
      "epoch": 6.063941299790357,
      "grad_norm": 3.095189332962036,
      "learning_rate": 4.397207606423911e-05,
      "loss": 0.3229,
      "num_input_tokens_seen": 7558344,
      "step": 11570
    },
    {
      "epoch": 6.066561844863732,
      "grad_norm": 2.697866916656494,
      "learning_rate": 4.396462776573775e-05,
      "loss": 0.273,
      "num_input_tokens_seen": 7561704,
      "step": 11575
    },
    {
      "epoch": 6.069182389937107,
      "grad_norm": 3.890191078186035,
      "learning_rate": 4.395717550005262e-05,
      "loss": 0.2957,
      "num_input_tokens_seen": 7563816,
      "step": 11580
    },
    {
      "epoch": 6.071802935010482,
      "grad_norm": 1.0568714141845703,
      "learning_rate": 4.394971926874265e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 7566408,
      "step": 11585
    },
    {
      "epoch": 6.0744234800838575,
      "grad_norm": 1.217642903327942,
      "learning_rate": 4.3942259073367604e-05,
      "loss": 0.2287,
      "num_input_tokens_seen": 7569960,
      "step": 11590
    },
    {
      "epoch": 6.077044025157233,
      "grad_norm": 3.8475818634033203,
      "learning_rate": 4.3934794915488065e-05,
      "loss": 0.2299,
      "num_input_tokens_seen": 7572872,
      "step": 11595
    },
    {
      "epoch": 6.079664570230608,
      "grad_norm": 2.6121206283569336,
      "learning_rate": 4.392732679666544e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 7576104,
      "step": 11600
    },
    {
      "epoch": 6.082285115303983,
      "grad_norm": 1.0928539037704468,
      "learning_rate": 4.3919854718461986e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 7578856,
      "step": 11605
    },
    {
      "epoch": 6.084905660377358,
      "grad_norm": 1.4869531393051147,
      "learning_rate": 4.391237868244078e-05,
      "loss": 0.2746,
      "num_input_tokens_seen": 7582408,
      "step": 11610
    },
    {
      "epoch": 6.087526205450734,
      "grad_norm": 1.2383270263671875,
      "learning_rate": 4.390489869016571e-05,
      "loss": 0.2477,
      "num_input_tokens_seen": 7585224,
      "step": 11615
    },
    {
      "epoch": 6.090146750524109,
      "grad_norm": 1.7002452611923218,
      "learning_rate": 4.389741474320152e-05,
      "loss": 0.2408,
      "num_input_tokens_seen": 7587880,
      "step": 11620
    },
    {
      "epoch": 6.0927672955974845,
      "grad_norm": 2.422562599182129,
      "learning_rate": 4.388992684311376e-05,
      "loss": 0.3273,
      "num_input_tokens_seen": 7591784,
      "step": 11625
    },
    {
      "epoch": 6.09538784067086,
      "grad_norm": 1.9452810287475586,
      "learning_rate": 4.388243499146881e-05,
      "loss": 0.2613,
      "num_input_tokens_seen": 7594920,
      "step": 11630
    },
    {
      "epoch": 6.098008385744235,
      "grad_norm": 2.870295524597168,
      "learning_rate": 4.3874939189833876e-05,
      "loss": 0.3908,
      "num_input_tokens_seen": 7598504,
      "step": 11635
    },
    {
      "epoch": 6.10062893081761,
      "grad_norm": 0.9883649945259094,
      "learning_rate": 4.3867439439777005e-05,
      "loss": 0.325,
      "num_input_tokens_seen": 7601640,
      "step": 11640
    },
    {
      "epoch": 6.103249475890985,
      "grad_norm": 1.3404697179794312,
      "learning_rate": 4.385993574286705e-05,
      "loss": 0.2826,
      "num_input_tokens_seen": 7604520,
      "step": 11645
    },
    {
      "epoch": 6.10587002096436,
      "grad_norm": 1.0301520824432373,
      "learning_rate": 4.3852428100673705e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 7608136,
      "step": 11650
    },
    {
      "epoch": 6.1084905660377355,
      "grad_norm": 2.4370272159576416,
      "learning_rate": 4.3844916514767484e-05,
      "loss": 0.3052,
      "num_input_tokens_seen": 7611592,
      "step": 11655
    },
    {
      "epoch": 6.111111111111111,
      "grad_norm": 2.3006129264831543,
      "learning_rate": 4.383740098671972e-05,
      "loss": 0.2676,
      "num_input_tokens_seen": 7614184,
      "step": 11660
    },
    {
      "epoch": 6.113731656184487,
      "grad_norm": 1.7118265628814697,
      "learning_rate": 4.3829881518102576e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 7617064,
      "step": 11665
    },
    {
      "epoch": 6.116352201257862,
      "grad_norm": 1.140360713005066,
      "learning_rate": 4.3822358110489045e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 7620680,
      "step": 11670
    },
    {
      "epoch": 6.118972746331237,
      "grad_norm": 1.640647053718567,
      "learning_rate": 4.381483076545294e-05,
      "loss": 0.1959,
      "num_input_tokens_seen": 7623944,
      "step": 11675
    },
    {
      "epoch": 6.121593291404612,
      "grad_norm": 3.3364157676696777,
      "learning_rate": 4.380729948456888e-05,
      "loss": 0.215,
      "num_input_tokens_seen": 7627144,
      "step": 11680
    },
    {
      "epoch": 6.1242138364779874,
      "grad_norm": 1.5771626234054565,
      "learning_rate": 4.3799764269412336e-05,
      "loss": 0.2836,
      "num_input_tokens_seen": 7630216,
      "step": 11685
    },
    {
      "epoch": 6.126834381551363,
      "grad_norm": 2.708223581314087,
      "learning_rate": 4.379222512155958e-05,
      "loss": 0.2468,
      "num_input_tokens_seen": 7633000,
      "step": 11690
    },
    {
      "epoch": 6.129454926624738,
      "grad_norm": 2.647523880004883,
      "learning_rate": 4.378468204258773e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 7636104,
      "step": 11695
    },
    {
      "epoch": 6.132075471698113,
      "grad_norm": 2.02941632270813,
      "learning_rate": 4.37771350340747e-05,
      "loss": 0.3896,
      "num_input_tokens_seen": 7639208,
      "step": 11700
    },
    {
      "epoch": 6.134696016771488,
      "grad_norm": 0.8152154684066772,
      "learning_rate": 4.376958409759925e-05,
      "loss": 0.2606,
      "num_input_tokens_seen": 7642184,
      "step": 11705
    },
    {
      "epoch": 6.137316561844864,
      "grad_norm": 2.6649422645568848,
      "learning_rate": 4.376202923474093e-05,
      "loss": 0.3036,
      "num_input_tokens_seen": 7645480,
      "step": 11710
    },
    {
      "epoch": 6.139937106918239,
      "grad_norm": 1.055091142654419,
      "learning_rate": 4.3754470447080145e-05,
      "loss": 0.2537,
      "num_input_tokens_seen": 7648936,
      "step": 11715
    },
    {
      "epoch": 6.1425576519916145,
      "grad_norm": 1.4250770807266235,
      "learning_rate": 4.374690773619811e-05,
      "loss": 0.2423,
      "num_input_tokens_seen": 7651816,
      "step": 11720
    },
    {
      "epoch": 6.14517819706499,
      "grad_norm": 1.3823630809783936,
      "learning_rate": 4.373934110367685e-05,
      "loss": 0.2944,
      "num_input_tokens_seen": 7654568,
      "step": 11725
    },
    {
      "epoch": 6.147798742138365,
      "grad_norm": 1.0948214530944824,
      "learning_rate": 4.373177055109922e-05,
      "loss": 0.1676,
      "num_input_tokens_seen": 7657544,
      "step": 11730
    },
    {
      "epoch": 6.15041928721174,
      "grad_norm": 1.925729751586914,
      "learning_rate": 4.3724196080048896e-05,
      "loss": 0.3453,
      "num_input_tokens_seen": 7660936,
      "step": 11735
    },
    {
      "epoch": 6.153039832285115,
      "grad_norm": 1.746381163597107,
      "learning_rate": 4.3716617692110364e-05,
      "loss": 0.2836,
      "num_input_tokens_seen": 7663336,
      "step": 11740
    },
    {
      "epoch": 6.15566037735849,
      "grad_norm": 0.9144853949546814,
      "learning_rate": 4.3709035388868943e-05,
      "loss": 0.2569,
      "num_input_tokens_seen": 7666312,
      "step": 11745
    },
    {
      "epoch": 6.1582809224318655,
      "grad_norm": 1.2236599922180176,
      "learning_rate": 4.3701449171910767e-05,
      "loss": 0.3123,
      "num_input_tokens_seen": 7669224,
      "step": 11750
    },
    {
      "epoch": 6.160901467505241,
      "grad_norm": 1.7369624376296997,
      "learning_rate": 4.3693859042822774e-05,
      "loss": 0.2396,
      "num_input_tokens_seen": 7672264,
      "step": 11755
    },
    {
      "epoch": 6.163522012578617,
      "grad_norm": 2.738693952560425,
      "learning_rate": 4.368626500319275e-05,
      "loss": 0.3193,
      "num_input_tokens_seen": 7674824,
      "step": 11760
    },
    {
      "epoch": 6.166142557651992,
      "grad_norm": 1.6683809757232666,
      "learning_rate": 4.367866705460927e-05,
      "loss": 0.1965,
      "num_input_tokens_seen": 7677608,
      "step": 11765
    },
    {
      "epoch": 6.168763102725367,
      "grad_norm": 1.3798167705535889,
      "learning_rate": 4.367106519866173e-05,
      "loss": 0.2809,
      "num_input_tokens_seen": 7680680,
      "step": 11770
    },
    {
      "epoch": 6.171383647798742,
      "grad_norm": 1.7549785375595093,
      "learning_rate": 4.366345943694037e-05,
      "loss": 0.2947,
      "num_input_tokens_seen": 7684072,
      "step": 11775
    },
    {
      "epoch": 6.174004192872117,
      "grad_norm": 0.9754730463027954,
      "learning_rate": 4.365584977103622e-05,
      "loss": 0.2752,
      "num_input_tokens_seen": 7687624,
      "step": 11780
    },
    {
      "epoch": 6.176624737945493,
      "grad_norm": 1.2617613077163696,
      "learning_rate": 4.364823620254113e-05,
      "loss": 0.353,
      "num_input_tokens_seen": 7691656,
      "step": 11785
    },
    {
      "epoch": 6.179245283018868,
      "grad_norm": 1.4952903985977173,
      "learning_rate": 4.364061873304778e-05,
      "loss": 0.283,
      "num_input_tokens_seen": 7694568,
      "step": 11790
    },
    {
      "epoch": 6.181865828092243,
      "grad_norm": 0.9075688123703003,
      "learning_rate": 4.363299736414965e-05,
      "loss": 0.243,
      "num_input_tokens_seen": 7697544,
      "step": 11795
    },
    {
      "epoch": 6.184486373165618,
      "grad_norm": 2.4815845489501953,
      "learning_rate": 4.3625372097441045e-05,
      "loss": 0.5004,
      "num_input_tokens_seen": 7700456,
      "step": 11800
    },
    {
      "epoch": 6.187106918238993,
      "grad_norm": 1.537275791168213,
      "learning_rate": 4.3617742934517094e-05,
      "loss": 0.2499,
      "num_input_tokens_seen": 7703112,
      "step": 11805
    },
    {
      "epoch": 6.189727463312369,
      "grad_norm": 2.9678750038146973,
      "learning_rate": 4.361010987697373e-05,
      "loss": 0.2328,
      "num_input_tokens_seen": 7705992,
      "step": 11810
    },
    {
      "epoch": 6.1923480083857445,
      "grad_norm": 0.9947319626808167,
      "learning_rate": 4.360247292640768e-05,
      "loss": 0.2719,
      "num_input_tokens_seen": 7709576,
      "step": 11815
    },
    {
      "epoch": 6.19496855345912,
      "grad_norm": 1.57572603225708,
      "learning_rate": 4.359483208441653e-05,
      "loss": 0.2712,
      "num_input_tokens_seen": 7712712,
      "step": 11820
    },
    {
      "epoch": 6.197589098532495,
      "grad_norm": 2.572828769683838,
      "learning_rate": 4.3587187352598646e-05,
      "loss": 0.2514,
      "num_input_tokens_seen": 7715432,
      "step": 11825
    },
    {
      "epoch": 6.20020964360587,
      "grad_norm": 2.0652804374694824,
      "learning_rate": 4.357953873255323e-05,
      "loss": 0.2199,
      "num_input_tokens_seen": 7718216,
      "step": 11830
    },
    {
      "epoch": 6.202830188679245,
      "grad_norm": 1.3605382442474365,
      "learning_rate": 4.357188622588027e-05,
      "loss": 0.2737,
      "num_input_tokens_seen": 7721096,
      "step": 11835
    },
    {
      "epoch": 6.20545073375262,
      "grad_norm": 1.4643765687942505,
      "learning_rate": 4.356422983418059e-05,
      "loss": 0.2722,
      "num_input_tokens_seen": 7724264,
      "step": 11840
    },
    {
      "epoch": 6.2080712788259955,
      "grad_norm": 1.5872061252593994,
      "learning_rate": 4.3556569559055824e-05,
      "loss": 0.2349,
      "num_input_tokens_seen": 7727176,
      "step": 11845
    },
    {
      "epoch": 6.210691823899371,
      "grad_norm": 1.3892039060592651,
      "learning_rate": 4.3548905402108406e-05,
      "loss": 0.2818,
      "num_input_tokens_seen": 7729544,
      "step": 11850
    },
    {
      "epoch": 6.213312368972747,
      "grad_norm": 1.0951178073883057,
      "learning_rate": 4.35412373649416e-05,
      "loss": 0.3428,
      "num_input_tokens_seen": 7733288,
      "step": 11855
    },
    {
      "epoch": 6.215932914046122,
      "grad_norm": 1.7382709980010986,
      "learning_rate": 4.353356544915947e-05,
      "loss": 0.2456,
      "num_input_tokens_seen": 7737096,
      "step": 11860
    },
    {
      "epoch": 6.218553459119497,
      "grad_norm": 2.3427157402038574,
      "learning_rate": 4.352588965636689e-05,
      "loss": 0.2659,
      "num_input_tokens_seen": 7741512,
      "step": 11865
    },
    {
      "epoch": 6.221174004192872,
      "grad_norm": 1.3677173852920532,
      "learning_rate": 4.351820998816954e-05,
      "loss": 0.251,
      "num_input_tokens_seen": 7744648,
      "step": 11870
    },
    {
      "epoch": 6.223794549266247,
      "grad_norm": 1.931267499923706,
      "learning_rate": 4.3510526446173935e-05,
      "loss": 0.2101,
      "num_input_tokens_seen": 7748392,
      "step": 11875
    },
    {
      "epoch": 6.226415094339623,
      "grad_norm": 1.1974358558654785,
      "learning_rate": 4.350283903198737e-05,
      "loss": 0.2209,
      "num_input_tokens_seen": 7752360,
      "step": 11880
    },
    {
      "epoch": 6.229035639412998,
      "grad_norm": 2.0612246990203857,
      "learning_rate": 4.349514774721797e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 7755016,
      "step": 11885
    },
    {
      "epoch": 6.231656184486373,
      "grad_norm": 1.7673709392547607,
      "learning_rate": 4.348745259347467e-05,
      "loss": 0.31,
      "num_input_tokens_seen": 7758120,
      "step": 11890
    },
    {
      "epoch": 6.234276729559748,
      "grad_norm": 1.6764895915985107,
      "learning_rate": 4.347975357236721e-05,
      "loss": 0.3484,
      "num_input_tokens_seen": 7761032,
      "step": 11895
    },
    {
      "epoch": 6.236897274633123,
      "grad_norm": 1.526545763015747,
      "learning_rate": 4.347205068550612e-05,
      "loss": 0.2303,
      "num_input_tokens_seen": 7763240,
      "step": 11900
    },
    {
      "epoch": 6.239517819706499,
      "grad_norm": 1.709629774093628,
      "learning_rate": 4.3464343934502774e-05,
      "loss": 0.2431,
      "num_input_tokens_seen": 7766664,
      "step": 11905
    },
    {
      "epoch": 6.2421383647798745,
      "grad_norm": 1.659745216369629,
      "learning_rate": 4.345663332096932e-05,
      "loss": 0.3223,
      "num_input_tokens_seen": 7769448,
      "step": 11910
    },
    {
      "epoch": 6.24475890985325,
      "grad_norm": 2.1329150199890137,
      "learning_rate": 4.344891884651875e-05,
      "loss": 0.339,
      "num_input_tokens_seen": 7772296,
      "step": 11915
    },
    {
      "epoch": 6.247379454926625,
      "grad_norm": 1.7812100648880005,
      "learning_rate": 4.344120051276483e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 7775912,
      "step": 11920
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.8926135897636414,
      "learning_rate": 4.3433478321322155e-05,
      "loss": 0.3132,
      "num_input_tokens_seen": 7779784,
      "step": 11925
    },
    {
      "epoch": 6.252620545073375,
      "grad_norm": 1.7783867120742798,
      "learning_rate": 4.3425752273806106e-05,
      "loss": 0.2136,
      "num_input_tokens_seen": 7782504,
      "step": 11930
    },
    {
      "epoch": 6.25524109014675,
      "grad_norm": 3.780365228652954,
      "learning_rate": 4.34180223718329e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 7785864,
      "step": 11935
    },
    {
      "epoch": 6.2578616352201255,
      "grad_norm": 2.027233600616455,
      "learning_rate": 4.341028861701953e-05,
      "loss": 0.2428,
      "num_input_tokens_seen": 7788808,
      "step": 11940
    },
    {
      "epoch": 6.260482180293501,
      "grad_norm": 1.7229104042053223,
      "learning_rate": 4.3402551010983825e-05,
      "loss": 0.2566,
      "num_input_tokens_seen": 7792232,
      "step": 11945
    },
    {
      "epoch": 6.263102725366876,
      "grad_norm": 1.2692128419876099,
      "learning_rate": 4.33948095553444e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 7794856,
      "step": 11950
    },
    {
      "epoch": 6.265723270440252,
      "grad_norm": 1.8257763385772705,
      "learning_rate": 4.338706425172067e-05,
      "loss": 0.3608,
      "num_input_tokens_seen": 7797800,
      "step": 11955
    },
    {
      "epoch": 6.268343815513627,
      "grad_norm": 1.187098503112793,
      "learning_rate": 4.3379315101732866e-05,
      "loss": 0.2096,
      "num_input_tokens_seen": 7801256,
      "step": 11960
    },
    {
      "epoch": 6.270964360587002,
      "grad_norm": 2.8222897052764893,
      "learning_rate": 4.3371562107002026e-05,
      "loss": 0.2337,
      "num_input_tokens_seen": 7804296,
      "step": 11965
    },
    {
      "epoch": 6.273584905660377,
      "grad_norm": 1.4065907001495361,
      "learning_rate": 4.3363805269149984e-05,
      "loss": 0.2467,
      "num_input_tokens_seen": 7807464,
      "step": 11970
    },
    {
      "epoch": 6.276205450733753,
      "grad_norm": 0.9274473190307617,
      "learning_rate": 4.335604458979939e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 7812008,
      "step": 11975
    },
    {
      "epoch": 6.278825995807128,
      "grad_norm": 2.050987958908081,
      "learning_rate": 4.3348280070573686e-05,
      "loss": 0.2339,
      "num_input_tokens_seen": 7815336,
      "step": 11980
    },
    {
      "epoch": 6.281446540880503,
      "grad_norm": 0.9690232276916504,
      "learning_rate": 4.334051171309712e-05,
      "loss": 0.1897,
      "num_input_tokens_seen": 7818856,
      "step": 11985
    },
    {
      "epoch": 6.284067085953878,
      "grad_norm": 1.5148372650146484,
      "learning_rate": 4.333273951899474e-05,
      "loss": 0.328,
      "num_input_tokens_seen": 7822024,
      "step": 11990
    },
    {
      "epoch": 6.286687631027253,
      "grad_norm": 1.4238110780715942,
      "learning_rate": 4.33249634898924e-05,
      "loss": 0.3479,
      "num_input_tokens_seen": 7825640,
      "step": 11995
    },
    {
      "epoch": 6.289308176100629,
      "grad_norm": 1.0785630941390991,
      "learning_rate": 4.331718362741677e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 7827944,
      "step": 12000
    },
    {
      "epoch": 6.2919287211740045,
      "grad_norm": 1.3367255926132202,
      "learning_rate": 4.330939993319529e-05,
      "loss": 0.2929,
      "num_input_tokens_seen": 7830600,
      "step": 12005
    },
    {
      "epoch": 6.29454926624738,
      "grad_norm": 0.8205295205116272,
      "learning_rate": 4.330161240885623e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 7834056,
      "step": 12010
    },
    {
      "epoch": 6.297169811320755,
      "grad_norm": 2.026334762573242,
      "learning_rate": 4.3293821056028654e-05,
      "loss": 0.2764,
      "num_input_tokens_seen": 7836904,
      "step": 12015
    },
    {
      "epoch": 6.29979035639413,
      "grad_norm": 2.1078715324401855,
      "learning_rate": 4.3286025876342415e-05,
      "loss": 0.2747,
      "num_input_tokens_seen": 7839880,
      "step": 12020
    },
    {
      "epoch": 6.302410901467505,
      "grad_norm": 1.503339409828186,
      "learning_rate": 4.327822687142819e-05,
      "loss": 0.278,
      "num_input_tokens_seen": 7843432,
      "step": 12025
    },
    {
      "epoch": 6.30503144654088,
      "grad_norm": 1.6843677759170532,
      "learning_rate": 4.327042404291742e-05,
      "loss": 0.2388,
      "num_input_tokens_seen": 7846216,
      "step": 12030
    },
    {
      "epoch": 6.3076519916142555,
      "grad_norm": 1.4855704307556152,
      "learning_rate": 4.326261739244239e-05,
      "loss": 0.2962,
      "num_input_tokens_seen": 7849544,
      "step": 12035
    },
    {
      "epoch": 6.310272536687631,
      "grad_norm": 1.723224401473999,
      "learning_rate": 4.325480692163615e-05,
      "loss": 0.2827,
      "num_input_tokens_seen": 7852552,
      "step": 12040
    },
    {
      "epoch": 6.312893081761006,
      "grad_norm": 2.119061231613159,
      "learning_rate": 4.324699263213256e-05,
      "loss": 0.3058,
      "num_input_tokens_seen": 7855336,
      "step": 12045
    },
    {
      "epoch": 6.315513626834382,
      "grad_norm": 1.3349932432174683,
      "learning_rate": 4.3239174525566285e-05,
      "loss": 0.2963,
      "num_input_tokens_seen": 7858184,
      "step": 12050
    },
    {
      "epoch": 6.318134171907757,
      "grad_norm": 3.0146260261535645,
      "learning_rate": 4.3231352603572784e-05,
      "loss": 0.247,
      "num_input_tokens_seen": 7861192,
      "step": 12055
    },
    {
      "epoch": 6.320754716981132,
      "grad_norm": 1.4994943141937256,
      "learning_rate": 4.322352686778832e-05,
      "loss": 0.2907,
      "num_input_tokens_seen": 7864520,
      "step": 12060
    },
    {
      "epoch": 6.323375262054507,
      "grad_norm": 1.64491605758667,
      "learning_rate": 4.3215697319849926e-05,
      "loss": 0.3099,
      "num_input_tokens_seen": 7867240,
      "step": 12065
    },
    {
      "epoch": 6.325995807127883,
      "grad_norm": 2.1452391147613525,
      "learning_rate": 4.3207863961395476e-05,
      "loss": 0.2985,
      "num_input_tokens_seen": 7870120,
      "step": 12070
    },
    {
      "epoch": 6.328616352201258,
      "grad_norm": 1.8821187019348145,
      "learning_rate": 4.320002679406361e-05,
      "loss": 0.2313,
      "num_input_tokens_seen": 7873736,
      "step": 12075
    },
    {
      "epoch": 6.331236897274633,
      "grad_norm": 1.530409812927246,
      "learning_rate": 4.319218581949378e-05,
      "loss": 0.2941,
      "num_input_tokens_seen": 7877000,
      "step": 12080
    },
    {
      "epoch": 6.333857442348008,
      "grad_norm": 1.6310795545578003,
      "learning_rate": 4.318434103932622e-05,
      "loss": 0.179,
      "num_input_tokens_seen": 7879784,
      "step": 12085
    },
    {
      "epoch": 6.336477987421383,
      "grad_norm": 2.4733681678771973,
      "learning_rate": 4.317649245520197e-05,
      "loss": 0.2337,
      "num_input_tokens_seen": 7882792,
      "step": 12090
    },
    {
      "epoch": 6.339098532494759,
      "grad_norm": 2.0909957885742188,
      "learning_rate": 4.3168640068762875e-05,
      "loss": 0.3081,
      "num_input_tokens_seen": 7885416,
      "step": 12095
    },
    {
      "epoch": 6.3417190775681345,
      "grad_norm": 1.363242506980896,
      "learning_rate": 4.316078388165155e-05,
      "loss": 0.285,
      "num_input_tokens_seen": 7888936,
      "step": 12100
    },
    {
      "epoch": 6.34433962264151,
      "grad_norm": 1.5521645545959473,
      "learning_rate": 4.315292389551143e-05,
      "loss": 0.3356,
      "num_input_tokens_seen": 7891976,
      "step": 12105
    },
    {
      "epoch": 6.346960167714885,
      "grad_norm": 4.015725612640381,
      "learning_rate": 4.3145060111986726e-05,
      "loss": 0.3222,
      "num_input_tokens_seen": 7895848,
      "step": 12110
    },
    {
      "epoch": 6.34958071278826,
      "grad_norm": 1.5559990406036377,
      "learning_rate": 4.313719253272246e-05,
      "loss": 0.3689,
      "num_input_tokens_seen": 7898760,
      "step": 12115
    },
    {
      "epoch": 6.352201257861635,
      "grad_norm": 1.3843885660171509,
      "learning_rate": 4.3129321159364436e-05,
      "loss": 0.2668,
      "num_input_tokens_seen": 7902120,
      "step": 12120
    },
    {
      "epoch": 6.35482180293501,
      "grad_norm": 3.7650935649871826,
      "learning_rate": 4.312144599355925e-05,
      "loss": 0.4266,
      "num_input_tokens_seen": 7905224,
      "step": 12125
    },
    {
      "epoch": 6.3574423480083855,
      "grad_norm": 1.5660909414291382,
      "learning_rate": 4.31135670369543e-05,
      "loss": 0.3278,
      "num_input_tokens_seen": 7908424,
      "step": 12130
    },
    {
      "epoch": 6.360062893081761,
      "grad_norm": 0.8470793962478638,
      "learning_rate": 4.310568429119779e-05,
      "loss": 0.2627,
      "num_input_tokens_seen": 7911336,
      "step": 12135
    },
    {
      "epoch": 6.362683438155136,
      "grad_norm": 1.1071510314941406,
      "learning_rate": 4.309779775793868e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 7913704,
      "step": 12140
    },
    {
      "epoch": 6.365303983228512,
      "grad_norm": 1.89142906665802,
      "learning_rate": 4.308990743882675e-05,
      "loss": 0.2886,
      "num_input_tokens_seen": 7916584,
      "step": 12145
    },
    {
      "epoch": 6.367924528301887,
      "grad_norm": 1.374040126800537,
      "learning_rate": 4.308201333551256e-05,
      "loss": 0.2887,
      "num_input_tokens_seen": 7919880,
      "step": 12150
    },
    {
      "epoch": 6.370545073375262,
      "grad_norm": 3.4218852519989014,
      "learning_rate": 4.3074115449647476e-05,
      "loss": 0.3066,
      "num_input_tokens_seen": 7923752,
      "step": 12155
    },
    {
      "epoch": 6.373165618448637,
      "grad_norm": 1.2716587781906128,
      "learning_rate": 4.306621378288364e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 7927464,
      "step": 12160
    },
    {
      "epoch": 6.3757861635220126,
      "grad_norm": 1.3284531831741333,
      "learning_rate": 4.305830833687399e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 7930792,
      "step": 12165
    },
    {
      "epoch": 6.378406708595388,
      "grad_norm": 1.113320231437683,
      "learning_rate": 4.305039911327225e-05,
      "loss": 0.3156,
      "num_input_tokens_seen": 7935112,
      "step": 12170
    },
    {
      "epoch": 6.381027253668763,
      "grad_norm": 1.3982548713684082,
      "learning_rate": 4.304248611373296e-05,
      "loss": 0.2676,
      "num_input_tokens_seen": 7938152,
      "step": 12175
    },
    {
      "epoch": 6.383647798742138,
      "grad_norm": 0.9729936122894287,
      "learning_rate": 4.30345693399114e-05,
      "loss": 0.2802,
      "num_input_tokens_seen": 7942536,
      "step": 12180
    },
    {
      "epoch": 6.386268343815513,
      "grad_norm": 1.3694777488708496,
      "learning_rate": 4.30266487934637e-05,
      "loss": 0.2054,
      "num_input_tokens_seen": 7945224,
      "step": 12185
    },
    {
      "epoch": 6.388888888888889,
      "grad_norm": 1.100031852722168,
      "learning_rate": 4.3018724476046714e-05,
      "loss": 0.238,
      "num_input_tokens_seen": 7948520,
      "step": 12190
    },
    {
      "epoch": 6.3915094339622645,
      "grad_norm": 2.078993797302246,
      "learning_rate": 4.301079638931814e-05,
      "loss": 0.3246,
      "num_input_tokens_seen": 7951368,
      "step": 12195
    },
    {
      "epoch": 6.39412997903564,
      "grad_norm": 1.4539856910705566,
      "learning_rate": 4.3002864534936444e-05,
      "loss": 0.3638,
      "num_input_tokens_seen": 7954472,
      "step": 12200
    },
    {
      "epoch": 6.396750524109015,
      "grad_norm": 1.7024867534637451,
      "learning_rate": 4.2994928914560874e-05,
      "loss": 0.2468,
      "num_input_tokens_seen": 7957000,
      "step": 12205
    },
    {
      "epoch": 6.39937106918239,
      "grad_norm": 1.4686390161514282,
      "learning_rate": 4.298698952985147e-05,
      "loss": 0.2237,
      "num_input_tokens_seen": 7959400,
      "step": 12210
    },
    {
      "epoch": 6.401991614255765,
      "grad_norm": 5.15472936630249,
      "learning_rate": 4.297904638246906e-05,
      "loss": 0.3323,
      "num_input_tokens_seen": 7962312,
      "step": 12215
    },
    {
      "epoch": 6.40461215932914,
      "grad_norm": 2.076664924621582,
      "learning_rate": 4.297109947407527e-05,
      "loss": 0.3312,
      "num_input_tokens_seen": 7965768,
      "step": 12220
    },
    {
      "epoch": 6.4072327044025155,
      "grad_norm": 1.1486055850982666,
      "learning_rate": 4.2963148806332486e-05,
      "loss": 0.2283,
      "num_input_tokens_seen": 7972840,
      "step": 12225
    },
    {
      "epoch": 6.409853249475891,
      "grad_norm": 1.9943296909332275,
      "learning_rate": 4.2955194380903915e-05,
      "loss": 0.2949,
      "num_input_tokens_seen": 7975304,
      "step": 12230
    },
    {
      "epoch": 6.412473794549266,
      "grad_norm": 1.2462079524993896,
      "learning_rate": 4.2947236199453524e-05,
      "loss": 0.23,
      "num_input_tokens_seen": 7977928,
      "step": 12235
    },
    {
      "epoch": 6.415094339622642,
      "grad_norm": 1.8466261625289917,
      "learning_rate": 4.2939274263646065e-05,
      "loss": 0.2683,
      "num_input_tokens_seen": 7982120,
      "step": 12240
    },
    {
      "epoch": 6.417714884696017,
      "grad_norm": 1.8584386110305786,
      "learning_rate": 4.29313085751471e-05,
      "loss": 0.2714,
      "num_input_tokens_seen": 7985032,
      "step": 12245
    },
    {
      "epoch": 6.420335429769392,
      "grad_norm": 2.0092668533325195,
      "learning_rate": 4.292333913562294e-05,
      "loss": 0.2457,
      "num_input_tokens_seen": 7987272,
      "step": 12250
    },
    {
      "epoch": 6.422955974842767,
      "grad_norm": 1.2220025062561035,
      "learning_rate": 4.291536594674073e-05,
      "loss": 0.3235,
      "num_input_tokens_seen": 7990152,
      "step": 12255
    },
    {
      "epoch": 6.4255765199161425,
      "grad_norm": 1.8166441917419434,
      "learning_rate": 4.2907389010168344e-05,
      "loss": 0.3062,
      "num_input_tokens_seen": 7992648,
      "step": 12260
    },
    {
      "epoch": 6.428197064989518,
      "grad_norm": 1.417116403579712,
      "learning_rate": 4.2899408327574475e-05,
      "loss": 0.2359,
      "num_input_tokens_seen": 7995880,
      "step": 12265
    },
    {
      "epoch": 6.430817610062893,
      "grad_norm": 1.470475673675537,
      "learning_rate": 4.289142390062859e-05,
      "loss": 0.2601,
      "num_input_tokens_seen": 7998312,
      "step": 12270
    },
    {
      "epoch": 6.433438155136268,
      "grad_norm": 1.3367140293121338,
      "learning_rate": 4.2883435731000946e-05,
      "loss": 0.2834,
      "num_input_tokens_seen": 8001768,
      "step": 12275
    },
    {
      "epoch": 6.436058700209643,
      "grad_norm": 1.7765475511550903,
      "learning_rate": 4.287544382036256e-05,
      "loss": 0.2373,
      "num_input_tokens_seen": 8004392,
      "step": 12280
    },
    {
      "epoch": 6.438679245283019,
      "grad_norm": 1.355229139328003,
      "learning_rate": 4.286744817038527e-05,
      "loss": 0.2516,
      "num_input_tokens_seen": 8006984,
      "step": 12285
    },
    {
      "epoch": 6.441299790356394,
      "grad_norm": 1.8480072021484375,
      "learning_rate": 4.2859448782741666e-05,
      "loss": 0.4069,
      "num_input_tokens_seen": 8010920,
      "step": 12290
    },
    {
      "epoch": 6.44392033542977,
      "grad_norm": 1.3797364234924316,
      "learning_rate": 4.2851445659105126e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 8013800,
      "step": 12295
    },
    {
      "epoch": 6.446540880503145,
      "grad_norm": 1.3224233388900757,
      "learning_rate": 4.284343880114982e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 8016616,
      "step": 12300
    },
    {
      "epoch": 6.44916142557652,
      "grad_norm": 0.9279593229293823,
      "learning_rate": 4.2835428210550676e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 8020040,
      "step": 12305
    },
    {
      "epoch": 6.451781970649895,
      "grad_norm": 1.1808990240097046,
      "learning_rate": 4.282741388898344e-05,
      "loss": 0.2291,
      "num_input_tokens_seen": 8024008,
      "step": 12310
    },
    {
      "epoch": 6.45440251572327,
      "grad_norm": 1.7261543273925781,
      "learning_rate": 4.281939583812459e-05,
      "loss": 0.289,
      "num_input_tokens_seen": 8027528,
      "step": 12315
    },
    {
      "epoch": 6.4570230607966455,
      "grad_norm": 1.7068586349487305,
      "learning_rate": 4.281137405965143e-05,
      "loss": 0.3064,
      "num_input_tokens_seen": 8030312,
      "step": 12320
    },
    {
      "epoch": 6.459643605870021,
      "grad_norm": 1.9658271074295044,
      "learning_rate": 4.280334855524203e-05,
      "loss": 0.2288,
      "num_input_tokens_seen": 8034280,
      "step": 12325
    },
    {
      "epoch": 6.462264150943396,
      "grad_norm": 1.5543782711029053,
      "learning_rate": 4.279531932657521e-05,
      "loss": 0.2947,
      "num_input_tokens_seen": 8038408,
      "step": 12330
    },
    {
      "epoch": 6.464884696016772,
      "grad_norm": 4.626752853393555,
      "learning_rate": 4.278728637533062e-05,
      "loss": 0.3053,
      "num_input_tokens_seen": 8041544,
      "step": 12335
    },
    {
      "epoch": 6.467505241090147,
      "grad_norm": 1.5837608575820923,
      "learning_rate": 4.2779249703188626e-05,
      "loss": 0.1847,
      "num_input_tokens_seen": 8045096,
      "step": 12340
    },
    {
      "epoch": 6.470125786163522,
      "grad_norm": 1.0021919012069702,
      "learning_rate": 4.277120931183045e-05,
      "loss": 0.2726,
      "num_input_tokens_seen": 8049064,
      "step": 12345
    },
    {
      "epoch": 6.472746331236897,
      "grad_norm": 1.670751690864563,
      "learning_rate": 4.276316520293803e-05,
      "loss": 0.3007,
      "num_input_tokens_seen": 8051816,
      "step": 12350
    },
    {
      "epoch": 6.4753668763102725,
      "grad_norm": 1.4664561748504639,
      "learning_rate": 4.27551173781941e-05,
      "loss": 0.2642,
      "num_input_tokens_seen": 8055528,
      "step": 12355
    },
    {
      "epoch": 6.477987421383648,
      "grad_norm": 1.1349947452545166,
      "learning_rate": 4.274706583928216e-05,
      "loss": 0.2545,
      "num_input_tokens_seen": 8058472,
      "step": 12360
    },
    {
      "epoch": 6.480607966457023,
      "grad_norm": 1.5857595205307007,
      "learning_rate": 4.273901058788654e-05,
      "loss": 0.4326,
      "num_input_tokens_seen": 8061608,
      "step": 12365
    },
    {
      "epoch": 6.483228511530398,
      "grad_norm": 0.9425234198570251,
      "learning_rate": 4.2730951625692275e-05,
      "loss": 0.303,
      "num_input_tokens_seen": 8064648,
      "step": 12370
    },
    {
      "epoch": 6.485849056603773,
      "grad_norm": 1.419378399848938,
      "learning_rate": 4.27228889543852e-05,
      "loss": 0.3746,
      "num_input_tokens_seen": 8067496,
      "step": 12375
    },
    {
      "epoch": 6.488469601677149,
      "grad_norm": 1.6711561679840088,
      "learning_rate": 4.2714822575651966e-05,
      "loss": 0.24,
      "num_input_tokens_seen": 8069864,
      "step": 12380
    },
    {
      "epoch": 6.491090146750524,
      "grad_norm": 1.8333756923675537,
      "learning_rate": 4.270675249117994e-05,
      "loss": 0.3244,
      "num_input_tokens_seen": 8072936,
      "step": 12385
    },
    {
      "epoch": 6.4937106918239,
      "grad_norm": 2.963012456893921,
      "learning_rate": 4.269867870265731e-05,
      "loss": 0.3089,
      "num_input_tokens_seen": 8075784,
      "step": 12390
    },
    {
      "epoch": 6.496331236897275,
      "grad_norm": 2.8861918449401855,
      "learning_rate": 4.269060121177301e-05,
      "loss": 0.3411,
      "num_input_tokens_seen": 8078280,
      "step": 12395
    },
    {
      "epoch": 6.49895178197065,
      "grad_norm": 2.6276702880859375,
      "learning_rate": 4.2682520020216764e-05,
      "loss": 0.2094,
      "num_input_tokens_seen": 8081992,
      "step": 12400
    },
    {
      "epoch": 6.501572327044025,
      "grad_norm": 1.467923879623413,
      "learning_rate": 4.267443512967906e-05,
      "loss": 0.2697,
      "num_input_tokens_seen": 8085672,
      "step": 12405
    },
    {
      "epoch": 6.5041928721174,
      "grad_norm": 1.443921446800232,
      "learning_rate": 4.2666346541851165e-05,
      "loss": 0.2394,
      "num_input_tokens_seen": 8089000,
      "step": 12410
    },
    {
      "epoch": 6.506813417190775,
      "grad_norm": 1.2869291305541992,
      "learning_rate": 4.265825425842512e-05,
      "loss": 0.2355,
      "num_input_tokens_seen": 8092168,
      "step": 12415
    },
    {
      "epoch": 6.509433962264151,
      "grad_norm": 1.2564918994903564,
      "learning_rate": 4.265015828109375e-05,
      "loss": 0.3175,
      "num_input_tokens_seen": 8094728,
      "step": 12420
    },
    {
      "epoch": 6.512054507337526,
      "grad_norm": 2.2132208347320557,
      "learning_rate": 4.264205861155063e-05,
      "loss": 0.3051,
      "num_input_tokens_seen": 8098088,
      "step": 12425
    },
    {
      "epoch": 6.514675052410902,
      "grad_norm": 1.6050896644592285,
      "learning_rate": 4.263395525149011e-05,
      "loss": 0.3324,
      "num_input_tokens_seen": 8100904,
      "step": 12430
    },
    {
      "epoch": 6.517295597484277,
      "grad_norm": 1.4846627712249756,
      "learning_rate": 4.262584820260734e-05,
      "loss": 0.2336,
      "num_input_tokens_seen": 8104264,
      "step": 12435
    },
    {
      "epoch": 6.519916142557652,
      "grad_norm": 0.9688090085983276,
      "learning_rate": 4.261773746659821e-05,
      "loss": 0.2483,
      "num_input_tokens_seen": 8107144,
      "step": 12440
    },
    {
      "epoch": 6.522536687631027,
      "grad_norm": 1.6851344108581543,
      "learning_rate": 4.260962304515939e-05,
      "loss": 0.2234,
      "num_input_tokens_seen": 8111304,
      "step": 12445
    },
    {
      "epoch": 6.5251572327044025,
      "grad_norm": 1.2602648735046387,
      "learning_rate": 4.2601504939988335e-05,
      "loss": 0.3023,
      "num_input_tokens_seen": 8114056,
      "step": 12450
    },
    {
      "epoch": 6.527777777777778,
      "grad_norm": 1.8778568506240845,
      "learning_rate": 4.259338315278325e-05,
      "loss": 0.3511,
      "num_input_tokens_seen": 8116808,
      "step": 12455
    },
    {
      "epoch": 6.530398322851153,
      "grad_norm": 2.92879056930542,
      "learning_rate": 4.258525768524313e-05,
      "loss": 0.2647,
      "num_input_tokens_seen": 8119496,
      "step": 12460
    },
    {
      "epoch": 6.533018867924528,
      "grad_norm": 2.6445040702819824,
      "learning_rate": 4.2577128539067725e-05,
      "loss": 0.2927,
      "num_input_tokens_seen": 8122728,
      "step": 12465
    },
    {
      "epoch": 6.535639412997903,
      "grad_norm": 2.282378911972046,
      "learning_rate": 4.256899571595756e-05,
      "loss": 0.3815,
      "num_input_tokens_seen": 8126504,
      "step": 12470
    },
    {
      "epoch": 6.538259958071279,
      "grad_norm": 3.1743009090423584,
      "learning_rate": 4.256085921761393e-05,
      "loss": 0.352,
      "num_input_tokens_seen": 8128776,
      "step": 12475
    },
    {
      "epoch": 6.540880503144654,
      "grad_norm": 1.050147533416748,
      "learning_rate": 4.25527190457389e-05,
      "loss": 0.3517,
      "num_input_tokens_seen": 8131720,
      "step": 12480
    },
    {
      "epoch": 6.54350104821803,
      "grad_norm": 2.080209970474243,
      "learning_rate": 4.2544575202035285e-05,
      "loss": 0.3279,
      "num_input_tokens_seen": 8135624,
      "step": 12485
    },
    {
      "epoch": 6.546121593291405,
      "grad_norm": 1.0929621458053589,
      "learning_rate": 4.253642768820671e-05,
      "loss": 0.2536,
      "num_input_tokens_seen": 8138632,
      "step": 12490
    },
    {
      "epoch": 6.54874213836478,
      "grad_norm": 1.6435495615005493,
      "learning_rate": 4.252827650595752e-05,
      "loss": 0.2908,
      "num_input_tokens_seen": 8142184,
      "step": 12495
    },
    {
      "epoch": 6.551362683438155,
      "grad_norm": 1.2304885387420654,
      "learning_rate": 4.252012165699286e-05,
      "loss": 0.305,
      "num_input_tokens_seen": 8146792,
      "step": 12500
    },
    {
      "epoch": 6.55398322851153,
      "grad_norm": 1.4907995462417603,
      "learning_rate": 4.2511963143018634e-05,
      "loss": 0.2019,
      "num_input_tokens_seen": 8150504,
      "step": 12505
    },
    {
      "epoch": 6.556603773584905,
      "grad_norm": 3.3261849880218506,
      "learning_rate": 4.25038009657415e-05,
      "loss": 0.2865,
      "num_input_tokens_seen": 8153000,
      "step": 12510
    },
    {
      "epoch": 6.559224318658281,
      "grad_norm": 2.6134414672851562,
      "learning_rate": 4.24956351268689e-05,
      "loss": 0.1911,
      "num_input_tokens_seen": 8155720,
      "step": 12515
    },
    {
      "epoch": 6.561844863731656,
      "grad_norm": 1.7555924654006958,
      "learning_rate": 4.2487465628109036e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 8159400,
      "step": 12520
    },
    {
      "epoch": 6.564465408805032,
      "grad_norm": 1.4209052324295044,
      "learning_rate": 4.247929247117087e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 8162952,
      "step": 12525
    },
    {
      "epoch": 6.567085953878407,
      "grad_norm": 1.458818793296814,
      "learning_rate": 4.247111565776413e-05,
      "loss": 0.2676,
      "num_input_tokens_seen": 8166376,
      "step": 12530
    },
    {
      "epoch": 6.569706498951782,
      "grad_norm": 2.90069317817688,
      "learning_rate": 4.246293518959932e-05,
      "loss": 0.3302,
      "num_input_tokens_seen": 8169064,
      "step": 12535
    },
    {
      "epoch": 6.572327044025157,
      "grad_norm": 1.2638450860977173,
      "learning_rate": 4.245475106838769e-05,
      "loss": 0.3346,
      "num_input_tokens_seen": 8171976,
      "step": 12540
    },
    {
      "epoch": 6.5749475890985325,
      "grad_norm": 1.6022471189498901,
      "learning_rate": 4.2446563295841266e-05,
      "loss": 0.2664,
      "num_input_tokens_seen": 8176200,
      "step": 12545
    },
    {
      "epoch": 6.577568134171908,
      "grad_norm": 2.90366268157959,
      "learning_rate": 4.243837187367285e-05,
      "loss": 0.291,
      "num_input_tokens_seen": 8179368,
      "step": 12550
    },
    {
      "epoch": 6.580188679245283,
      "grad_norm": 1.9441763162612915,
      "learning_rate": 4.2430176803595987e-05,
      "loss": 0.2135,
      "num_input_tokens_seen": 8182248,
      "step": 12555
    },
    {
      "epoch": 6.582809224318658,
      "grad_norm": 1.2938474416732788,
      "learning_rate": 4.2421978087324985e-05,
      "loss": 0.3276,
      "num_input_tokens_seen": 8184776,
      "step": 12560
    },
    {
      "epoch": 6.585429769392033,
      "grad_norm": 1.511635661125183,
      "learning_rate": 4.241377572657493e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 8187976,
      "step": 12565
    },
    {
      "epoch": 6.588050314465409,
      "grad_norm": 1.5941429138183594,
      "learning_rate": 4.2405569723061654e-05,
      "loss": 0.2628,
      "num_input_tokens_seen": 8190984,
      "step": 12570
    },
    {
      "epoch": 6.590670859538784,
      "grad_norm": 1.5940167903900146,
      "learning_rate": 4.239736007850178e-05,
      "loss": 0.2283,
      "num_input_tokens_seen": 8194568,
      "step": 12575
    },
    {
      "epoch": 6.59329140461216,
      "grad_norm": 1.4322911500930786,
      "learning_rate": 4.238914679461264e-05,
      "loss": 0.3383,
      "num_input_tokens_seen": 8197512,
      "step": 12580
    },
    {
      "epoch": 6.595911949685535,
      "grad_norm": 1.4513955116271973,
      "learning_rate": 4.2380929873112386e-05,
      "loss": 0.2475,
      "num_input_tokens_seen": 8201128,
      "step": 12585
    },
    {
      "epoch": 6.59853249475891,
      "grad_norm": 2.7217676639556885,
      "learning_rate": 4.2372709315719896e-05,
      "loss": 0.1674,
      "num_input_tokens_seen": 8204392,
      "step": 12590
    },
    {
      "epoch": 6.601153039832285,
      "grad_norm": 1.849694013595581,
      "learning_rate": 4.2364485124154806e-05,
      "loss": 0.2608,
      "num_input_tokens_seen": 8206952,
      "step": 12595
    },
    {
      "epoch": 6.60377358490566,
      "grad_norm": 1.6386483907699585,
      "learning_rate": 4.235625730013754e-05,
      "loss": 0.1977,
      "num_input_tokens_seen": 8209768,
      "step": 12600
    },
    {
      "epoch": 6.606394129979035,
      "grad_norm": 1.3899013996124268,
      "learning_rate": 4.234802584538926e-05,
      "loss": 0.2855,
      "num_input_tokens_seen": 8216680,
      "step": 12605
    },
    {
      "epoch": 6.609014675052411,
      "grad_norm": 1.4764809608459473,
      "learning_rate": 4.2339790761631884e-05,
      "loss": 0.2606,
      "num_input_tokens_seen": 8220552,
      "step": 12610
    },
    {
      "epoch": 6.611635220125786,
      "grad_norm": 1.2671594619750977,
      "learning_rate": 4.233155205058811e-05,
      "loss": 0.3028,
      "num_input_tokens_seen": 8225064,
      "step": 12615
    },
    {
      "epoch": 6.614255765199162,
      "grad_norm": 0.9878057241439819,
      "learning_rate": 4.232330971398137e-05,
      "loss": 0.3221,
      "num_input_tokens_seen": 8228168,
      "step": 12620
    },
    {
      "epoch": 6.616876310272537,
      "grad_norm": 2.4151077270507812,
      "learning_rate": 4.2315063753535875e-05,
      "loss": 0.2227,
      "num_input_tokens_seen": 8231944,
      "step": 12625
    },
    {
      "epoch": 6.619496855345912,
      "grad_norm": 2.639868974685669,
      "learning_rate": 4.230681417097658e-05,
      "loss": 0.2899,
      "num_input_tokens_seen": 8235848,
      "step": 12630
    },
    {
      "epoch": 6.622117400419287,
      "grad_norm": 1.1458889245986938,
      "learning_rate": 4.229856096802921e-05,
      "loss": 0.3829,
      "num_input_tokens_seen": 8238792,
      "step": 12635
    },
    {
      "epoch": 6.6247379454926625,
      "grad_norm": 1.7714638710021973,
      "learning_rate": 4.229030414642024e-05,
      "loss": 0.2778,
      "num_input_tokens_seen": 8242088,
      "step": 12640
    },
    {
      "epoch": 6.627358490566038,
      "grad_norm": 1.7425589561462402,
      "learning_rate": 4.2282043707876896e-05,
      "loss": 0.3159,
      "num_input_tokens_seen": 8244616,
      "step": 12645
    },
    {
      "epoch": 6.629979035639413,
      "grad_norm": 3.8026349544525146,
      "learning_rate": 4.2273779654127175e-05,
      "loss": 0.2363,
      "num_input_tokens_seen": 8247912,
      "step": 12650
    },
    {
      "epoch": 6.632599580712788,
      "grad_norm": 1.5749609470367432,
      "learning_rate": 4.226551198689982e-05,
      "loss": 0.3073,
      "num_input_tokens_seen": 8251816,
      "step": 12655
    },
    {
      "epoch": 6.635220125786163,
      "grad_norm": 1.6417242288589478,
      "learning_rate": 4.225724070792434e-05,
      "loss": 0.3228,
      "num_input_tokens_seen": 8255432,
      "step": 12660
    },
    {
      "epoch": 6.637840670859539,
      "grad_norm": 0.9041414260864258,
      "learning_rate": 4.2248965818930975e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 8258824,
      "step": 12665
    },
    {
      "epoch": 6.640461215932914,
      "grad_norm": 1.526542067527771,
      "learning_rate": 4.224068732165075e-05,
      "loss": 0.3028,
      "num_input_tokens_seen": 8262888,
      "step": 12670
    },
    {
      "epoch": 6.6430817610062896,
      "grad_norm": 0.9339457154273987,
      "learning_rate": 4.223240521781543e-05,
      "loss": 0.2064,
      "num_input_tokens_seen": 8266568,
      "step": 12675
    },
    {
      "epoch": 6.645702306079665,
      "grad_norm": 2.0915753841400146,
      "learning_rate": 4.2224119509157535e-05,
      "loss": 0.3412,
      "num_input_tokens_seen": 8269800,
      "step": 12680
    },
    {
      "epoch": 6.64832285115304,
      "grad_norm": 1.277955412864685,
      "learning_rate": 4.221583019741033e-05,
      "loss": 0.2452,
      "num_input_tokens_seen": 8272936,
      "step": 12685
    },
    {
      "epoch": 6.650943396226415,
      "grad_norm": 1.7629319429397583,
      "learning_rate": 4.2207537284307876e-05,
      "loss": 0.2739,
      "num_input_tokens_seen": 8276104,
      "step": 12690
    },
    {
      "epoch": 6.65356394129979,
      "grad_norm": 0.894109308719635,
      "learning_rate": 4.219924077158493e-05,
      "loss": 0.2265,
      "num_input_tokens_seen": 8278728,
      "step": 12695
    },
    {
      "epoch": 6.656184486373165,
      "grad_norm": 1.3642524480819702,
      "learning_rate": 4.219094066097703e-05,
      "loss": 0.253,
      "num_input_tokens_seen": 8282312,
      "step": 12700
    },
    {
      "epoch": 6.658805031446541,
      "grad_norm": 1.6871901750564575,
      "learning_rate": 4.218263695422047e-05,
      "loss": 0.3002,
      "num_input_tokens_seen": 8285800,
      "step": 12705
    },
    {
      "epoch": 6.661425576519916,
      "grad_norm": 1.4869717359542847,
      "learning_rate": 4.217432965305229e-05,
      "loss": 0.3345,
      "num_input_tokens_seen": 8288648,
      "step": 12710
    },
    {
      "epoch": 6.664046121593291,
      "grad_norm": 1.6637070178985596,
      "learning_rate": 4.2166018759210286e-05,
      "loss": 0.2286,
      "num_input_tokens_seen": 8291688,
      "step": 12715
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 2.1998095512390137,
      "learning_rate": 4.215770427443299e-05,
      "loss": 0.301,
      "num_input_tokens_seen": 8295400,
      "step": 12720
    },
    {
      "epoch": 6.669287211740042,
      "grad_norm": 1.2079724073410034,
      "learning_rate": 4.214938620045971e-05,
      "loss": 0.2812,
      "num_input_tokens_seen": 8297704,
      "step": 12725
    },
    {
      "epoch": 6.671907756813417,
      "grad_norm": 2.095078706741333,
      "learning_rate": 4.214106453903048e-05,
      "loss": 0.3792,
      "num_input_tokens_seen": 8300808,
      "step": 12730
    },
    {
      "epoch": 6.6745283018867925,
      "grad_norm": 2.336210012435913,
      "learning_rate": 4.2132739291886114e-05,
      "loss": 0.3534,
      "num_input_tokens_seen": 8304104,
      "step": 12735
    },
    {
      "epoch": 6.677148846960168,
      "grad_norm": 2.5688564777374268,
      "learning_rate": 4.212441046076814e-05,
      "loss": 0.3173,
      "num_input_tokens_seen": 8307784,
      "step": 12740
    },
    {
      "epoch": 6.679769392033543,
      "grad_norm": 1.0210243463516235,
      "learning_rate": 4.211607804741887e-05,
      "loss": 0.2519,
      "num_input_tokens_seen": 8311848,
      "step": 12745
    },
    {
      "epoch": 6.682389937106918,
      "grad_norm": 1.8634892702102661,
      "learning_rate": 4.2107742053581344e-05,
      "loss": 0.255,
      "num_input_tokens_seen": 8315272,
      "step": 12750
    },
    {
      "epoch": 6.685010482180293,
      "grad_norm": 1.7090781927108765,
      "learning_rate": 4.209940248099935e-05,
      "loss": 0.2916,
      "num_input_tokens_seen": 8318376,
      "step": 12755
    },
    {
      "epoch": 6.687631027253669,
      "grad_norm": 2.0867257118225098,
      "learning_rate": 4.2091059331417445e-05,
      "loss": 0.2365,
      "num_input_tokens_seen": 8322248,
      "step": 12760
    },
    {
      "epoch": 6.690251572327044,
      "grad_norm": 0.8203480839729309,
      "learning_rate": 4.208271260658091e-05,
      "loss": 0.1782,
      "num_input_tokens_seen": 8326056,
      "step": 12765
    },
    {
      "epoch": 6.6928721174004195,
      "grad_norm": 1.417693018913269,
      "learning_rate": 4.2074362308235785e-05,
      "loss": 0.3362,
      "num_input_tokens_seen": 8329608,
      "step": 12770
    },
    {
      "epoch": 6.695492662473795,
      "grad_norm": 1.6425116062164307,
      "learning_rate": 4.206600843812886e-05,
      "loss": 0.2389,
      "num_input_tokens_seen": 8332072,
      "step": 12775
    },
    {
      "epoch": 6.69811320754717,
      "grad_norm": 2.226950168609619,
      "learning_rate": 4.205765099800767e-05,
      "loss": 0.2847,
      "num_input_tokens_seen": 8334760,
      "step": 12780
    },
    {
      "epoch": 6.700733752620545,
      "grad_norm": 1.8789070844650269,
      "learning_rate": 4.204928998962049e-05,
      "loss": 0.2125,
      "num_input_tokens_seen": 8337736,
      "step": 12785
    },
    {
      "epoch": 6.70335429769392,
      "grad_norm": 1.969480037689209,
      "learning_rate": 4.204092541471636e-05,
      "loss": 0.2106,
      "num_input_tokens_seen": 8340488,
      "step": 12790
    },
    {
      "epoch": 6.705974842767295,
      "grad_norm": 1.9055540561676025,
      "learning_rate": 4.203255727504504e-05,
      "loss": 0.2982,
      "num_input_tokens_seen": 8343176,
      "step": 12795
    },
    {
      "epoch": 6.7085953878406706,
      "grad_norm": 1.5746808052062988,
      "learning_rate": 4.202418557235706e-05,
      "loss": 0.3286,
      "num_input_tokens_seen": 8346376,
      "step": 12800
    },
    {
      "epoch": 6.711215932914046,
      "grad_norm": 1.171759009361267,
      "learning_rate": 4.201581030840368e-05,
      "loss": 0.2493,
      "num_input_tokens_seen": 8349672,
      "step": 12805
    },
    {
      "epoch": 6.713836477987421,
      "grad_norm": 1.4421346187591553,
      "learning_rate": 4.200743148493691e-05,
      "loss": 0.2627,
      "num_input_tokens_seen": 8352232,
      "step": 12810
    },
    {
      "epoch": 6.716457023060797,
      "grad_norm": 1.2456048727035522,
      "learning_rate": 4.199904910370951e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 8355592,
      "step": 12815
    },
    {
      "epoch": 6.719077568134172,
      "grad_norm": 1.3549059629440308,
      "learning_rate": 4.1990663166474964e-05,
      "loss": 0.1844,
      "num_input_tokens_seen": 8357736,
      "step": 12820
    },
    {
      "epoch": 6.721698113207547,
      "grad_norm": 1.6069929599761963,
      "learning_rate": 4.1982273674987524e-05,
      "loss": 0.2092,
      "num_input_tokens_seen": 8362568,
      "step": 12825
    },
    {
      "epoch": 6.7243186582809225,
      "grad_norm": 1.3635263442993164,
      "learning_rate": 4.197388063100218e-05,
      "loss": 0.3137,
      "num_input_tokens_seen": 8366152,
      "step": 12830
    },
    {
      "epoch": 6.726939203354298,
      "grad_norm": 2.338322639465332,
      "learning_rate": 4.196548403627465e-05,
      "loss": 0.248,
      "num_input_tokens_seen": 8368456,
      "step": 12835
    },
    {
      "epoch": 6.729559748427673,
      "grad_norm": 2.9612505435943604,
      "learning_rate": 4.195708389256142e-05,
      "loss": 0.2949,
      "num_input_tokens_seen": 8371560,
      "step": 12840
    },
    {
      "epoch": 6.732180293501048,
      "grad_norm": 2.2582666873931885,
      "learning_rate": 4.194868020161969e-05,
      "loss": 0.2416,
      "num_input_tokens_seen": 8374792,
      "step": 12845
    },
    {
      "epoch": 6.734800838574423,
      "grad_norm": 2.1213879585266113,
      "learning_rate": 4.1940272965207425e-05,
      "loss": 0.2761,
      "num_input_tokens_seen": 8377832,
      "step": 12850
    },
    {
      "epoch": 6.737421383647799,
      "grad_norm": 1.6384868621826172,
      "learning_rate": 4.1931862185083326e-05,
      "loss": 0.3178,
      "num_input_tokens_seen": 8380872,
      "step": 12855
    },
    {
      "epoch": 6.740041928721174,
      "grad_norm": 1.561309576034546,
      "learning_rate": 4.192344786300682e-05,
      "loss": 0.3055,
      "num_input_tokens_seen": 8384264,
      "step": 12860
    },
    {
      "epoch": 6.7426624737945495,
      "grad_norm": 1.1436091661453247,
      "learning_rate": 4.19150300007381e-05,
      "loss": 0.3096,
      "num_input_tokens_seen": 8389064,
      "step": 12865
    },
    {
      "epoch": 6.745283018867925,
      "grad_norm": 1.3049664497375488,
      "learning_rate": 4.190660860003806e-05,
      "loss": 0.3341,
      "num_input_tokens_seen": 8391976,
      "step": 12870
    },
    {
      "epoch": 6.7479035639413,
      "grad_norm": 1.1929517984390259,
      "learning_rate": 4.189818366266841e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 8395176,
      "step": 12875
    },
    {
      "epoch": 6.750524109014675,
      "grad_norm": 1.6493761539459229,
      "learning_rate": 4.188975519039151e-05,
      "loss": 0.3019,
      "num_input_tokens_seen": 8398216,
      "step": 12880
    },
    {
      "epoch": 6.75314465408805,
      "grad_norm": 1.3368233442306519,
      "learning_rate": 4.188132318497051e-05,
      "loss": 0.3194,
      "num_input_tokens_seen": 8402120,
      "step": 12885
    },
    {
      "epoch": 6.755765199161425,
      "grad_norm": 3.146425247192383,
      "learning_rate": 4.1872887648169304e-05,
      "loss": 0.3275,
      "num_input_tokens_seen": 8405736,
      "step": 12890
    },
    {
      "epoch": 6.7583857442348005,
      "grad_norm": 3.0682661533355713,
      "learning_rate": 4.186444858175248e-05,
      "loss": 0.2645,
      "num_input_tokens_seen": 8408808,
      "step": 12895
    },
    {
      "epoch": 6.761006289308176,
      "grad_norm": 0.840604841709137,
      "learning_rate": 4.1856005987485424e-05,
      "loss": 0.2001,
      "num_input_tokens_seen": 8411400,
      "step": 12900
    },
    {
      "epoch": 6.763626834381551,
      "grad_norm": 1.7012189626693726,
      "learning_rate": 4.1847559867134215e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 8414632,
      "step": 12905
    },
    {
      "epoch": 6.766247379454927,
      "grad_norm": 1.727198839187622,
      "learning_rate": 4.1839110222465693e-05,
      "loss": 0.2764,
      "num_input_tokens_seen": 8417256,
      "step": 12910
    },
    {
      "epoch": 6.768867924528302,
      "grad_norm": 1.0606211423873901,
      "learning_rate": 4.183065705524742e-05,
      "loss": 0.2552,
      "num_input_tokens_seen": 8420584,
      "step": 12915
    },
    {
      "epoch": 6.771488469601677,
      "grad_norm": 1.295638918876648,
      "learning_rate": 4.1822200367247696e-05,
      "loss": 0.3421,
      "num_input_tokens_seen": 8423912,
      "step": 12920
    },
    {
      "epoch": 6.774109014675052,
      "grad_norm": 2.1194097995758057,
      "learning_rate": 4.181374016023559e-05,
      "loss": 0.2434,
      "num_input_tokens_seen": 8427368,
      "step": 12925
    },
    {
      "epoch": 6.776729559748428,
      "grad_norm": 1.6543519496917725,
      "learning_rate": 4.1805276435980854e-05,
      "loss": 0.2819,
      "num_input_tokens_seen": 8431176,
      "step": 12930
    },
    {
      "epoch": 6.779350104821803,
      "grad_norm": 1.2499825954437256,
      "learning_rate": 4.1796809196254014e-05,
      "loss": 0.2668,
      "num_input_tokens_seen": 8433768,
      "step": 12935
    },
    {
      "epoch": 6.781970649895178,
      "grad_norm": 1.7534947395324707,
      "learning_rate": 4.178833844282632e-05,
      "loss": 0.2404,
      "num_input_tokens_seen": 8437384,
      "step": 12940
    },
    {
      "epoch": 6.784591194968553,
      "grad_norm": 1.4078006744384766,
      "learning_rate": 4.177986417746976e-05,
      "loss": 0.2496,
      "num_input_tokens_seen": 8441192,
      "step": 12945
    },
    {
      "epoch": 6.787211740041929,
      "grad_norm": 1.647047519683838,
      "learning_rate": 4.177138640195704e-05,
      "loss": 0.4044,
      "num_input_tokens_seen": 8446504,
      "step": 12950
    },
    {
      "epoch": 6.789832285115304,
      "grad_norm": 1.5886071920394897,
      "learning_rate": 4.176290511806162e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 8449384,
      "step": 12955
    },
    {
      "epoch": 6.7924528301886795,
      "grad_norm": 2.1169822216033936,
      "learning_rate": 4.1754420327557706e-05,
      "loss": 0.2604,
      "num_input_tokens_seen": 8452456,
      "step": 12960
    },
    {
      "epoch": 6.795073375262055,
      "grad_norm": 1.536686897277832,
      "learning_rate": 4.1745932032220195e-05,
      "loss": 0.3319,
      "num_input_tokens_seen": 8457000,
      "step": 12965
    },
    {
      "epoch": 6.79769392033543,
      "grad_norm": 1.7575995922088623,
      "learning_rate": 4.173744023382474e-05,
      "loss": 0.2601,
      "num_input_tokens_seen": 8461576,
      "step": 12970
    },
    {
      "epoch": 6.800314465408805,
      "grad_norm": 2.925907611846924,
      "learning_rate": 4.172894493414776e-05,
      "loss": 0.2767,
      "num_input_tokens_seen": 8464712,
      "step": 12975
    },
    {
      "epoch": 6.80293501048218,
      "grad_norm": 1.600106120109558,
      "learning_rate": 4.172044613496634e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 8467880,
      "step": 12980
    },
    {
      "epoch": 6.805555555555555,
      "grad_norm": 1.7285585403442383,
      "learning_rate": 4.171194383805835e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 8471208,
      "step": 12985
    },
    {
      "epoch": 6.8081761006289305,
      "grad_norm": 1.3960273265838623,
      "learning_rate": 4.170343804520237e-05,
      "loss": 0.2497,
      "num_input_tokens_seen": 8474248,
      "step": 12990
    },
    {
      "epoch": 6.810796645702306,
      "grad_norm": 3.155156373977661,
      "learning_rate": 4.169492875817771e-05,
      "loss": 0.2552,
      "num_input_tokens_seen": 8477064,
      "step": 12995
    },
    {
      "epoch": 6.813417190775681,
      "grad_norm": 1.6678842306137085,
      "learning_rate": 4.1686415978764415e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 8479528,
      "step": 13000
    },
    {
      "epoch": 6.816037735849057,
      "grad_norm": 2.8110415935516357,
      "learning_rate": 4.167789970874327e-05,
      "loss": 0.1833,
      "num_input_tokens_seen": 8482856,
      "step": 13005
    },
    {
      "epoch": 6.818658280922432,
      "grad_norm": 1.4450207948684692,
      "learning_rate": 4.166937994989578e-05,
      "loss": 0.3118,
      "num_input_tokens_seen": 8486312,
      "step": 13010
    },
    {
      "epoch": 6.821278825995807,
      "grad_norm": 1.9624695777893066,
      "learning_rate": 4.166085670400418e-05,
      "loss": 0.2886,
      "num_input_tokens_seen": 8489096,
      "step": 13015
    },
    {
      "epoch": 6.823899371069182,
      "grad_norm": 1.3484712839126587,
      "learning_rate": 4.165232997285143e-05,
      "loss": 0.3123,
      "num_input_tokens_seen": 8493448,
      "step": 13020
    },
    {
      "epoch": 6.826519916142558,
      "grad_norm": 1.4208271503448486,
      "learning_rate": 4.164379975822124e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 8496968,
      "step": 13025
    },
    {
      "epoch": 6.829140461215933,
      "grad_norm": 1.4104911088943481,
      "learning_rate": 4.1635266061898015e-05,
      "loss": 0.2071,
      "num_input_tokens_seen": 8500616,
      "step": 13030
    },
    {
      "epoch": 6.831761006289308,
      "grad_norm": 1.8207145929336548,
      "learning_rate": 4.162672888566691e-05,
      "loss": 0.2963,
      "num_input_tokens_seen": 8504136,
      "step": 13035
    },
    {
      "epoch": 6.834381551362683,
      "grad_norm": 1.9374972581863403,
      "learning_rate": 4.1618188231313816e-05,
      "loss": 0.2989,
      "num_input_tokens_seen": 8507464,
      "step": 13040
    },
    {
      "epoch": 6.837002096436059,
      "grad_norm": 1.868208408355713,
      "learning_rate": 4.1609644100625335e-05,
      "loss": 0.3301,
      "num_input_tokens_seen": 8510216,
      "step": 13045
    },
    {
      "epoch": 6.839622641509434,
      "grad_norm": 2.0195202827453613,
      "learning_rate": 4.16010964953888e-05,
      "loss": 0.3098,
      "num_input_tokens_seen": 8513960,
      "step": 13050
    },
    {
      "epoch": 6.8422431865828095,
      "grad_norm": 1.4275753498077393,
      "learning_rate": 4.159254541739227e-05,
      "loss": 0.2897,
      "num_input_tokens_seen": 8517768,
      "step": 13055
    },
    {
      "epoch": 6.844863731656185,
      "grad_norm": 1.4089018106460571,
      "learning_rate": 4.158399086842454e-05,
      "loss": 0.3569,
      "num_input_tokens_seen": 8521640,
      "step": 13060
    },
    {
      "epoch": 6.84748427672956,
      "grad_norm": 1.7650729417800903,
      "learning_rate": 4.157543285027512e-05,
      "loss": 0.2488,
      "num_input_tokens_seen": 8524968,
      "step": 13065
    },
    {
      "epoch": 6.850104821802935,
      "grad_norm": 1.957087755203247,
      "learning_rate": 4.156687136473425e-05,
      "loss": 0.4216,
      "num_input_tokens_seen": 8528200,
      "step": 13070
    },
    {
      "epoch": 6.85272536687631,
      "grad_norm": 2.4328043460845947,
      "learning_rate": 4.1558306413592894e-05,
      "loss": 0.2652,
      "num_input_tokens_seen": 8531144,
      "step": 13075
    },
    {
      "epoch": 6.855345911949685,
      "grad_norm": 1.708906888961792,
      "learning_rate": 4.1549737998642745e-05,
      "loss": 0.2826,
      "num_input_tokens_seen": 8535240,
      "step": 13080
    },
    {
      "epoch": 6.8579664570230605,
      "grad_norm": 1.425681471824646,
      "learning_rate": 4.154116612167622e-05,
      "loss": 0.2771,
      "num_input_tokens_seen": 8537352,
      "step": 13085
    },
    {
      "epoch": 6.860587002096436,
      "grad_norm": 2.204054355621338,
      "learning_rate": 4.1532590784486444e-05,
      "loss": 0.2748,
      "num_input_tokens_seen": 8540776,
      "step": 13090
    },
    {
      "epoch": 6.863207547169811,
      "grad_norm": 1.8274130821228027,
      "learning_rate": 4.1524011988867294e-05,
      "loss": 0.2985,
      "num_input_tokens_seen": 8543944,
      "step": 13095
    },
    {
      "epoch": 6.865828092243187,
      "grad_norm": 2.292201280593872,
      "learning_rate": 4.151542973661336e-05,
      "loss": 0.2931,
      "num_input_tokens_seen": 8546984,
      "step": 13100
    },
    {
      "epoch": 6.868448637316562,
      "grad_norm": 4.775777339935303,
      "learning_rate": 4.150684402951994e-05,
      "loss": 0.6053,
      "num_input_tokens_seen": 8550376,
      "step": 13105
    },
    {
      "epoch": 6.871069182389937,
      "grad_norm": 1.98771071434021,
      "learning_rate": 4.149825486938306e-05,
      "loss": 0.2825,
      "num_input_tokens_seen": 8553512,
      "step": 13110
    },
    {
      "epoch": 6.873689727463312,
      "grad_norm": 1.5379201173782349,
      "learning_rate": 4.148966225799949e-05,
      "loss": 0.2693,
      "num_input_tokens_seen": 8557320,
      "step": 13115
    },
    {
      "epoch": 6.876310272536688,
      "grad_norm": 1.5422550439834595,
      "learning_rate": 4.14810661971667e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 8559848,
      "step": 13120
    },
    {
      "epoch": 6.878930817610063,
      "grad_norm": 1.1311442852020264,
      "learning_rate": 4.147246668868289e-05,
      "loss": 0.1855,
      "num_input_tokens_seen": 8563400,
      "step": 13125
    },
    {
      "epoch": 6.881551362683438,
      "grad_norm": 1.3600283861160278,
      "learning_rate": 4.1463863734346964e-05,
      "loss": 0.2809,
      "num_input_tokens_seen": 8566728,
      "step": 13130
    },
    {
      "epoch": 6.884171907756813,
      "grad_norm": 2.083635091781616,
      "learning_rate": 4.145525733595859e-05,
      "loss": 0.2826,
      "num_input_tokens_seen": 8570056,
      "step": 13135
    },
    {
      "epoch": 6.886792452830189,
      "grad_norm": 1.4235951900482178,
      "learning_rate": 4.14466474953181e-05,
      "loss": 0.2468,
      "num_input_tokens_seen": 8573064,
      "step": 13140
    },
    {
      "epoch": 6.889412997903564,
      "grad_norm": 1.7629798650741577,
      "learning_rate": 4.14380342142266e-05,
      "loss": 0.3488,
      "num_input_tokens_seen": 8577640,
      "step": 13145
    },
    {
      "epoch": 6.8920335429769395,
      "grad_norm": 1.5241570472717285,
      "learning_rate": 4.142941749448587e-05,
      "loss": 0.3434,
      "num_input_tokens_seen": 8581256,
      "step": 13150
    },
    {
      "epoch": 6.894654088050315,
      "grad_norm": 1.7028497457504272,
      "learning_rate": 4.142079733789843e-05,
      "loss": 0.224,
      "num_input_tokens_seen": 8584168,
      "step": 13155
    },
    {
      "epoch": 6.89727463312369,
      "grad_norm": 0.9876751899719238,
      "learning_rate": 4.141217374626753e-05,
      "loss": 0.3297,
      "num_input_tokens_seen": 8587432,
      "step": 13160
    },
    {
      "epoch": 6.899895178197065,
      "grad_norm": 1.2671433687210083,
      "learning_rate": 4.140354672139713e-05,
      "loss": 0.2819,
      "num_input_tokens_seen": 8590728,
      "step": 13165
    },
    {
      "epoch": 6.90251572327044,
      "grad_norm": 1.5535407066345215,
      "learning_rate": 4.1394916265091886e-05,
      "loss": 0.2185,
      "num_input_tokens_seen": 8594120,
      "step": 13170
    },
    {
      "epoch": 6.905136268343815,
      "grad_norm": 1.0596994161605835,
      "learning_rate": 4.1386282379157214e-05,
      "loss": 0.3097,
      "num_input_tokens_seen": 8596904,
      "step": 13175
    },
    {
      "epoch": 6.9077568134171905,
      "grad_norm": 1.2723702192306519,
      "learning_rate": 4.137764506539921e-05,
      "loss": 0.2598,
      "num_input_tokens_seen": 8599848,
      "step": 13180
    },
    {
      "epoch": 6.910377358490566,
      "grad_norm": 1.7993396520614624,
      "learning_rate": 4.136900432562471e-05,
      "loss": 0.2741,
      "num_input_tokens_seen": 8603272,
      "step": 13185
    },
    {
      "epoch": 6.912997903563941,
      "grad_norm": 1.270626187324524,
      "learning_rate": 4.136036016164123e-05,
      "loss": 0.2881,
      "num_input_tokens_seen": 8607240,
      "step": 13190
    },
    {
      "epoch": 6.915618448637317,
      "grad_norm": 5.20631217956543,
      "learning_rate": 4.135171257525707e-05,
      "loss": 0.3748,
      "num_input_tokens_seen": 8613160,
      "step": 13195
    },
    {
      "epoch": 6.918238993710692,
      "grad_norm": 1.251295804977417,
      "learning_rate": 4.1343061568281195e-05,
      "loss": 0.2958,
      "num_input_tokens_seen": 8617480,
      "step": 13200
    },
    {
      "epoch": 6.920859538784067,
      "grad_norm": 1.3586007356643677,
      "learning_rate": 4.133440714252328e-05,
      "loss": 0.2165,
      "num_input_tokens_seen": 8620520,
      "step": 13205
    },
    {
      "epoch": 6.923480083857442,
      "grad_norm": 1.3155406713485718,
      "learning_rate": 4.132574929979375e-05,
      "loss": 0.1776,
      "num_input_tokens_seen": 8626568,
      "step": 13210
    },
    {
      "epoch": 6.926100628930818,
      "grad_norm": 1.508451223373413,
      "learning_rate": 4.1317088041903714e-05,
      "loss": 0.2801,
      "num_input_tokens_seen": 8630184,
      "step": 13215
    },
    {
      "epoch": 6.928721174004193,
      "grad_norm": 2.166332721710205,
      "learning_rate": 4.1308423370665015e-05,
      "loss": 0.2462,
      "num_input_tokens_seen": 8633576,
      "step": 13220
    },
    {
      "epoch": 6.931341719077568,
      "grad_norm": 1.1902278661727905,
      "learning_rate": 4.1299755287890206e-05,
      "loss": 0.2808,
      "num_input_tokens_seen": 8636520,
      "step": 13225
    },
    {
      "epoch": 6.933962264150943,
      "grad_norm": 2.229491949081421,
      "learning_rate": 4.129108379539255e-05,
      "loss": 0.2172,
      "num_input_tokens_seen": 8639816,
      "step": 13230
    },
    {
      "epoch": 6.936582809224319,
      "grad_norm": 1.843408226966858,
      "learning_rate": 4.128240889498602e-05,
      "loss": 0.3058,
      "num_input_tokens_seen": 8642888,
      "step": 13235
    },
    {
      "epoch": 6.939203354297694,
      "grad_norm": 1.1751211881637573,
      "learning_rate": 4.127373058848531e-05,
      "loss": 0.3106,
      "num_input_tokens_seen": 8645704,
      "step": 13240
    },
    {
      "epoch": 6.9418238993710695,
      "grad_norm": 1.5604777336120605,
      "learning_rate": 4.126504887770582e-05,
      "loss": 0.2359,
      "num_input_tokens_seen": 8648648,
      "step": 13245
    },
    {
      "epoch": 6.944444444444445,
      "grad_norm": 1.7482985258102417,
      "learning_rate": 4.1256363764463666e-05,
      "loss": 0.236,
      "num_input_tokens_seen": 8651560,
      "step": 13250
    },
    {
      "epoch": 6.94706498951782,
      "grad_norm": 1.2603728771209717,
      "learning_rate": 4.1247675250575676e-05,
      "loss": 0.3272,
      "num_input_tokens_seen": 8655208,
      "step": 13255
    },
    {
      "epoch": 6.949685534591195,
      "grad_norm": 1.385318636894226,
      "learning_rate": 4.1238983337859396e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 8657576,
      "step": 13260
    },
    {
      "epoch": 6.95230607966457,
      "grad_norm": 1.698155164718628,
      "learning_rate": 4.123028802813305e-05,
      "loss": 0.4335,
      "num_input_tokens_seen": 8660872,
      "step": 13265
    },
    {
      "epoch": 6.954926624737945,
      "grad_norm": 1.0648013353347778,
      "learning_rate": 4.1221589323215625e-05,
      "loss": 0.2687,
      "num_input_tokens_seen": 8665192,
      "step": 13270
    },
    {
      "epoch": 6.9575471698113205,
      "grad_norm": 0.8416537642478943,
      "learning_rate": 4.121288722492678e-05,
      "loss": 0.2466,
      "num_input_tokens_seen": 8670088,
      "step": 13275
    },
    {
      "epoch": 6.960167714884696,
      "grad_norm": 3.0837278366088867,
      "learning_rate": 4.120418173508688e-05,
      "loss": 0.2684,
      "num_input_tokens_seen": 8673576,
      "step": 13280
    },
    {
      "epoch": 6.962788259958071,
      "grad_norm": 1.4219658374786377,
      "learning_rate": 4.1195472855517045e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 8676424,
      "step": 13285
    },
    {
      "epoch": 6.965408805031447,
      "grad_norm": 2.2557952404022217,
      "learning_rate": 4.1186760588039045e-05,
      "loss": 0.2674,
      "num_input_tokens_seen": 8679912,
      "step": 13290
    },
    {
      "epoch": 6.968029350104822,
      "grad_norm": 1.2888444662094116,
      "learning_rate": 4.11780449344754e-05,
      "loss": 0.4953,
      "num_input_tokens_seen": 8683880,
      "step": 13295
    },
    {
      "epoch": 6.970649895178197,
      "grad_norm": 1.6715527772903442,
      "learning_rate": 4.1169325896649326e-05,
      "loss": 0.3659,
      "num_input_tokens_seen": 8687784,
      "step": 13300
    },
    {
      "epoch": 6.973270440251572,
      "grad_norm": 1.9801082611083984,
      "learning_rate": 4.116060347638474e-05,
      "loss": 0.271,
      "num_input_tokens_seen": 8690920,
      "step": 13305
    },
    {
      "epoch": 6.975890985324948,
      "grad_norm": 1.5705554485321045,
      "learning_rate": 4.115187767550628e-05,
      "loss": 0.2252,
      "num_input_tokens_seen": 8693960,
      "step": 13310
    },
    {
      "epoch": 6.978511530398323,
      "grad_norm": 2.6798312664031982,
      "learning_rate": 4.114314849583928e-05,
      "loss": 0.2686,
      "num_input_tokens_seen": 8697640,
      "step": 13315
    },
    {
      "epoch": 6.981132075471698,
      "grad_norm": 1.6194096803665161,
      "learning_rate": 4.113441593920979e-05,
      "loss": 0.4185,
      "num_input_tokens_seen": 8699912,
      "step": 13320
    },
    {
      "epoch": 6.983752620545073,
      "grad_norm": 1.8868660926818848,
      "learning_rate": 4.112568000744454e-05,
      "loss": 0.3583,
      "num_input_tokens_seen": 8703240,
      "step": 13325
    },
    {
      "epoch": 6.986373165618449,
      "grad_norm": 1.2749799489974976,
      "learning_rate": 4.1116940702371025e-05,
      "loss": 0.2146,
      "num_input_tokens_seen": 8706376,
      "step": 13330
    },
    {
      "epoch": 6.988993710691824,
      "grad_norm": 1.4692397117614746,
      "learning_rate": 4.1108198025817376e-05,
      "loss": 0.3494,
      "num_input_tokens_seen": 8709224,
      "step": 13335
    },
    {
      "epoch": 6.9916142557651995,
      "grad_norm": 3.2128031253814697,
      "learning_rate": 4.1099451979612476e-05,
      "loss": 0.3415,
      "num_input_tokens_seen": 8711816,
      "step": 13340
    },
    {
      "epoch": 6.994234800838575,
      "grad_norm": 2.2780046463012695,
      "learning_rate": 4.109070256558589e-05,
      "loss": 0.2762,
      "num_input_tokens_seen": 8715496,
      "step": 13345
    },
    {
      "epoch": 6.99685534591195,
      "grad_norm": 1.3875999450683594,
      "learning_rate": 4.108194978556791e-05,
      "loss": 0.2766,
      "num_input_tokens_seen": 8718312,
      "step": 13350
    },
    {
      "epoch": 6.999475890985325,
      "grad_norm": 2.0061089992523193,
      "learning_rate": 4.1073193641389495e-05,
      "loss": 0.2651,
      "num_input_tokens_seen": 8722632,
      "step": 13355
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.5636921525001526,
      "eval_runtime": 11.6595,
      "eval_samples_per_second": 72.73,
      "eval_steps_per_second": 18.183,
      "num_input_tokens_seen": 8722744,
      "step": 13356
    },
    {
      "epoch": 7.0020964360587,
      "grad_norm": 1.9269328117370605,
      "learning_rate": 4.1064434134882345e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 8725944,
      "step": 13360
    },
    {
      "epoch": 7.004716981132075,
      "grad_norm": 1.9232940673828125,
      "learning_rate": 4.1055671267878847e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 8729592,
      "step": 13365
    },
    {
      "epoch": 7.0073375262054505,
      "grad_norm": 1.3219385147094727,
      "learning_rate": 4.1046905042212106e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 8734264,
      "step": 13370
    },
    {
      "epoch": 7.009958071278826,
      "grad_norm": 1.0897793769836426,
      "learning_rate": 4.103813545971589e-05,
      "loss": 0.1807,
      "num_input_tokens_seen": 8738744,
      "step": 13375
    },
    {
      "epoch": 7.012578616352202,
      "grad_norm": 1.0511285066604614,
      "learning_rate": 4.10293625222247e-05,
      "loss": 0.2002,
      "num_input_tokens_seen": 8741720,
      "step": 13380
    },
    {
      "epoch": 7.015199161425577,
      "grad_norm": 1.6165016889572144,
      "learning_rate": 4.102058623157376e-05,
      "loss": 0.1993,
      "num_input_tokens_seen": 8745112,
      "step": 13385
    },
    {
      "epoch": 7.017819706498952,
      "grad_norm": 3.187776803970337,
      "learning_rate": 4.101180658959894e-05,
      "loss": 0.1775,
      "num_input_tokens_seen": 8747960,
      "step": 13390
    },
    {
      "epoch": 7.020440251572327,
      "grad_norm": 1.5987720489501953,
      "learning_rate": 4.1003023598136866e-05,
      "loss": 0.2448,
      "num_input_tokens_seen": 8750808,
      "step": 13395
    },
    {
      "epoch": 7.023060796645702,
      "grad_norm": 2.948115587234497,
      "learning_rate": 4.099423725902482e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 8753368,
      "step": 13400
    },
    {
      "epoch": 7.0256813417190775,
      "grad_norm": 2.588610887527466,
      "learning_rate": 4.098544757410081e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 8756056,
      "step": 13405
    },
    {
      "epoch": 7.028301886792453,
      "grad_norm": 1.9871143102645874,
      "learning_rate": 4.097665454520354e-05,
      "loss": 0.2249,
      "num_input_tokens_seen": 8759256,
      "step": 13410
    },
    {
      "epoch": 7.030922431865828,
      "grad_norm": 6.531321048736572,
      "learning_rate": 4.096785817417241e-05,
      "loss": 0.1641,
      "num_input_tokens_seen": 8763032,
      "step": 13415
    },
    {
      "epoch": 7.033542976939203,
      "grad_norm": 2.4663891792297363,
      "learning_rate": 4.0959058462847524e-05,
      "loss": 0.1342,
      "num_input_tokens_seen": 8766072,
      "step": 13420
    },
    {
      "epoch": 7.036163522012578,
      "grad_norm": 4.3105621337890625,
      "learning_rate": 4.095025541306968e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 8769976,
      "step": 13425
    },
    {
      "epoch": 7.038784067085954,
      "grad_norm": 1.4357964992523193,
      "learning_rate": 4.0941449026680374e-05,
      "loss": 0.2502,
      "num_input_tokens_seen": 8773528,
      "step": 13430
    },
    {
      "epoch": 7.0414046121593294,
      "grad_norm": 4.100120544433594,
      "learning_rate": 4.09326393055218e-05,
      "loss": 0.217,
      "num_input_tokens_seen": 8776440,
      "step": 13435
    },
    {
      "epoch": 7.044025157232705,
      "grad_norm": 1.603991985321045,
      "learning_rate": 4.0923826251436845e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 8780408,
      "step": 13440
    },
    {
      "epoch": 7.04664570230608,
      "grad_norm": 2.1377320289611816,
      "learning_rate": 4.0915009866269115e-05,
      "loss": 0.2224,
      "num_input_tokens_seen": 8782872,
      "step": 13445
    },
    {
      "epoch": 7.049266247379455,
      "grad_norm": 1.0914058685302734,
      "learning_rate": 4.090619015186288e-05,
      "loss": 0.2269,
      "num_input_tokens_seen": 8786776,
      "step": 13450
    },
    {
      "epoch": 7.05188679245283,
      "grad_norm": 1.2959452867507935,
      "learning_rate": 4.0897367110063145e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 8791576,
      "step": 13455
    },
    {
      "epoch": 7.054507337526205,
      "grad_norm": 2.2785799503326416,
      "learning_rate": 4.088854074271557e-05,
      "loss": 0.159,
      "num_input_tokens_seen": 8794072,
      "step": 13460
    },
    {
      "epoch": 7.0571278825995805,
      "grad_norm": 0.9538244605064392,
      "learning_rate": 4.0879711051666534e-05,
      "loss": 0.2387,
      "num_input_tokens_seen": 8798424,
      "step": 13465
    },
    {
      "epoch": 7.059748427672956,
      "grad_norm": 1.9503052234649658,
      "learning_rate": 4.087087803876313e-05,
      "loss": 0.183,
      "num_input_tokens_seen": 8801208,
      "step": 13470
    },
    {
      "epoch": 7.062368972746331,
      "grad_norm": 4.343796730041504,
      "learning_rate": 4.086204170585308e-05,
      "loss": 0.2291,
      "num_input_tokens_seen": 8804472,
      "step": 13475
    },
    {
      "epoch": 7.064989517819707,
      "grad_norm": 1.9013184309005737,
      "learning_rate": 4.085320205478488e-05,
      "loss": 0.2636,
      "num_input_tokens_seen": 8808152,
      "step": 13480
    },
    {
      "epoch": 7.067610062893082,
      "grad_norm": 2.2614572048187256,
      "learning_rate": 4.084435908740768e-05,
      "loss": 0.1885,
      "num_input_tokens_seen": 8811352,
      "step": 13485
    },
    {
      "epoch": 7.070230607966457,
      "grad_norm": 2.769265651702881,
      "learning_rate": 4.0835512805571306e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 8814328,
      "step": 13490
    },
    {
      "epoch": 7.072851153039832,
      "grad_norm": 1.368281602859497,
      "learning_rate": 4.082666321112633e-05,
      "loss": 0.253,
      "num_input_tokens_seen": 8817976,
      "step": 13495
    },
    {
      "epoch": 7.0754716981132075,
      "grad_norm": 0.9077532291412354,
      "learning_rate": 4.081781030592396e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 8820856,
      "step": 13500
    },
    {
      "epoch": 7.078092243186583,
      "grad_norm": 1.8136279582977295,
      "learning_rate": 4.080895409181613e-05,
      "loss": 0.3171,
      "num_input_tokens_seen": 8824120,
      "step": 13505
    },
    {
      "epoch": 7.080712788259958,
      "grad_norm": 1.6227971315383911,
      "learning_rate": 4.080009457065547e-05,
      "loss": 0.1943,
      "num_input_tokens_seen": 8827192,
      "step": 13510
    },
    {
      "epoch": 7.083333333333333,
      "grad_norm": 2.5032801628112793,
      "learning_rate": 4.0791231744295286e-05,
      "loss": 0.174,
      "num_input_tokens_seen": 8831064,
      "step": 13515
    },
    {
      "epoch": 7.085953878406708,
      "grad_norm": 1.3899593353271484,
      "learning_rate": 4.078236561458957e-05,
      "loss": 0.2832,
      "num_input_tokens_seen": 8834200,
      "step": 13520
    },
    {
      "epoch": 7.088574423480084,
      "grad_norm": 1.470560073852539,
      "learning_rate": 4.077349618339302e-05,
      "loss": 0.1506,
      "num_input_tokens_seen": 8837240,
      "step": 13525
    },
    {
      "epoch": 7.091194968553459,
      "grad_norm": 1.3000929355621338,
      "learning_rate": 4.076462345256103e-05,
      "loss": 0.3065,
      "num_input_tokens_seen": 8841976,
      "step": 13530
    },
    {
      "epoch": 7.093815513626835,
      "grad_norm": 2.531395673751831,
      "learning_rate": 4.0755747423949655e-05,
      "loss": 0.1592,
      "num_input_tokens_seen": 8845336,
      "step": 13535
    },
    {
      "epoch": 7.09643605870021,
      "grad_norm": 1.7036936283111572,
      "learning_rate": 4.074686809941568e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 8848184,
      "step": 13540
    },
    {
      "epoch": 7.099056603773585,
      "grad_norm": 3.189998149871826,
      "learning_rate": 4.0737985480816546e-05,
      "loss": 0.2047,
      "num_input_tokens_seen": 8850840,
      "step": 13545
    },
    {
      "epoch": 7.10167714884696,
      "grad_norm": 4.299417972564697,
      "learning_rate": 4.0729099570010395e-05,
      "loss": 0.3042,
      "num_input_tokens_seen": 8854296,
      "step": 13550
    },
    {
      "epoch": 7.104297693920335,
      "grad_norm": 1.8404052257537842,
      "learning_rate": 4.072021036885607e-05,
      "loss": 0.1617,
      "num_input_tokens_seen": 8857560,
      "step": 13555
    },
    {
      "epoch": 7.1069182389937104,
      "grad_norm": 1.2318637371063232,
      "learning_rate": 4.0711317879213075e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 8861496,
      "step": 13560
    },
    {
      "epoch": 7.109538784067086,
      "grad_norm": 1.5560120344161987,
      "learning_rate": 4.070242210294162e-05,
      "loss": 0.2068,
      "num_input_tokens_seen": 8864344,
      "step": 13565
    },
    {
      "epoch": 7.112159329140461,
      "grad_norm": 1.1313741207122803,
      "learning_rate": 4.069352304190261e-05,
      "loss": 0.2895,
      "num_input_tokens_seen": 8868152,
      "step": 13570
    },
    {
      "epoch": 7.114779874213837,
      "grad_norm": 1.9794386625289917,
      "learning_rate": 4.068462069795761e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 8870776,
      "step": 13575
    },
    {
      "epoch": 7.117400419287212,
      "grad_norm": 2.0226240158081055,
      "learning_rate": 4.067571507296891e-05,
      "loss": 0.2387,
      "num_input_tokens_seen": 8873944,
      "step": 13580
    },
    {
      "epoch": 7.120020964360587,
      "grad_norm": 4.09745979309082,
      "learning_rate": 4.0666806168799455e-05,
      "loss": 0.208,
      "num_input_tokens_seen": 8876792,
      "step": 13585
    },
    {
      "epoch": 7.122641509433962,
      "grad_norm": 1.5053068399429321,
      "learning_rate": 4.0657893987312876e-05,
      "loss": 0.2355,
      "num_input_tokens_seen": 8879672,
      "step": 13590
    },
    {
      "epoch": 7.1252620545073375,
      "grad_norm": 2.6710946559906006,
      "learning_rate": 4.064897853037352e-05,
      "loss": 0.1697,
      "num_input_tokens_seen": 8882904,
      "step": 13595
    },
    {
      "epoch": 7.127882599580713,
      "grad_norm": 1.5820269584655762,
      "learning_rate": 4.064005979984639e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 8886328,
      "step": 13600
    },
    {
      "epoch": 7.130503144654088,
      "grad_norm": 4.41533899307251,
      "learning_rate": 4.0631137797597176e-05,
      "loss": 0.1858,
      "num_input_tokens_seen": 8888888,
      "step": 13605
    },
    {
      "epoch": 7.133123689727463,
      "grad_norm": 3.1726064682006836,
      "learning_rate": 4.0622212525492266e-05,
      "loss": 0.1891,
      "num_input_tokens_seen": 8892536,
      "step": 13610
    },
    {
      "epoch": 7.135744234800838,
      "grad_norm": 1.183375358581543,
      "learning_rate": 4.061328398539872e-05,
      "loss": 0.1992,
      "num_input_tokens_seen": 8895768,
      "step": 13615
    },
    {
      "epoch": 7.138364779874214,
      "grad_norm": 1.9536625146865845,
      "learning_rate": 4.060435217918429e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 8899320,
      "step": 13620
    },
    {
      "epoch": 7.140985324947589,
      "grad_norm": 2.564108371734619,
      "learning_rate": 4.0595417108717405e-05,
      "loss": 0.233,
      "num_input_tokens_seen": 8901816,
      "step": 13625
    },
    {
      "epoch": 7.143605870020965,
      "grad_norm": 1.4689816236495972,
      "learning_rate": 4.0586478775867185e-05,
      "loss": 0.1684,
      "num_input_tokens_seen": 8904984,
      "step": 13630
    },
    {
      "epoch": 7.14622641509434,
      "grad_norm": 1.4340054988861084,
      "learning_rate": 4.057753718250343e-05,
      "loss": 0.2459,
      "num_input_tokens_seen": 8909272,
      "step": 13635
    },
    {
      "epoch": 7.148846960167715,
      "grad_norm": 1.582764983177185,
      "learning_rate": 4.056859233049661e-05,
      "loss": 0.2116,
      "num_input_tokens_seen": 8912088,
      "step": 13640
    },
    {
      "epoch": 7.15146750524109,
      "grad_norm": 1.6158392429351807,
      "learning_rate": 4.05596442217179e-05,
      "loss": 0.2539,
      "num_input_tokens_seen": 8915160,
      "step": 13645
    },
    {
      "epoch": 7.154088050314465,
      "grad_norm": 1.3646061420440674,
      "learning_rate": 4.055069285803912e-05,
      "loss": 0.1883,
      "num_input_tokens_seen": 8918200,
      "step": 13650
    },
    {
      "epoch": 7.15670859538784,
      "grad_norm": 1.6634880304336548,
      "learning_rate": 4.0541738241332815e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 8921464,
      "step": 13655
    },
    {
      "epoch": 7.159329140461216,
      "grad_norm": 3.5264570713043213,
      "learning_rate": 4.053278037347217e-05,
      "loss": 0.2018,
      "num_input_tokens_seen": 8924344,
      "step": 13660
    },
    {
      "epoch": 7.161949685534591,
      "grad_norm": 2.1657392978668213,
      "learning_rate": 4.052381925633109e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 8927352,
      "step": 13665
    },
    {
      "epoch": 7.164570230607967,
      "grad_norm": 4.276818752288818,
      "learning_rate": 4.051485489178412e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 8930936,
      "step": 13670
    },
    {
      "epoch": 7.167190775681342,
      "grad_norm": 2.803510904312134,
      "learning_rate": 4.05058872817065e-05,
      "loss": 0.2837,
      "num_input_tokens_seen": 8934296,
      "step": 13675
    },
    {
      "epoch": 7.169811320754717,
      "grad_norm": 1.192264199256897,
      "learning_rate": 4.0496916427974177e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 8937016,
      "step": 13680
    },
    {
      "epoch": 7.172431865828092,
      "grad_norm": 1.7104854583740234,
      "learning_rate": 4.048794233246373e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 8940792,
      "step": 13685
    },
    {
      "epoch": 7.1750524109014675,
      "grad_norm": 1.623905897140503,
      "learning_rate": 4.0478964997052435e-05,
      "loss": 0.2723,
      "num_input_tokens_seen": 8944024,
      "step": 13690
    },
    {
      "epoch": 7.177672955974843,
      "grad_norm": 2.7138848304748535,
      "learning_rate": 4.046998442361827e-05,
      "loss": 0.2152,
      "num_input_tokens_seen": 8946872,
      "step": 13695
    },
    {
      "epoch": 7.180293501048218,
      "grad_norm": 1.469689965248108,
      "learning_rate": 4.046100061403984e-05,
      "loss": 0.1939,
      "num_input_tokens_seen": 8950168,
      "step": 13700
    },
    {
      "epoch": 7.182914046121593,
      "grad_norm": 6.379891395568848,
      "learning_rate": 4.045201357019648e-05,
      "loss": 0.2274,
      "num_input_tokens_seen": 8953112,
      "step": 13705
    },
    {
      "epoch": 7.185534591194968,
      "grad_norm": 1.9657180309295654,
      "learning_rate": 4.044302329396816e-05,
      "loss": 0.1843,
      "num_input_tokens_seen": 8956312,
      "step": 13710
    },
    {
      "epoch": 7.188155136268344,
      "grad_norm": 2.6565651893615723,
      "learning_rate": 4.0434029787235564e-05,
      "loss": 0.1987,
      "num_input_tokens_seen": 8959032,
      "step": 13715
    },
    {
      "epoch": 7.190775681341719,
      "grad_norm": 2.831611156463623,
      "learning_rate": 4.042503305188e-05,
      "loss": 0.2824,
      "num_input_tokens_seen": 8962744,
      "step": 13720
    },
    {
      "epoch": 7.193396226415095,
      "grad_norm": 2.320228338241577,
      "learning_rate": 4.041603308978351e-05,
      "loss": 0.2527,
      "num_input_tokens_seen": 8966520,
      "step": 13725
    },
    {
      "epoch": 7.19601677148847,
      "grad_norm": 1.6439062356948853,
      "learning_rate": 4.040702990282878e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 8969304,
      "step": 13730
    },
    {
      "epoch": 7.198637316561845,
      "grad_norm": 1.3848230838775635,
      "learning_rate": 4.039802349289914e-05,
      "loss": 0.3044,
      "num_input_tokens_seen": 8973432,
      "step": 13735
    },
    {
      "epoch": 7.20125786163522,
      "grad_norm": 2.7234368324279785,
      "learning_rate": 4.038901386187868e-05,
      "loss": 0.2141,
      "num_input_tokens_seen": 8976248,
      "step": 13740
    },
    {
      "epoch": 7.203878406708595,
      "grad_norm": 1.420428991317749,
      "learning_rate": 4.038000101165208e-05,
      "loss": 0.2232,
      "num_input_tokens_seen": 8980792,
      "step": 13745
    },
    {
      "epoch": 7.20649895178197,
      "grad_norm": 1.5987966060638428,
      "learning_rate": 4.0370984944104725e-05,
      "loss": 0.2477,
      "num_input_tokens_seen": 8983480,
      "step": 13750
    },
    {
      "epoch": 7.209119496855346,
      "grad_norm": 1.732732892036438,
      "learning_rate": 4.0361965661122695e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 8986200,
      "step": 13755
    },
    {
      "epoch": 7.211740041928721,
      "grad_norm": 2.3156862258911133,
      "learning_rate": 4.035294316459271e-05,
      "loss": 0.2065,
      "num_input_tokens_seen": 8989816,
      "step": 13760
    },
    {
      "epoch": 7.214360587002097,
      "grad_norm": 2.0514590740203857,
      "learning_rate": 4.034391745640215e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 8992600,
      "step": 13765
    },
    {
      "epoch": 7.216981132075472,
      "grad_norm": 2.466654062271118,
      "learning_rate": 4.033488853843913e-05,
      "loss": 0.2191,
      "num_input_tokens_seen": 8997432,
      "step": 13770
    },
    {
      "epoch": 7.219601677148847,
      "grad_norm": 2.778792142868042,
      "learning_rate": 4.0325856412592363e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 8999864,
      "step": 13775
    },
    {
      "epoch": 7.222222222222222,
      "grad_norm": 1.0664052963256836,
      "learning_rate": 4.031682108075129e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 9003160,
      "step": 13780
    },
    {
      "epoch": 7.2248427672955975,
      "grad_norm": 2.756289005279541,
      "learning_rate": 4.030778254480599e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 9006040,
      "step": 13785
    },
    {
      "epoch": 7.227463312368973,
      "grad_norm": 2.611870765686035,
      "learning_rate": 4.029874080664722e-05,
      "loss": 0.2179,
      "num_input_tokens_seen": 9009624,
      "step": 13790
    },
    {
      "epoch": 7.230083857442348,
      "grad_norm": 1.0890480279922485,
      "learning_rate": 4.0289695868166414e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 9013912,
      "step": 13795
    },
    {
      "epoch": 7.232704402515723,
      "grad_norm": 2.109097719192505,
      "learning_rate": 4.028064773125566e-05,
      "loss": 0.2221,
      "num_input_tokens_seen": 9017016,
      "step": 13800
    },
    {
      "epoch": 7.235324947589098,
      "grad_norm": 3.097383975982666,
      "learning_rate": 4.0271596397807734e-05,
      "loss": 0.2134,
      "num_input_tokens_seen": 9019000,
      "step": 13805
    },
    {
      "epoch": 7.237945492662474,
      "grad_norm": 2.1657986640930176,
      "learning_rate": 4.026254186971607e-05,
      "loss": 0.3029,
      "num_input_tokens_seen": 9022520,
      "step": 13810
    },
    {
      "epoch": 7.240566037735849,
      "grad_norm": 4.176776885986328,
      "learning_rate": 4.025348414887477e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 9025496,
      "step": 13815
    },
    {
      "epoch": 7.243186582809225,
      "grad_norm": 1.7511155605316162,
      "learning_rate": 4.024442323717861e-05,
      "loss": 0.205,
      "num_input_tokens_seen": 9029464,
      "step": 13820
    },
    {
      "epoch": 7.2458071278826,
      "grad_norm": 2.3414435386657715,
      "learning_rate": 4.023535913652302e-05,
      "loss": 0.2654,
      "num_input_tokens_seen": 9032088,
      "step": 13825
    },
    {
      "epoch": 7.248427672955975,
      "grad_norm": 1.6221563816070557,
      "learning_rate": 4.022629184880412e-05,
      "loss": 0.2091,
      "num_input_tokens_seen": 9035768,
      "step": 13830
    },
    {
      "epoch": 7.25104821802935,
      "grad_norm": 2.307288646697998,
      "learning_rate": 4.021722137591867e-05,
      "loss": 0.2115,
      "num_input_tokens_seen": 9038840,
      "step": 13835
    },
    {
      "epoch": 7.253668763102725,
      "grad_norm": 2.9499361515045166,
      "learning_rate": 4.020814771976412e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 9042328,
      "step": 13840
    },
    {
      "epoch": 7.2562893081761,
      "grad_norm": 1.6573642492294312,
      "learning_rate": 4.019907088223857e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 9045336,
      "step": 13845
    },
    {
      "epoch": 7.258909853249476,
      "grad_norm": 2.714646100997925,
      "learning_rate": 4.018999086524079e-05,
      "loss": 0.2455,
      "num_input_tokens_seen": 9048312,
      "step": 13850
    },
    {
      "epoch": 7.261530398322851,
      "grad_norm": 1.957669973373413,
      "learning_rate": 4.018090767067022e-05,
      "loss": 0.2806,
      "num_input_tokens_seen": 9052760,
      "step": 13855
    },
    {
      "epoch": 7.264150943396227,
      "grad_norm": 5.931875705718994,
      "learning_rate": 4.017182130042696e-05,
      "loss": 0.3982,
      "num_input_tokens_seen": 9055864,
      "step": 13860
    },
    {
      "epoch": 7.266771488469602,
      "grad_norm": 1.6778552532196045,
      "learning_rate": 4.0162731756411764e-05,
      "loss": 0.1759,
      "num_input_tokens_seen": 9058776,
      "step": 13865
    },
    {
      "epoch": 7.269392033542977,
      "grad_norm": 1.6809239387512207,
      "learning_rate": 4.015363904052608e-05,
      "loss": 0.2724,
      "num_input_tokens_seen": 9062360,
      "step": 13870
    },
    {
      "epoch": 7.272012578616352,
      "grad_norm": 1.4800384044647217,
      "learning_rate": 4.014454315467199e-05,
      "loss": 0.2892,
      "num_input_tokens_seen": 9065016,
      "step": 13875
    },
    {
      "epoch": 7.2746331236897275,
      "grad_norm": 1.874128818511963,
      "learning_rate": 4.013544410075224e-05,
      "loss": 0.2203,
      "num_input_tokens_seen": 9071992,
      "step": 13880
    },
    {
      "epoch": 7.277253668763103,
      "grad_norm": 1.3385062217712402,
      "learning_rate": 4.0126341880670273e-05,
      "loss": 0.2451,
      "num_input_tokens_seen": 9075672,
      "step": 13885
    },
    {
      "epoch": 7.279874213836478,
      "grad_norm": 1.5186740159988403,
      "learning_rate": 4.011723649633014e-05,
      "loss": 0.2222,
      "num_input_tokens_seen": 9078936,
      "step": 13890
    },
    {
      "epoch": 7.282494758909853,
      "grad_norm": 3.936445474624634,
      "learning_rate": 4.0108127949636606e-05,
      "loss": 0.3074,
      "num_input_tokens_seen": 9082168,
      "step": 13895
    },
    {
      "epoch": 7.285115303983228,
      "grad_norm": 2.828125476837158,
      "learning_rate": 4.009901624249507e-05,
      "loss": 0.2937,
      "num_input_tokens_seen": 9084664,
      "step": 13900
    },
    {
      "epoch": 7.287735849056604,
      "grad_norm": 1.2939515113830566,
      "learning_rate": 4.008990137681159e-05,
      "loss": 0.2188,
      "num_input_tokens_seen": 9087864,
      "step": 13905
    },
    {
      "epoch": 7.290356394129979,
      "grad_norm": 1.5254303216934204,
      "learning_rate": 4.00807833544929e-05,
      "loss": 0.2532,
      "num_input_tokens_seen": 9091416,
      "step": 13910
    },
    {
      "epoch": 7.2929769392033545,
      "grad_norm": 3.2694931030273438,
      "learning_rate": 4.0071662177446376e-05,
      "loss": 0.2495,
      "num_input_tokens_seen": 9094296,
      "step": 13915
    },
    {
      "epoch": 7.29559748427673,
      "grad_norm": 1.4579683542251587,
      "learning_rate": 4.006253784758008e-05,
      "loss": 0.2434,
      "num_input_tokens_seen": 9097112,
      "step": 13920
    },
    {
      "epoch": 7.298218029350105,
      "grad_norm": 1.894858956336975,
      "learning_rate": 4.0053410366802704e-05,
      "loss": 0.1989,
      "num_input_tokens_seen": 9100632,
      "step": 13925
    },
    {
      "epoch": 7.30083857442348,
      "grad_norm": 2.253549098968506,
      "learning_rate": 4.004427973702362e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 9103864,
      "step": 13930
    },
    {
      "epoch": 7.303459119496855,
      "grad_norm": 2.048334836959839,
      "learning_rate": 4.003514596015285e-05,
      "loss": 0.2145,
      "num_input_tokens_seen": 9107928,
      "step": 13935
    },
    {
      "epoch": 7.30607966457023,
      "grad_norm": 1.4765433073043823,
      "learning_rate": 4.002600903810108e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 9111384,
      "step": 13940
    },
    {
      "epoch": 7.308700209643606,
      "grad_norm": 2.3944590091705322,
      "learning_rate": 4.001686897277963e-05,
      "loss": 0.2424,
      "num_input_tokens_seen": 9114488,
      "step": 13945
    },
    {
      "epoch": 7.311320754716981,
      "grad_norm": 1.2168973684310913,
      "learning_rate": 4.000772576610052e-05,
      "loss": 0.2033,
      "num_input_tokens_seen": 9117208,
      "step": 13950
    },
    {
      "epoch": 7.313941299790357,
      "grad_norm": 3.719318151473999,
      "learning_rate": 3.99985794199764e-05,
      "loss": 0.2149,
      "num_input_tokens_seen": 9121176,
      "step": 13955
    },
    {
      "epoch": 7.316561844863732,
      "grad_norm": 3.789658546447754,
      "learning_rate": 3.998942993632058e-05,
      "loss": 0.2813,
      "num_input_tokens_seen": 9124216,
      "step": 13960
    },
    {
      "epoch": 7.319182389937107,
      "grad_norm": 3.2193565368652344,
      "learning_rate": 3.998027731704702e-05,
      "loss": 0.2322,
      "num_input_tokens_seen": 9126936,
      "step": 13965
    },
    {
      "epoch": 7.321802935010482,
      "grad_norm": 1.7277497053146362,
      "learning_rate": 3.9971121564070354e-05,
      "loss": 0.2645,
      "num_input_tokens_seen": 9129688,
      "step": 13970
    },
    {
      "epoch": 7.3244234800838575,
      "grad_norm": 1.4955769777297974,
      "learning_rate": 3.996196267930585e-05,
      "loss": 0.1584,
      "num_input_tokens_seen": 9132152,
      "step": 13975
    },
    {
      "epoch": 7.327044025157233,
      "grad_norm": 1.7845392227172852,
      "learning_rate": 3.995280066466945e-05,
      "loss": 0.2905,
      "num_input_tokens_seen": 9135352,
      "step": 13980
    },
    {
      "epoch": 7.329664570230608,
      "grad_norm": 2.004671335220337,
      "learning_rate": 3.9943635522077746e-05,
      "loss": 0.2998,
      "num_input_tokens_seen": 9139064,
      "step": 13985
    },
    {
      "epoch": 7.332285115303983,
      "grad_norm": 1.4374299049377441,
      "learning_rate": 3.993446725344797e-05,
      "loss": 0.196,
      "num_input_tokens_seen": 9141624,
      "step": 13990
    },
    {
      "epoch": 7.334905660377358,
      "grad_norm": 1.2918102741241455,
      "learning_rate": 3.992529586069803e-05,
      "loss": 0.1546,
      "num_input_tokens_seen": 9144120,
      "step": 13995
    },
    {
      "epoch": 7.337526205450734,
      "grad_norm": 1.6119787693023682,
      "learning_rate": 3.991612134574648e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 9147384,
      "step": 14000
    },
    {
      "epoch": 7.340146750524109,
      "grad_norm": 1.2282700538635254,
      "learning_rate": 3.99069437105125e-05,
      "loss": 0.24,
      "num_input_tokens_seen": 9150744,
      "step": 14005
    },
    {
      "epoch": 7.3427672955974845,
      "grad_norm": 1.424468755722046,
      "learning_rate": 3.9897762956915976e-05,
      "loss": 0.2109,
      "num_input_tokens_seen": 9153560,
      "step": 14010
    },
    {
      "epoch": 7.34538784067086,
      "grad_norm": 1.9655914306640625,
      "learning_rate": 3.9888579086877403e-05,
      "loss": 0.1496,
      "num_input_tokens_seen": 9156696,
      "step": 14015
    },
    {
      "epoch": 7.348008385744235,
      "grad_norm": 1.2545174360275269,
      "learning_rate": 3.987939210231794e-05,
      "loss": 0.1731,
      "num_input_tokens_seen": 9159416,
      "step": 14020
    },
    {
      "epoch": 7.35062893081761,
      "grad_norm": 1.712816834449768,
      "learning_rate": 3.987020200515941e-05,
      "loss": 0.2119,
      "num_input_tokens_seen": 9162584,
      "step": 14025
    },
    {
      "epoch": 7.353249475890985,
      "grad_norm": 1.8804478645324707,
      "learning_rate": 3.9861008797324255e-05,
      "loss": 0.1983,
      "num_input_tokens_seen": 9165848,
      "step": 14030
    },
    {
      "epoch": 7.35587002096436,
      "grad_norm": 2.8788902759552,
      "learning_rate": 3.9851812480735605e-05,
      "loss": 0.2542,
      "num_input_tokens_seen": 9168600,
      "step": 14035
    },
    {
      "epoch": 7.3584905660377355,
      "grad_norm": 2.1809935569763184,
      "learning_rate": 3.9842613057317235e-05,
      "loss": 0.1926,
      "num_input_tokens_seen": 9173336,
      "step": 14040
    },
    {
      "epoch": 7.361111111111111,
      "grad_norm": 1.3881961107254028,
      "learning_rate": 3.9833410528993534e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 9176376,
      "step": 14045
    },
    {
      "epoch": 7.363731656184487,
      "grad_norm": 1.8756237030029297,
      "learning_rate": 3.9824204897689585e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 9180088,
      "step": 14050
    },
    {
      "epoch": 7.366352201257862,
      "grad_norm": 2.6962575912475586,
      "learning_rate": 3.98149961653311e-05,
      "loss": 0.1737,
      "num_input_tokens_seen": 9183064,
      "step": 14055
    },
    {
      "epoch": 7.368972746331237,
      "grad_norm": 2.5148165225982666,
      "learning_rate": 3.980578433384442e-05,
      "loss": 0.2294,
      "num_input_tokens_seen": 9186040,
      "step": 14060
    },
    {
      "epoch": 7.371593291404612,
      "grad_norm": 1.8280278444290161,
      "learning_rate": 3.979656940515659e-05,
      "loss": 0.2566,
      "num_input_tokens_seen": 9189080,
      "step": 14065
    },
    {
      "epoch": 7.3742138364779874,
      "grad_norm": 1.443115234375,
      "learning_rate": 3.9787351381195235e-05,
      "loss": 0.3479,
      "num_input_tokens_seen": 9191992,
      "step": 14070
    },
    {
      "epoch": 7.376834381551363,
      "grad_norm": 1.9425902366638184,
      "learning_rate": 3.977813026388868e-05,
      "loss": 0.2032,
      "num_input_tokens_seen": 9194520,
      "step": 14075
    },
    {
      "epoch": 7.379454926624738,
      "grad_norm": 2.926790952682495,
      "learning_rate": 3.976890605516587e-05,
      "loss": 0.2495,
      "num_input_tokens_seen": 9196632,
      "step": 14080
    },
    {
      "epoch": 7.382075471698113,
      "grad_norm": 2.2560949325561523,
      "learning_rate": 3.97596787569564e-05,
      "loss": 0.2401,
      "num_input_tokens_seen": 9199832,
      "step": 14085
    },
    {
      "epoch": 7.384696016771488,
      "grad_norm": 1.6613647937774658,
      "learning_rate": 3.975044837119053e-05,
      "loss": 0.2396,
      "num_input_tokens_seen": 9203384,
      "step": 14090
    },
    {
      "epoch": 7.387316561844864,
      "grad_norm": 1.8001182079315186,
      "learning_rate": 3.974121489979914e-05,
      "loss": 0.191,
      "num_input_tokens_seen": 9206840,
      "step": 14095
    },
    {
      "epoch": 7.389937106918239,
      "grad_norm": 1.8423717021942139,
      "learning_rate": 3.973197834471377e-05,
      "loss": 0.2783,
      "num_input_tokens_seen": 9209432,
      "step": 14100
    },
    {
      "epoch": 7.3925576519916145,
      "grad_norm": 2.775785446166992,
      "learning_rate": 3.9722738707866595e-05,
      "loss": 0.2213,
      "num_input_tokens_seen": 9212056,
      "step": 14105
    },
    {
      "epoch": 7.39517819706499,
      "grad_norm": 1.8044534921646118,
      "learning_rate": 3.971349599119046e-05,
      "loss": 0.1573,
      "num_input_tokens_seen": 9215224,
      "step": 14110
    },
    {
      "epoch": 7.397798742138365,
      "grad_norm": 1.9094080924987793,
      "learning_rate": 3.970425019661881e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 9219672,
      "step": 14115
    },
    {
      "epoch": 7.40041928721174,
      "grad_norm": 2.2794406414031982,
      "learning_rate": 3.9695001326085785e-05,
      "loss": 0.2033,
      "num_input_tokens_seen": 9222616,
      "step": 14120
    },
    {
      "epoch": 7.403039832285115,
      "grad_norm": 1.3576565980911255,
      "learning_rate": 3.968574938152613e-05,
      "loss": 0.3727,
      "num_input_tokens_seen": 9227032,
      "step": 14125
    },
    {
      "epoch": 7.40566037735849,
      "grad_norm": 4.8790364265441895,
      "learning_rate": 3.9676494364875247e-05,
      "loss": 0.251,
      "num_input_tokens_seen": 9229688,
      "step": 14130
    },
    {
      "epoch": 7.4082809224318655,
      "grad_norm": 2.2105259895324707,
      "learning_rate": 3.966723627806918e-05,
      "loss": 0.2661,
      "num_input_tokens_seen": 9233016,
      "step": 14135
    },
    {
      "epoch": 7.410901467505241,
      "grad_norm": 1.9413254261016846,
      "learning_rate": 3.9657975123044624e-05,
      "loss": 0.2138,
      "num_input_tokens_seen": 9236504,
      "step": 14140
    },
    {
      "epoch": 7.413522012578617,
      "grad_norm": 2.4998443126678467,
      "learning_rate": 3.964871090173888e-05,
      "loss": 0.2309,
      "num_input_tokens_seen": 9240344,
      "step": 14145
    },
    {
      "epoch": 7.416142557651992,
      "grad_norm": 2.2660326957702637,
      "learning_rate": 3.963944361608995e-05,
      "loss": 0.185,
      "num_input_tokens_seen": 9242904,
      "step": 14150
    },
    {
      "epoch": 7.418763102725367,
      "grad_norm": 2.936939001083374,
      "learning_rate": 3.963017326803644e-05,
      "loss": 0.2878,
      "num_input_tokens_seen": 9246072,
      "step": 14155
    },
    {
      "epoch": 7.421383647798742,
      "grad_norm": 2.265719413757324,
      "learning_rate": 3.962089985951758e-05,
      "loss": 0.2064,
      "num_input_tokens_seen": 9248696,
      "step": 14160
    },
    {
      "epoch": 7.424004192872117,
      "grad_norm": 3.000471353530884,
      "learning_rate": 3.961162339247327e-05,
      "loss": 0.3795,
      "num_input_tokens_seen": 9251160,
      "step": 14165
    },
    {
      "epoch": 7.426624737945493,
      "grad_norm": 1.6994645595550537,
      "learning_rate": 3.9602343868844055e-05,
      "loss": 0.2428,
      "num_input_tokens_seen": 9254040,
      "step": 14170
    },
    {
      "epoch": 7.429245283018868,
      "grad_norm": 1.306282639503479,
      "learning_rate": 3.959306129057109e-05,
      "loss": 0.2239,
      "num_input_tokens_seen": 9257816,
      "step": 14175
    },
    {
      "epoch": 7.431865828092243,
      "grad_norm": 2.383876323699951,
      "learning_rate": 3.958377565959619e-05,
      "loss": 0.2119,
      "num_input_tokens_seen": 9260184,
      "step": 14180
    },
    {
      "epoch": 7.434486373165618,
      "grad_norm": 1.9144054651260376,
      "learning_rate": 3.95744869778618e-05,
      "loss": 0.3073,
      "num_input_tokens_seen": 9263032,
      "step": 14185
    },
    {
      "epoch": 7.437106918238994,
      "grad_norm": 1.5207916498184204,
      "learning_rate": 3.956519524731101e-05,
      "loss": 0.2918,
      "num_input_tokens_seen": 9267448,
      "step": 14190
    },
    {
      "epoch": 7.439727463312369,
      "grad_norm": 1.73007333278656,
      "learning_rate": 3.9555900469887555e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 9270200,
      "step": 14195
    },
    {
      "epoch": 7.4423480083857445,
      "grad_norm": 2.261655807495117,
      "learning_rate": 3.954660264753577e-05,
      "loss": 0.2248,
      "num_input_tokens_seen": 9272856,
      "step": 14200
    },
    {
      "epoch": 7.44496855345912,
      "grad_norm": 1.4734394550323486,
      "learning_rate": 3.953730178220067e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 9276504,
      "step": 14205
    },
    {
      "epoch": 7.447589098532495,
      "grad_norm": 2.642159938812256,
      "learning_rate": 3.952799787582789e-05,
      "loss": 0.28,
      "num_input_tokens_seen": 9279736,
      "step": 14210
    },
    {
      "epoch": 7.45020964360587,
      "grad_norm": 1.9617210626602173,
      "learning_rate": 3.9518690930363694e-05,
      "loss": 0.2268,
      "num_input_tokens_seen": 9282680,
      "step": 14215
    },
    {
      "epoch": 7.452830188679245,
      "grad_norm": 3.9047086238861084,
      "learning_rate": 3.9509380947754994e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 9285912,
      "step": 14220
    },
    {
      "epoch": 7.45545073375262,
      "grad_norm": 1.5252034664154053,
      "learning_rate": 3.9500067929949336e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 9289176,
      "step": 14225
    },
    {
      "epoch": 7.4580712788259955,
      "grad_norm": 2.8555445671081543,
      "learning_rate": 3.9490751878894895e-05,
      "loss": 0.2157,
      "num_input_tokens_seen": 9291736,
      "step": 14230
    },
    {
      "epoch": 7.460691823899371,
      "grad_norm": 1.2664142847061157,
      "learning_rate": 3.948143279654047e-05,
      "loss": 0.2035,
      "num_input_tokens_seen": 9296312,
      "step": 14235
    },
    {
      "epoch": 7.463312368972747,
      "grad_norm": 2.5054211616516113,
      "learning_rate": 3.947211068483553e-05,
      "loss": 0.2267,
      "num_input_tokens_seen": 9298904,
      "step": 14240
    },
    {
      "epoch": 7.465932914046122,
      "grad_norm": 2.2430479526519775,
      "learning_rate": 3.946278554573013e-05,
      "loss": 0.2182,
      "num_input_tokens_seen": 9301432,
      "step": 14245
    },
    {
      "epoch": 7.468553459119497,
      "grad_norm": 2.622699737548828,
      "learning_rate": 3.9453457381175e-05,
      "loss": 0.2447,
      "num_input_tokens_seen": 9304376,
      "step": 14250
    },
    {
      "epoch": 7.471174004192872,
      "grad_norm": 2.8385024070739746,
      "learning_rate": 3.9444126193121474e-05,
      "loss": 0.2469,
      "num_input_tokens_seen": 9307864,
      "step": 14255
    },
    {
      "epoch": 7.473794549266247,
      "grad_norm": 2.2528862953186035,
      "learning_rate": 3.943479198352153e-05,
      "loss": 0.2619,
      "num_input_tokens_seen": 9311032,
      "step": 14260
    },
    {
      "epoch": 7.476415094339623,
      "grad_norm": 1.9814574718475342,
      "learning_rate": 3.94254547543278e-05,
      "loss": 0.1928,
      "num_input_tokens_seen": 9314264,
      "step": 14265
    },
    {
      "epoch": 7.479035639412998,
      "grad_norm": 2.5033366680145264,
      "learning_rate": 3.9416114507493504e-05,
      "loss": 0.2089,
      "num_input_tokens_seen": 9317272,
      "step": 14270
    },
    {
      "epoch": 7.481656184486373,
      "grad_norm": 2.567181348800659,
      "learning_rate": 3.940677124497252e-05,
      "loss": 0.2167,
      "num_input_tokens_seen": 9320504,
      "step": 14275
    },
    {
      "epoch": 7.484276729559748,
      "grad_norm": 2.091782808303833,
      "learning_rate": 3.939742496871935e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 9324792,
      "step": 14280
    },
    {
      "epoch": 7.486897274633124,
      "grad_norm": 1.0982681512832642,
      "learning_rate": 3.9388075680689136e-05,
      "loss": 0.1447,
      "num_input_tokens_seen": 9327544,
      "step": 14285
    },
    {
      "epoch": 7.489517819706499,
      "grad_norm": 2.147439956665039,
      "learning_rate": 3.9378723382837634e-05,
      "loss": 0.2448,
      "num_input_tokens_seen": 9330488,
      "step": 14290
    },
    {
      "epoch": 7.4921383647798745,
      "grad_norm": 1.8428542613983154,
      "learning_rate": 3.936936807712124e-05,
      "loss": 0.2944,
      "num_input_tokens_seen": 9335288,
      "step": 14295
    },
    {
      "epoch": 7.49475890985325,
      "grad_norm": 1.2758122682571411,
      "learning_rate": 3.936000976549698e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 9339384,
      "step": 14300
    },
    {
      "epoch": 7.497379454926625,
      "grad_norm": 1.6123183965682983,
      "learning_rate": 3.935064844992251e-05,
      "loss": 0.2639,
      "num_input_tokens_seen": 9342872,
      "step": 14305
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.4796215295791626,
      "learning_rate": 3.93412841323561e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 9345976,
      "step": 14310
    },
    {
      "epoch": 7.502620545073375,
      "grad_norm": 1.712274193763733,
      "learning_rate": 3.933191681475666e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 9349208,
      "step": 14315
    },
    {
      "epoch": 7.50524109014675,
      "grad_norm": 2.1891138553619385,
      "learning_rate": 3.932254649908373e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 9352184,
      "step": 14320
    },
    {
      "epoch": 7.5078616352201255,
      "grad_norm": 1.7967965602874756,
      "learning_rate": 3.931317318729748e-05,
      "loss": 0.2304,
      "num_input_tokens_seen": 9356216,
      "step": 14325
    },
    {
      "epoch": 7.510482180293501,
      "grad_norm": 1.8096818923950195,
      "learning_rate": 3.930379688135869e-05,
      "loss": 0.2148,
      "num_input_tokens_seen": 9359544,
      "step": 14330
    },
    {
      "epoch": 7.513102725366876,
      "grad_norm": 1.3195527791976929,
      "learning_rate": 3.929441758322878e-05,
      "loss": 0.2071,
      "num_input_tokens_seen": 9362200,
      "step": 14335
    },
    {
      "epoch": 7.515723270440252,
      "grad_norm": 1.4157075881958008,
      "learning_rate": 3.928503529486979e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 9366520,
      "step": 14340
    },
    {
      "epoch": 7.518343815513627,
      "grad_norm": 1.1853363513946533,
      "learning_rate": 3.927565001824438e-05,
      "loss": 0.2321,
      "num_input_tokens_seen": 9372440,
      "step": 14345
    },
    {
      "epoch": 7.520964360587002,
      "grad_norm": 2.0507543087005615,
      "learning_rate": 3.926626175531586e-05,
      "loss": 0.1588,
      "num_input_tokens_seen": 9375064,
      "step": 14350
    },
    {
      "epoch": 7.523584905660377,
      "grad_norm": 2.5746631622314453,
      "learning_rate": 3.925687050804814e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 9377720,
      "step": 14355
    },
    {
      "epoch": 7.526205450733753,
      "grad_norm": 1.3082455396652222,
      "learning_rate": 3.924747627840575e-05,
      "loss": 0.218,
      "num_input_tokens_seen": 9380984,
      "step": 14360
    },
    {
      "epoch": 7.528825995807128,
      "grad_norm": 2.1556007862091064,
      "learning_rate": 3.923807906835388e-05,
      "loss": 0.1989,
      "num_input_tokens_seen": 9383992,
      "step": 14365
    },
    {
      "epoch": 7.531446540880503,
      "grad_norm": 3.654747247695923,
      "learning_rate": 3.92286788798583e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 9386360,
      "step": 14370
    },
    {
      "epoch": 7.534067085953878,
      "grad_norm": 2.285158634185791,
      "learning_rate": 3.921927571488543e-05,
      "loss": 0.2935,
      "num_input_tokens_seen": 9389144,
      "step": 14375
    },
    {
      "epoch": 7.536687631027254,
      "grad_norm": 1.3951915502548218,
      "learning_rate": 3.9209869575402296e-05,
      "loss": 0.1628,
      "num_input_tokens_seen": 9392024,
      "step": 14380
    },
    {
      "epoch": 7.539308176100629,
      "grad_norm": 2.7300426959991455,
      "learning_rate": 3.9200460463376566e-05,
      "loss": 0.2589,
      "num_input_tokens_seen": 9394872,
      "step": 14385
    },
    {
      "epoch": 7.5419287211740045,
      "grad_norm": 1.7863558530807495,
      "learning_rate": 3.9191048380776515e-05,
      "loss": 0.2685,
      "num_input_tokens_seen": 9397304,
      "step": 14390
    },
    {
      "epoch": 7.54454926624738,
      "grad_norm": 2.4436070919036865,
      "learning_rate": 3.918163332957104e-05,
      "loss": 0.2395,
      "num_input_tokens_seen": 9400664,
      "step": 14395
    },
    {
      "epoch": 7.547169811320755,
      "grad_norm": 2.065861701965332,
      "learning_rate": 3.9172215311729665e-05,
      "loss": 0.2783,
      "num_input_tokens_seen": 9403864,
      "step": 14400
    },
    {
      "epoch": 7.54979035639413,
      "grad_norm": 2.3473641872406006,
      "learning_rate": 3.916279432922254e-05,
      "loss": 0.2084,
      "num_input_tokens_seen": 9407544,
      "step": 14405
    },
    {
      "epoch": 7.552410901467505,
      "grad_norm": 1.3095663785934448,
      "learning_rate": 3.915337038402041e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 9410968,
      "step": 14410
    },
    {
      "epoch": 7.55503144654088,
      "grad_norm": 2.891296863555908,
      "learning_rate": 3.914394347809467e-05,
      "loss": 0.2571,
      "num_input_tokens_seen": 9413400,
      "step": 14415
    },
    {
      "epoch": 7.5576519916142555,
      "grad_norm": 1.7687398195266724,
      "learning_rate": 3.9134513613417315e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 9415832,
      "step": 14420
    },
    {
      "epoch": 7.560272536687631,
      "grad_norm": 2.132863998413086,
      "learning_rate": 3.912508079196097e-05,
      "loss": 0.3008,
      "num_input_tokens_seen": 9419064,
      "step": 14425
    },
    {
      "epoch": 7.562893081761006,
      "grad_norm": 2.0413637161254883,
      "learning_rate": 3.9115645015698866e-05,
      "loss": 0.2622,
      "num_input_tokens_seen": 9421752,
      "step": 14430
    },
    {
      "epoch": 7.565513626834382,
      "grad_norm": 1.6502292156219482,
      "learning_rate": 3.9106206286604865e-05,
      "loss": 0.2301,
      "num_input_tokens_seen": 9424984,
      "step": 14435
    },
    {
      "epoch": 7.568134171907757,
      "grad_norm": 1.5252158641815186,
      "learning_rate": 3.909676460665345e-05,
      "loss": 0.1832,
      "num_input_tokens_seen": 9428216,
      "step": 14440
    },
    {
      "epoch": 7.570754716981132,
      "grad_norm": 1.7483282089233398,
      "learning_rate": 3.908731997781969e-05,
      "loss": 0.2063,
      "num_input_tokens_seen": 9431576,
      "step": 14445
    },
    {
      "epoch": 7.573375262054507,
      "grad_norm": 4.217467308044434,
      "learning_rate": 3.907787240207932e-05,
      "loss": 0.2581,
      "num_input_tokens_seen": 9434680,
      "step": 14450
    },
    {
      "epoch": 7.575995807127883,
      "grad_norm": 2.7018678188323975,
      "learning_rate": 3.9068421881408645e-05,
      "loss": 0.1367,
      "num_input_tokens_seen": 9437432,
      "step": 14455
    },
    {
      "epoch": 7.578616352201258,
      "grad_norm": 1.0462374687194824,
      "learning_rate": 3.9058968417784614e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 9441208,
      "step": 14460
    },
    {
      "epoch": 7.581236897274633,
      "grad_norm": 2.3193209171295166,
      "learning_rate": 3.9049512013184785e-05,
      "loss": 0.2162,
      "num_input_tokens_seen": 9444056,
      "step": 14465
    },
    {
      "epoch": 7.583857442348008,
      "grad_norm": 6.115753173828125,
      "learning_rate": 3.9040052669587325e-05,
      "loss": 0.2689,
      "num_input_tokens_seen": 9446968,
      "step": 14470
    },
    {
      "epoch": 7.586477987421384,
      "grad_norm": 1.8804951906204224,
      "learning_rate": 3.9030590388971026e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 9450712,
      "step": 14475
    },
    {
      "epoch": 7.589098532494759,
      "grad_norm": 2.0062527656555176,
      "learning_rate": 3.902112517331529e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 9453240,
      "step": 14480
    },
    {
      "epoch": 7.5917190775681345,
      "grad_norm": 1.2903074026107788,
      "learning_rate": 3.901165702460012e-05,
      "loss": 0.1728,
      "num_input_tokens_seen": 9456728,
      "step": 14485
    },
    {
      "epoch": 7.59433962264151,
      "grad_norm": 3.269315004348755,
      "learning_rate": 3.900218594480616e-05,
      "loss": 0.2544,
      "num_input_tokens_seen": 9459800,
      "step": 14490
    },
    {
      "epoch": 7.596960167714885,
      "grad_norm": 1.372464656829834,
      "learning_rate": 3.899271193591465e-05,
      "loss": 0.1962,
      "num_input_tokens_seen": 9463384,
      "step": 14495
    },
    {
      "epoch": 7.59958071278826,
      "grad_norm": 2.8009192943573,
      "learning_rate": 3.8983234999907435e-05,
      "loss": 0.1869,
      "num_input_tokens_seen": 9466520,
      "step": 14500
    },
    {
      "epoch": 7.602201257861635,
      "grad_norm": 2.1053824424743652,
      "learning_rate": 3.897375513876699e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 9469944,
      "step": 14505
    },
    {
      "epoch": 7.60482180293501,
      "grad_norm": 2.06569242477417,
      "learning_rate": 3.8964272354476396e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 9472280,
      "step": 14510
    },
    {
      "epoch": 7.6074423480083855,
      "grad_norm": 2.3684966564178467,
      "learning_rate": 3.895478664901933e-05,
      "loss": 0.2288,
      "num_input_tokens_seen": 9474808,
      "step": 14515
    },
    {
      "epoch": 7.610062893081761,
      "grad_norm": 1.5382663011550903,
      "learning_rate": 3.89452980243801e-05,
      "loss": 0.2516,
      "num_input_tokens_seen": 9479064,
      "step": 14520
    },
    {
      "epoch": 7.612683438155136,
      "grad_norm": 1.5620179176330566,
      "learning_rate": 3.893580648254362e-05,
      "loss": 0.2682,
      "num_input_tokens_seen": 9483128,
      "step": 14525
    },
    {
      "epoch": 7.615303983228512,
      "grad_norm": 1.7626601457595825,
      "learning_rate": 3.8926312025495414e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 9486136,
      "step": 14530
    },
    {
      "epoch": 7.617924528301887,
      "grad_norm": 2.340705633163452,
      "learning_rate": 3.891681465522161e-05,
      "loss": 0.4412,
      "num_input_tokens_seen": 9488568,
      "step": 14535
    },
    {
      "epoch": 7.620545073375262,
      "grad_norm": 2.6018760204315186,
      "learning_rate": 3.890731437370896e-05,
      "loss": 0.2152,
      "num_input_tokens_seen": 9492440,
      "step": 14540
    },
    {
      "epoch": 7.623165618448637,
      "grad_norm": 1.6504559516906738,
      "learning_rate": 3.88978111829448e-05,
      "loss": 0.1767,
      "num_input_tokens_seen": 9495192,
      "step": 14545
    },
    {
      "epoch": 7.6257861635220126,
      "grad_norm": 2.8330743312835693,
      "learning_rate": 3.88883050849171e-05,
      "loss": 0.2581,
      "num_input_tokens_seen": 9497688,
      "step": 14550
    },
    {
      "epoch": 7.628406708595388,
      "grad_norm": 1.5644439458847046,
      "learning_rate": 3.8878796081614417e-05,
      "loss": 0.213,
      "num_input_tokens_seen": 9500984,
      "step": 14555
    },
    {
      "epoch": 7.631027253668763,
      "grad_norm": 2.454005002975464,
      "learning_rate": 3.8869284175025935e-05,
      "loss": 0.2235,
      "num_input_tokens_seen": 9504152,
      "step": 14560
    },
    {
      "epoch": 7.633647798742138,
      "grad_norm": 1.3543874025344849,
      "learning_rate": 3.885976936714143e-05,
      "loss": 0.2381,
      "num_input_tokens_seen": 9506840,
      "step": 14565
    },
    {
      "epoch": 7.636268343815514,
      "grad_norm": 1.7953954935073853,
      "learning_rate": 3.885025165995129e-05,
      "loss": 0.2581,
      "num_input_tokens_seen": 9509880,
      "step": 14570
    },
    {
      "epoch": 7.638888888888889,
      "grad_norm": 2.1188836097717285,
      "learning_rate": 3.884073105544652e-05,
      "loss": 0.1914,
      "num_input_tokens_seen": 9513464,
      "step": 14575
    },
    {
      "epoch": 7.6415094339622645,
      "grad_norm": 2.6921234130859375,
      "learning_rate": 3.8831207555618723e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 9516216,
      "step": 14580
    },
    {
      "epoch": 7.64412997903564,
      "grad_norm": 2.385012149810791,
      "learning_rate": 3.882168116246009e-05,
      "loss": 0.2028,
      "num_input_tokens_seen": 9519032,
      "step": 14585
    },
    {
      "epoch": 7.646750524109015,
      "grad_norm": 1.075760841369629,
      "learning_rate": 3.881215187796344e-05,
      "loss": 0.1937,
      "num_input_tokens_seen": 9523640,
      "step": 14590
    },
    {
      "epoch": 7.64937106918239,
      "grad_norm": 1.9219887256622314,
      "learning_rate": 3.8802619704122204e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 9526616,
      "step": 14595
    },
    {
      "epoch": 7.651991614255765,
      "grad_norm": 1.7310949563980103,
      "learning_rate": 3.8793084642930387e-05,
      "loss": 0.2351,
      "num_input_tokens_seen": 9529656,
      "step": 14600
    },
    {
      "epoch": 7.65461215932914,
      "grad_norm": 1.3954969644546509,
      "learning_rate": 3.878354669638263e-05,
      "loss": 0.2069,
      "num_input_tokens_seen": 9532984,
      "step": 14605
    },
    {
      "epoch": 7.6572327044025155,
      "grad_norm": 2.109367847442627,
      "learning_rate": 3.8774005866474155e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 9535928,
      "step": 14610
    },
    {
      "epoch": 7.659853249475891,
      "grad_norm": 1.3492944240570068,
      "learning_rate": 3.876446215520079e-05,
      "loss": 0.2455,
      "num_input_tokens_seen": 9539256,
      "step": 14615
    },
    {
      "epoch": 7.662473794549266,
      "grad_norm": 1.5622831583023071,
      "learning_rate": 3.875491556455898e-05,
      "loss": 0.2562,
      "num_input_tokens_seen": 9542296,
      "step": 14620
    },
    {
      "epoch": 7.665094339622642,
      "grad_norm": 0.9023475050926208,
      "learning_rate": 3.8745366096545754e-05,
      "loss": 0.1802,
      "num_input_tokens_seen": 9545368,
      "step": 14625
    },
    {
      "epoch": 7.667714884696017,
      "grad_norm": 1.1907565593719482,
      "learning_rate": 3.873581375315876e-05,
      "loss": 0.2037,
      "num_input_tokens_seen": 9548472,
      "step": 14630
    },
    {
      "epoch": 7.670335429769392,
      "grad_norm": 1.6268216371536255,
      "learning_rate": 3.872625853639623e-05,
      "loss": 0.1978,
      "num_input_tokens_seen": 9552248,
      "step": 14635
    },
    {
      "epoch": 7.672955974842767,
      "grad_norm": 1.655364751815796,
      "learning_rate": 3.871670044825701e-05,
      "loss": 0.409,
      "num_input_tokens_seen": 9555320,
      "step": 14640
    },
    {
      "epoch": 7.6755765199161425,
      "grad_norm": 1.569029688835144,
      "learning_rate": 3.870713949074056e-05,
      "loss": 0.2233,
      "num_input_tokens_seen": 9558904,
      "step": 14645
    },
    {
      "epoch": 7.678197064989518,
      "grad_norm": 2.530158281326294,
      "learning_rate": 3.86975756658469e-05,
      "loss": 0.2154,
      "num_input_tokens_seen": 9561624,
      "step": 14650
    },
    {
      "epoch": 7.680817610062893,
      "grad_norm": 4.270571231842041,
      "learning_rate": 3.868800897557667e-05,
      "loss": 0.3135,
      "num_input_tokens_seen": 9564568,
      "step": 14655
    },
    {
      "epoch": 7.683438155136268,
      "grad_norm": 1.4427671432495117,
      "learning_rate": 3.867843942193113e-05,
      "loss": 0.2303,
      "num_input_tokens_seen": 9567704,
      "step": 14660
    },
    {
      "epoch": 7.686058700209644,
      "grad_norm": 2.045862913131714,
      "learning_rate": 3.866886700691212e-05,
      "loss": 0.1966,
      "num_input_tokens_seen": 9571512,
      "step": 14665
    },
    {
      "epoch": 7.688679245283019,
      "grad_norm": 1.9962961673736572,
      "learning_rate": 3.8659291732522075e-05,
      "loss": 0.3102,
      "num_input_tokens_seen": 9575448,
      "step": 14670
    },
    {
      "epoch": 7.691299790356394,
      "grad_norm": 1.1003458499908447,
      "learning_rate": 3.8649713600764026e-05,
      "loss": 0.1396,
      "num_input_tokens_seen": 9583704,
      "step": 14675
    },
    {
      "epoch": 7.69392033542977,
      "grad_norm": 1.7181346416473389,
      "learning_rate": 3.864013261364162e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 9586392,
      "step": 14680
    },
    {
      "epoch": 7.696540880503145,
      "grad_norm": 3.5230960845947266,
      "learning_rate": 3.8630548773159084e-05,
      "loss": 0.3194,
      "num_input_tokens_seen": 9589336,
      "step": 14685
    },
    {
      "epoch": 7.69916142557652,
      "grad_norm": 1.039353847503662,
      "learning_rate": 3.862096208132125e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 9593016,
      "step": 14690
    },
    {
      "epoch": 7.701781970649895,
      "grad_norm": 1.925734519958496,
      "learning_rate": 3.861137254013355e-05,
      "loss": 0.2324,
      "num_input_tokens_seen": 9595512,
      "step": 14695
    },
    {
      "epoch": 7.70440251572327,
      "grad_norm": 1.7819671630859375,
      "learning_rate": 3.8601780151601996e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 9598904,
      "step": 14700
    },
    {
      "epoch": 7.7070230607966455,
      "grad_norm": 1.2717152833938599,
      "learning_rate": 3.859218491773321e-05,
      "loss": 0.1359,
      "num_input_tokens_seen": 9602712,
      "step": 14705
    },
    {
      "epoch": 7.709643605870021,
      "grad_norm": 1.311065673828125,
      "learning_rate": 3.858258684053441e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 9605560,
      "step": 14710
    },
    {
      "epoch": 7.712264150943396,
      "grad_norm": 1.568639874458313,
      "learning_rate": 3.8572985922013405e-05,
      "loss": 0.3292,
      "num_input_tokens_seen": 9609048,
      "step": 14715
    },
    {
      "epoch": 7.714884696016772,
      "grad_norm": 1.667833924293518,
      "learning_rate": 3.856338216417859e-05,
      "loss": 0.222,
      "num_input_tokens_seen": 9612824,
      "step": 14720
    },
    {
      "epoch": 7.717505241090147,
      "grad_norm": 0.7338225245475769,
      "learning_rate": 3.855377556903897e-05,
      "loss": 0.2,
      "num_input_tokens_seen": 9615704,
      "step": 14725
    },
    {
      "epoch": 7.720125786163522,
      "grad_norm": 19.042781829833984,
      "learning_rate": 3.8544166138604124e-05,
      "loss": 0.2452,
      "num_input_tokens_seen": 9618840,
      "step": 14730
    },
    {
      "epoch": 7.722746331236897,
      "grad_norm": 1.3249685764312744,
      "learning_rate": 3.8534553874884244e-05,
      "loss": 0.152,
      "num_input_tokens_seen": 9622552,
      "step": 14735
    },
    {
      "epoch": 7.7253668763102725,
      "grad_norm": 0.8832669258117676,
      "learning_rate": 3.852493877989011e-05,
      "loss": 0.2868,
      "num_input_tokens_seen": 9625720,
      "step": 14740
    },
    {
      "epoch": 7.727987421383648,
      "grad_norm": 2.3830726146698,
      "learning_rate": 3.851532085563308e-05,
      "loss": 0.2154,
      "num_input_tokens_seen": 9628664,
      "step": 14745
    },
    {
      "epoch": 7.730607966457023,
      "grad_norm": 2.5122969150543213,
      "learning_rate": 3.850570010412511e-05,
      "loss": 0.2223,
      "num_input_tokens_seen": 9632088,
      "step": 14750
    },
    {
      "epoch": 7.733228511530398,
      "grad_norm": 3.3421695232391357,
      "learning_rate": 3.8496076527378766e-05,
      "loss": 0.2216,
      "num_input_tokens_seen": 9635160,
      "step": 14755
    },
    {
      "epoch": 7.735849056603773,
      "grad_norm": 2.136333465576172,
      "learning_rate": 3.848645012740718e-05,
      "loss": 0.2439,
      "num_input_tokens_seen": 9638776,
      "step": 14760
    },
    {
      "epoch": 7.738469601677149,
      "grad_norm": 2.0719716548919678,
      "learning_rate": 3.8476820906224096e-05,
      "loss": 0.2275,
      "num_input_tokens_seen": 9641624,
      "step": 14765
    },
    {
      "epoch": 7.741090146750524,
      "grad_norm": 1.5639086961746216,
      "learning_rate": 3.846718886584382e-05,
      "loss": 0.2785,
      "num_input_tokens_seen": 9649176,
      "step": 14770
    },
    {
      "epoch": 7.7437106918239,
      "grad_norm": 3.1389577388763428,
      "learning_rate": 3.845755400828128e-05,
      "loss": 0.3375,
      "num_input_tokens_seen": 9651768,
      "step": 14775
    },
    {
      "epoch": 7.746331236897275,
      "grad_norm": 1.5485354661941528,
      "learning_rate": 3.844791633555198e-05,
      "loss": 0.1532,
      "num_input_tokens_seen": 9654552,
      "step": 14780
    },
    {
      "epoch": 7.74895178197065,
      "grad_norm": 1.3471102714538574,
      "learning_rate": 3.8438275849671996e-05,
      "loss": 0.1955,
      "num_input_tokens_seen": 9657112,
      "step": 14785
    },
    {
      "epoch": 7.751572327044025,
      "grad_norm": 0.8826507925987244,
      "learning_rate": 3.842863255265802e-05,
      "loss": 0.2152,
      "num_input_tokens_seen": 9661176,
      "step": 14790
    },
    {
      "epoch": 7.7541928721174,
      "grad_norm": 1.9574223756790161,
      "learning_rate": 3.841898644652732e-05,
      "loss": 0.2239,
      "num_input_tokens_seen": 9663480,
      "step": 14795
    },
    {
      "epoch": 7.756813417190775,
      "grad_norm": 1.4810538291931152,
      "learning_rate": 3.840933753329774e-05,
      "loss": 0.2273,
      "num_input_tokens_seen": 9666808,
      "step": 14800
    },
    {
      "epoch": 7.759433962264151,
      "grad_norm": 1.7395572662353516,
      "learning_rate": 3.839968581498773e-05,
      "loss": 0.2955,
      "num_input_tokens_seen": 9669336,
      "step": 14805
    },
    {
      "epoch": 7.762054507337526,
      "grad_norm": 1.7723560333251953,
      "learning_rate": 3.839003129361632e-05,
      "loss": 0.2195,
      "num_input_tokens_seen": 9671928,
      "step": 14810
    },
    {
      "epoch": 7.764675052410902,
      "grad_norm": 1.3046278953552246,
      "learning_rate": 3.8380373971203134e-05,
      "loss": 0.3004,
      "num_input_tokens_seen": 9676120,
      "step": 14815
    },
    {
      "epoch": 7.767295597484277,
      "grad_norm": 1.8652433156967163,
      "learning_rate": 3.8370713849768346e-05,
      "loss": 0.2057,
      "num_input_tokens_seen": 9679032,
      "step": 14820
    },
    {
      "epoch": 7.769916142557652,
      "grad_norm": 1.7044767141342163,
      "learning_rate": 3.8361050931332774e-05,
      "loss": 0.1933,
      "num_input_tokens_seen": 9682744,
      "step": 14825
    },
    {
      "epoch": 7.772536687631027,
      "grad_norm": 2.7500641345977783,
      "learning_rate": 3.835138521791778e-05,
      "loss": 0.3217,
      "num_input_tokens_seen": 9686040,
      "step": 14830
    },
    {
      "epoch": 7.7751572327044025,
      "grad_norm": 2.565544843673706,
      "learning_rate": 3.8341716711545314e-05,
      "loss": 0.2262,
      "num_input_tokens_seen": 9690168,
      "step": 14835
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 1.6125187873840332,
      "learning_rate": 3.833204541423793e-05,
      "loss": 0.2229,
      "num_input_tokens_seen": 9693496,
      "step": 14840
    },
    {
      "epoch": 7.780398322851153,
      "grad_norm": 2.3027303218841553,
      "learning_rate": 3.832237132801874e-05,
      "loss": 0.2331,
      "num_input_tokens_seen": 9696760,
      "step": 14845
    },
    {
      "epoch": 7.783018867924528,
      "grad_norm": 1.140775442123413,
      "learning_rate": 3.831269445491146e-05,
      "loss": 0.2907,
      "num_input_tokens_seen": 9699672,
      "step": 14850
    },
    {
      "epoch": 7.785639412997903,
      "grad_norm": 2.6290647983551025,
      "learning_rate": 3.8303014796940375e-05,
      "loss": 0.1893,
      "num_input_tokens_seen": 9703384,
      "step": 14855
    },
    {
      "epoch": 7.788259958071279,
      "grad_norm": 1.452976107597351,
      "learning_rate": 3.829333235613036e-05,
      "loss": 0.2846,
      "num_input_tokens_seen": 9706328,
      "step": 14860
    },
    {
      "epoch": 7.790880503144654,
      "grad_norm": 2.281940460205078,
      "learning_rate": 3.828364713450688e-05,
      "loss": 0.2489,
      "num_input_tokens_seen": 9709432,
      "step": 14865
    },
    {
      "epoch": 7.79350104821803,
      "grad_norm": 2.969766139984131,
      "learning_rate": 3.827395913409597e-05,
      "loss": 0.3318,
      "num_input_tokens_seen": 9712696,
      "step": 14870
    },
    {
      "epoch": 7.796121593291405,
      "grad_norm": 2.2071874141693115,
      "learning_rate": 3.826426835692424e-05,
      "loss": 0.2325,
      "num_input_tokens_seen": 9715704,
      "step": 14875
    },
    {
      "epoch": 7.79874213836478,
      "grad_norm": 2.201414108276367,
      "learning_rate": 3.825457480501891e-05,
      "loss": 0.4263,
      "num_input_tokens_seen": 9720344,
      "step": 14880
    },
    {
      "epoch": 7.801362683438155,
      "grad_norm": 1.0216816663742065,
      "learning_rate": 3.824487848040773e-05,
      "loss": 0.2347,
      "num_input_tokens_seen": 9723544,
      "step": 14885
    },
    {
      "epoch": 7.80398322851153,
      "grad_norm": 1.6980414390563965,
      "learning_rate": 3.823517938511908e-05,
      "loss": 0.3844,
      "num_input_tokens_seen": 9726424,
      "step": 14890
    },
    {
      "epoch": 7.806603773584905,
      "grad_norm": 1.9803630113601685,
      "learning_rate": 3.8225477521181906e-05,
      "loss": 0.1825,
      "num_input_tokens_seen": 9729144,
      "step": 14895
    },
    {
      "epoch": 7.809224318658281,
      "grad_norm": 1.5697550773620605,
      "learning_rate": 3.821577289062571e-05,
      "loss": 0.1993,
      "num_input_tokens_seen": 9732504,
      "step": 14900
    },
    {
      "epoch": 7.811844863731656,
      "grad_norm": 3.469433307647705,
      "learning_rate": 3.82060654954806e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 9737048,
      "step": 14905
    },
    {
      "epoch": 7.814465408805032,
      "grad_norm": 3.384016752243042,
      "learning_rate": 3.819635533777726e-05,
      "loss": 0.3008,
      "num_input_tokens_seen": 9739960,
      "step": 14910
    },
    {
      "epoch": 7.817085953878407,
      "grad_norm": 1.6319843530654907,
      "learning_rate": 3.8186642419546924e-05,
      "loss": 0.1599,
      "num_input_tokens_seen": 9743288,
      "step": 14915
    },
    {
      "epoch": 7.819706498951782,
      "grad_norm": 2.1579723358154297,
      "learning_rate": 3.817692674282145e-05,
      "loss": 0.3448,
      "num_input_tokens_seen": 9746072,
      "step": 14920
    },
    {
      "epoch": 7.822327044025157,
      "grad_norm": 2.084280490875244,
      "learning_rate": 3.816720830963322e-05,
      "loss": 0.2869,
      "num_input_tokens_seen": 9748920,
      "step": 14925
    },
    {
      "epoch": 7.8249475890985325,
      "grad_norm": 1.345794677734375,
      "learning_rate": 3.815748712201523e-05,
      "loss": 0.1803,
      "num_input_tokens_seen": 9751864,
      "step": 14930
    },
    {
      "epoch": 7.827568134171908,
      "grad_norm": 2.1169872283935547,
      "learning_rate": 3.814776318200105e-05,
      "loss": 0.2632,
      "num_input_tokens_seen": 9755128,
      "step": 14935
    },
    {
      "epoch": 7.830188679245283,
      "grad_norm": 1.4828298091888428,
      "learning_rate": 3.8138036491624805e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 9757976,
      "step": 14940
    },
    {
      "epoch": 7.832809224318658,
      "grad_norm": 1.1369465589523315,
      "learning_rate": 3.812830705292122e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 9761464,
      "step": 14945
    },
    {
      "epoch": 7.835429769392033,
      "grad_norm": 2.957228899002075,
      "learning_rate": 3.8118574867925564e-05,
      "loss": 0.2424,
      "num_input_tokens_seen": 9768408,
      "step": 14950
    },
    {
      "epoch": 7.838050314465409,
      "grad_norm": 1.4949733018875122,
      "learning_rate": 3.810883993867371e-05,
      "loss": 0.2422,
      "num_input_tokens_seen": 9771832,
      "step": 14955
    },
    {
      "epoch": 7.840670859538784,
      "grad_norm": 2.021756887435913,
      "learning_rate": 3.80991022672021e-05,
      "loss": 0.2802,
      "num_input_tokens_seen": 9775736,
      "step": 14960
    },
    {
      "epoch": 7.84329140461216,
      "grad_norm": 3.18558669090271,
      "learning_rate": 3.8089361855547744e-05,
      "loss": 0.2636,
      "num_input_tokens_seen": 9779096,
      "step": 14965
    },
    {
      "epoch": 7.845911949685535,
      "grad_norm": 2.994007110595703,
      "learning_rate": 3.8079618705748214e-05,
      "loss": 0.3663,
      "num_input_tokens_seen": 9782264,
      "step": 14970
    },
    {
      "epoch": 7.84853249475891,
      "grad_norm": 1.8455685377120972,
      "learning_rate": 3.806987281984167e-05,
      "loss": 0.197,
      "num_input_tokens_seen": 9785368,
      "step": 14975
    },
    {
      "epoch": 7.851153039832285,
      "grad_norm": 2.352405548095703,
      "learning_rate": 3.806012419986684e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 9789304,
      "step": 14980
    },
    {
      "epoch": 7.85377358490566,
      "grad_norm": 2.1267809867858887,
      "learning_rate": 3.805037284786303e-05,
      "loss": 0.3164,
      "num_input_tokens_seen": 9792184,
      "step": 14985
    },
    {
      "epoch": 7.856394129979035,
      "grad_norm": 1.6215153932571411,
      "learning_rate": 3.804061876587009e-05,
      "loss": 0.305,
      "num_input_tokens_seen": 9795928,
      "step": 14990
    },
    {
      "epoch": 7.859014675052411,
      "grad_norm": 1.3033415079116821,
      "learning_rate": 3.80308619559285e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 9798776,
      "step": 14995
    },
    {
      "epoch": 7.861635220125786,
      "grad_norm": 1.5631524324417114,
      "learning_rate": 3.802110242007925e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 9802008,
      "step": 15000
    },
    {
      "epoch": 7.864255765199162,
      "grad_norm": 2.605191230773926,
      "learning_rate": 3.801134016036392e-05,
      "loss": 0.3106,
      "num_input_tokens_seen": 9805912,
      "step": 15005
    },
    {
      "epoch": 7.866876310272537,
      "grad_norm": 2.1150145530700684,
      "learning_rate": 3.8001575178824665e-05,
      "loss": 0.2187,
      "num_input_tokens_seen": 9808664,
      "step": 15010
    },
    {
      "epoch": 7.869496855345912,
      "grad_norm": 1.1386685371398926,
      "learning_rate": 3.799180747750422e-05,
      "loss": 0.1975,
      "num_input_tokens_seen": 9811256,
      "step": 15015
    },
    {
      "epoch": 7.872117400419287,
      "grad_norm": 1.683314561843872,
      "learning_rate": 3.798203705844586e-05,
      "loss": 0.2564,
      "num_input_tokens_seen": 9816440,
      "step": 15020
    },
    {
      "epoch": 7.8747379454926625,
      "grad_norm": 2.382423162460327,
      "learning_rate": 3.7972263923693476e-05,
      "loss": 0.2182,
      "num_input_tokens_seen": 9818776,
      "step": 15025
    },
    {
      "epoch": 7.877358490566038,
      "grad_norm": 2.46862530708313,
      "learning_rate": 3.796248807529145e-05,
      "loss": 0.238,
      "num_input_tokens_seen": 9822520,
      "step": 15030
    },
    {
      "epoch": 7.879979035639413,
      "grad_norm": 1.3654422760009766,
      "learning_rate": 3.795270951528482e-05,
      "loss": 0.1493,
      "num_input_tokens_seen": 9825496,
      "step": 15035
    },
    {
      "epoch": 7.882599580712788,
      "grad_norm": 1.674444317817688,
      "learning_rate": 3.7942928245719124e-05,
      "loss": 0.3192,
      "num_input_tokens_seen": 9829080,
      "step": 15040
    },
    {
      "epoch": 7.885220125786163,
      "grad_norm": 2.5293941497802734,
      "learning_rate": 3.7933144268640495e-05,
      "loss": 0.2299,
      "num_input_tokens_seen": 9832056,
      "step": 15045
    },
    {
      "epoch": 7.887840670859539,
      "grad_norm": 2.538691997528076,
      "learning_rate": 3.7923357586095645e-05,
      "loss": 0.209,
      "num_input_tokens_seen": 9834680,
      "step": 15050
    },
    {
      "epoch": 7.890461215932914,
      "grad_norm": 1.2304482460021973,
      "learning_rate": 3.7913568200131823e-05,
      "loss": 0.2884,
      "num_input_tokens_seen": 9838008,
      "step": 15055
    },
    {
      "epoch": 7.8930817610062896,
      "grad_norm": 3.684567451477051,
      "learning_rate": 3.790377611279685e-05,
      "loss": 0.2781,
      "num_input_tokens_seen": 9841528,
      "step": 15060
    },
    {
      "epoch": 7.895702306079665,
      "grad_norm": 3.2702035903930664,
      "learning_rate": 3.789398132613914e-05,
      "loss": 0.3515,
      "num_input_tokens_seen": 9844984,
      "step": 15065
    },
    {
      "epoch": 7.89832285115304,
      "grad_norm": 1.5911294221878052,
      "learning_rate": 3.788418384220763e-05,
      "loss": 0.201,
      "num_input_tokens_seen": 9848568,
      "step": 15070
    },
    {
      "epoch": 7.900943396226415,
      "grad_norm": 1.9357757568359375,
      "learning_rate": 3.7874383663051846e-05,
      "loss": 0.1495,
      "num_input_tokens_seen": 9851640,
      "step": 15075
    },
    {
      "epoch": 7.90356394129979,
      "grad_norm": 4.427441120147705,
      "learning_rate": 3.7864580790721894e-05,
      "loss": 0.1837,
      "num_input_tokens_seen": 9855192,
      "step": 15080
    },
    {
      "epoch": 7.906184486373165,
      "grad_norm": 1.7047239542007446,
      "learning_rate": 3.785477522726839e-05,
      "loss": 0.2397,
      "num_input_tokens_seen": 9858072,
      "step": 15085
    },
    {
      "epoch": 7.908805031446541,
      "grad_norm": 1.56654691696167,
      "learning_rate": 3.784496697474258e-05,
      "loss": 0.2038,
      "num_input_tokens_seen": 9861784,
      "step": 15090
    },
    {
      "epoch": 7.911425576519916,
      "grad_norm": 2.4225411415100098,
      "learning_rate": 3.783515603519622e-05,
      "loss": 0.2947,
      "num_input_tokens_seen": 9866104,
      "step": 15095
    },
    {
      "epoch": 7.914046121593291,
      "grad_norm": 3.7281808853149414,
      "learning_rate": 3.782534241068164e-05,
      "loss": 0.2928,
      "num_input_tokens_seen": 9869816,
      "step": 15100
    },
    {
      "epoch": 7.916666666666667,
      "grad_norm": 1.3608953952789307,
      "learning_rate": 3.781552610325175e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 9872504,
      "step": 15105
    },
    {
      "epoch": 7.919287211740042,
      "grad_norm": 0.8689777255058289,
      "learning_rate": 3.780570711496001e-05,
      "loss": 0.2319,
      "num_input_tokens_seen": 9875320,
      "step": 15110
    },
    {
      "epoch": 7.921907756813417,
      "grad_norm": 2.1943085193634033,
      "learning_rate": 3.779588544786044e-05,
      "loss": 0.206,
      "num_input_tokens_seen": 9878136,
      "step": 15115
    },
    {
      "epoch": 7.9245283018867925,
      "grad_norm": 2.0602056980133057,
      "learning_rate": 3.778606110400761e-05,
      "loss": 0.1995,
      "num_input_tokens_seen": 9880792,
      "step": 15120
    },
    {
      "epoch": 7.927148846960168,
      "grad_norm": 1.6929714679718018,
      "learning_rate": 3.7776234085456665e-05,
      "loss": 0.2247,
      "num_input_tokens_seen": 9884312,
      "step": 15125
    },
    {
      "epoch": 7.929769392033543,
      "grad_norm": 2.1597468852996826,
      "learning_rate": 3.776640439426332e-05,
      "loss": 0.2319,
      "num_input_tokens_seen": 9887416,
      "step": 15130
    },
    {
      "epoch": 7.932389937106918,
      "grad_norm": 1.7769081592559814,
      "learning_rate": 3.775657203248381e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 9890200,
      "step": 15135
    },
    {
      "epoch": 7.935010482180293,
      "grad_norm": 2.767322540283203,
      "learning_rate": 3.774673700217497e-05,
      "loss": 0.22,
      "num_input_tokens_seen": 9893368,
      "step": 15140
    },
    {
      "epoch": 7.937631027253669,
      "grad_norm": 3.313236713409424,
      "learning_rate": 3.773689930539418e-05,
      "loss": 0.2732,
      "num_input_tokens_seen": 9899224,
      "step": 15145
    },
    {
      "epoch": 7.940251572327044,
      "grad_norm": 2.9846808910369873,
      "learning_rate": 3.7727058944199354e-05,
      "loss": 0.2631,
      "num_input_tokens_seen": 9901944,
      "step": 15150
    },
    {
      "epoch": 7.9428721174004195,
      "grad_norm": 1.319358229637146,
      "learning_rate": 3.7717215920648994e-05,
      "loss": 0.3045,
      "num_input_tokens_seen": 9905720,
      "step": 15155
    },
    {
      "epoch": 7.945492662473795,
      "grad_norm": 1.8900494575500488,
      "learning_rate": 3.770737023680215e-05,
      "loss": 0.3141,
      "num_input_tokens_seen": 9908440,
      "step": 15160
    },
    {
      "epoch": 7.94811320754717,
      "grad_norm": 1.918620228767395,
      "learning_rate": 3.7697521894718436e-05,
      "loss": 0.258,
      "num_input_tokens_seen": 9911256,
      "step": 15165
    },
    {
      "epoch": 7.950733752620545,
      "grad_norm": 1.7009012699127197,
      "learning_rate": 3.7687670896458e-05,
      "loss": 0.2465,
      "num_input_tokens_seen": 9913336,
      "step": 15170
    },
    {
      "epoch": 7.95335429769392,
      "grad_norm": 1.0160871744155884,
      "learning_rate": 3.7677817244081556e-05,
      "loss": 0.1849,
      "num_input_tokens_seen": 9917048,
      "step": 15175
    },
    {
      "epoch": 7.955974842767295,
      "grad_norm": 3.6258060932159424,
      "learning_rate": 3.766796093965038e-05,
      "loss": 0.1895,
      "num_input_tokens_seen": 9920952,
      "step": 15180
    },
    {
      "epoch": 7.9585953878406706,
      "grad_norm": 1.9090543985366821,
      "learning_rate": 3.76581019852263e-05,
      "loss": 0.1694,
      "num_input_tokens_seen": 9924376,
      "step": 15185
    },
    {
      "epoch": 7.961215932914046,
      "grad_norm": 1.5857913494110107,
      "learning_rate": 3.7648240382871705e-05,
      "loss": 0.2556,
      "num_input_tokens_seen": 9927160,
      "step": 15190
    },
    {
      "epoch": 7.963836477987421,
      "grad_norm": 2.2250077724456787,
      "learning_rate": 3.763837613464951e-05,
      "loss": 0.2523,
      "num_input_tokens_seen": 9930840,
      "step": 15195
    },
    {
      "epoch": 7.966457023060797,
      "grad_norm": 3.100114345550537,
      "learning_rate": 3.7628509242623215e-05,
      "loss": 0.3002,
      "num_input_tokens_seen": 9934104,
      "step": 15200
    },
    {
      "epoch": 7.969077568134172,
      "grad_norm": 2.9997856616973877,
      "learning_rate": 3.761863970885686e-05,
      "loss": 0.2085,
      "num_input_tokens_seen": 9937016,
      "step": 15205
    },
    {
      "epoch": 7.971698113207547,
      "grad_norm": 4.593401908874512,
      "learning_rate": 3.760876753541504e-05,
      "loss": 0.2692,
      "num_input_tokens_seen": 9939736,
      "step": 15210
    },
    {
      "epoch": 7.9743186582809225,
      "grad_norm": 1.7351834774017334,
      "learning_rate": 3.7598892724362907e-05,
      "loss": 0.2496,
      "num_input_tokens_seen": 9942744,
      "step": 15215
    },
    {
      "epoch": 7.976939203354298,
      "grad_norm": 5.264279842376709,
      "learning_rate": 3.758901527776615e-05,
      "loss": 0.2761,
      "num_input_tokens_seen": 9946776,
      "step": 15220
    },
    {
      "epoch": 7.979559748427673,
      "grad_norm": 2.786059617996216,
      "learning_rate": 3.7579135197691015e-05,
      "loss": 0.2133,
      "num_input_tokens_seen": 9949304,
      "step": 15225
    },
    {
      "epoch": 7.982180293501048,
      "grad_norm": 1.802316665649414,
      "learning_rate": 3.7569252486204306e-05,
      "loss": 0.2267,
      "num_input_tokens_seen": 9953112,
      "step": 15230
    },
    {
      "epoch": 7.984800838574423,
      "grad_norm": 1.6626348495483398,
      "learning_rate": 3.755936714537338e-05,
      "loss": 0.2656,
      "num_input_tokens_seen": 9956056,
      "step": 15235
    },
    {
      "epoch": 7.987421383647799,
      "grad_norm": 5.79463529586792,
      "learning_rate": 3.754947917726612e-05,
      "loss": 0.2382,
      "num_input_tokens_seen": 9959416,
      "step": 15240
    },
    {
      "epoch": 7.990041928721174,
      "grad_norm": 2.4345858097076416,
      "learning_rate": 3.7539588583951e-05,
      "loss": 0.1839,
      "num_input_tokens_seen": 9962680,
      "step": 15245
    },
    {
      "epoch": 7.9926624737945495,
      "grad_norm": 1.2890878915786743,
      "learning_rate": 3.752969536749701e-05,
      "loss": 0.3443,
      "num_input_tokens_seen": 9966616,
      "step": 15250
    },
    {
      "epoch": 7.995283018867925,
      "grad_norm": 1.4950306415557861,
      "learning_rate": 3.7519799529973685e-05,
      "loss": 0.2542,
      "num_input_tokens_seen": 9971864,
      "step": 15255
    },
    {
      "epoch": 7.9979035639413,
      "grad_norm": 3.7759883403778076,
      "learning_rate": 3.7509901073451134e-05,
      "loss": 0.4997,
      "num_input_tokens_seen": 9975864,
      "step": 15260
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.6323651075363159,
      "eval_runtime": 11.7829,
      "eval_samples_per_second": 71.969,
      "eval_steps_per_second": 17.992,
      "num_input_tokens_seen": 9977520,
      "step": 15264
    },
    {
      "epoch": 8.000524109014675,
      "grad_norm": 2.1205413341522217,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.2409,
      "num_input_tokens_seen": 9978032,
      "step": 15265
    },
    {
      "epoch": 8.00314465408805,
      "grad_norm": 1.4914593696594238,
      "learning_rate": 3.749009631169147e-05,
      "loss": 0.2744,
      "num_input_tokens_seen": 9981616,
      "step": 15270
    },
    {
      "epoch": 8.005765199161425,
      "grad_norm": 0.9481394290924072,
      "learning_rate": 3.74801900105973e-05,
      "loss": 0.1285,
      "num_input_tokens_seen": 9984624,
      "step": 15275
    },
    {
      "epoch": 8.0083857442348,
      "grad_norm": 1.2564576864242554,
      "learning_rate": 3.747028109878975e-05,
      "loss": 0.1645,
      "num_input_tokens_seen": 9987568,
      "step": 15280
    },
    {
      "epoch": 8.011006289308176,
      "grad_norm": 3.5048916339874268,
      "learning_rate": 3.746036957834167e-05,
      "loss": 0.1591,
      "num_input_tokens_seen": 9991280,
      "step": 15285
    },
    {
      "epoch": 8.01362683438155,
      "grad_norm": 2.8030447959899902,
      "learning_rate": 3.7450455451326415e-05,
      "loss": 0.1619,
      "num_input_tokens_seen": 9994288,
      "step": 15290
    },
    {
      "epoch": 8.016247379454926,
      "grad_norm": 2.137347459793091,
      "learning_rate": 3.7440538719817935e-05,
      "loss": 0.1383,
      "num_input_tokens_seen": 9997200,
      "step": 15295
    },
    {
      "epoch": 8.018867924528301,
      "grad_norm": 3.9244117736816406,
      "learning_rate": 3.743061938589068e-05,
      "loss": 0.2369,
      "num_input_tokens_seen": 10000176,
      "step": 15300
    },
    {
      "epoch": 8.021488469601676,
      "grad_norm": 1.4840501546859741,
      "learning_rate": 3.742069745161966e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 10003504,
      "step": 15305
    },
    {
      "epoch": 8.024109014675052,
      "grad_norm": 4.614225387573242,
      "learning_rate": 3.741077291908045e-05,
      "loss": 0.1829,
      "num_input_tokens_seen": 10008688,
      "step": 15310
    },
    {
      "epoch": 8.026729559748428,
      "grad_norm": 1.7257683277130127,
      "learning_rate": 3.740084579034913e-05,
      "loss": 0.2083,
      "num_input_tokens_seen": 10011536,
      "step": 15315
    },
    {
      "epoch": 8.029350104821804,
      "grad_norm": 2.552321672439575,
      "learning_rate": 3.739091606750234e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 10014096,
      "step": 15320
    },
    {
      "epoch": 8.031970649895179,
      "grad_norm": 1.0359766483306885,
      "learning_rate": 3.7380983752617275e-05,
      "loss": 0.2121,
      "num_input_tokens_seen": 10017360,
      "step": 15325
    },
    {
      "epoch": 8.034591194968554,
      "grad_norm": 1.7853480577468872,
      "learning_rate": 3.737104884777166e-05,
      "loss": 0.2177,
      "num_input_tokens_seen": 10020560,
      "step": 15330
    },
    {
      "epoch": 8.03721174004193,
      "grad_norm": 4.943023681640625,
      "learning_rate": 3.736111135504376e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 10024176,
      "step": 15335
    },
    {
      "epoch": 8.039832285115304,
      "grad_norm": 1.010615348815918,
      "learning_rate": 3.735117127651239e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 10028336,
      "step": 15340
    },
    {
      "epoch": 8.04245283018868,
      "grad_norm": 2.0860748291015625,
      "learning_rate": 3.734122861425689e-05,
      "loss": 0.1838,
      "num_input_tokens_seen": 10031376,
      "step": 15345
    },
    {
      "epoch": 8.045073375262055,
      "grad_norm": 1.885791540145874,
      "learning_rate": 3.7331283370357173e-05,
      "loss": 0.1413,
      "num_input_tokens_seen": 10033904,
      "step": 15350
    },
    {
      "epoch": 8.04769392033543,
      "grad_norm": 1.5283700227737427,
      "learning_rate": 3.732133554689365e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 10036720,
      "step": 15355
    },
    {
      "epoch": 8.050314465408805,
      "grad_norm": 2.1328608989715576,
      "learning_rate": 3.73113851459473e-05,
      "loss": 0.1542,
      "num_input_tokens_seen": 10040016,
      "step": 15360
    },
    {
      "epoch": 8.05293501048218,
      "grad_norm": 3.3511805534362793,
      "learning_rate": 3.730143216959964e-05,
      "loss": 0.2229,
      "num_input_tokens_seen": 10043088,
      "step": 15365
    },
    {
      "epoch": 8.055555555555555,
      "grad_norm": 2.4066975116729736,
      "learning_rate": 3.729147661993271e-05,
      "loss": 0.2234,
      "num_input_tokens_seen": 10046160,
      "step": 15370
    },
    {
      "epoch": 8.05817610062893,
      "grad_norm": 1.5189944505691528,
      "learning_rate": 3.728151849902909e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 10048592,
      "step": 15375
    },
    {
      "epoch": 8.060796645702306,
      "grad_norm": 2.544337749481201,
      "learning_rate": 3.727155780897193e-05,
      "loss": 0.1666,
      "num_input_tokens_seen": 10051536,
      "step": 15380
    },
    {
      "epoch": 8.06341719077568,
      "grad_norm": 1.0828560590744019,
      "learning_rate": 3.7261594551844865e-05,
      "loss": 0.2317,
      "num_input_tokens_seen": 10055632,
      "step": 15385
    },
    {
      "epoch": 8.066037735849056,
      "grad_norm": 1.5634790658950806,
      "learning_rate": 3.7251628729732116e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 10060528,
      "step": 15390
    },
    {
      "epoch": 8.068658280922431,
      "grad_norm": 2.056555986404419,
      "learning_rate": 3.724166034471841e-05,
      "loss": 0.1001,
      "num_input_tokens_seen": 10063664,
      "step": 15395
    },
    {
      "epoch": 8.071278825995806,
      "grad_norm": 4.41472053527832,
      "learning_rate": 3.723168939888901e-05,
      "loss": 0.177,
      "num_input_tokens_seen": 10066288,
      "step": 15400
    },
    {
      "epoch": 8.073899371069182,
      "grad_norm": 2.3733227252960205,
      "learning_rate": 3.7221715894329747e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 10070064,
      "step": 15405
    },
    {
      "epoch": 8.076519916142558,
      "grad_norm": 2.214662551879883,
      "learning_rate": 3.7211739833126945e-05,
      "loss": 0.1381,
      "num_input_tokens_seen": 10074128,
      "step": 15410
    },
    {
      "epoch": 8.079140461215934,
      "grad_norm": 2.236394166946411,
      "learning_rate": 3.720176121736748e-05,
      "loss": 0.1308,
      "num_input_tokens_seen": 10077936,
      "step": 15415
    },
    {
      "epoch": 8.081761006289309,
      "grad_norm": 1.8825331926345825,
      "learning_rate": 3.719178004913879e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 10080976,
      "step": 15420
    },
    {
      "epoch": 8.084381551362684,
      "grad_norm": 1.8603942394256592,
      "learning_rate": 3.7181796330528796e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 10083984,
      "step": 15425
    },
    {
      "epoch": 8.08700209643606,
      "grad_norm": 3.764289379119873,
      "learning_rate": 3.717181006362599e-05,
      "loss": 0.1991,
      "num_input_tokens_seen": 10086800,
      "step": 15430
    },
    {
      "epoch": 8.089622641509434,
      "grad_norm": 1.3515328168869019,
      "learning_rate": 3.7161821250519384e-05,
      "loss": 0.1415,
      "num_input_tokens_seen": 10090544,
      "step": 15435
    },
    {
      "epoch": 8.09224318658281,
      "grad_norm": 2.6404895782470703,
      "learning_rate": 3.7151829893298515e-05,
      "loss": 0.1282,
      "num_input_tokens_seen": 10093360,
      "step": 15440
    },
    {
      "epoch": 8.094863731656185,
      "grad_norm": 1.8979902267456055,
      "learning_rate": 3.714183599405347e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 10097520,
      "step": 15445
    },
    {
      "epoch": 8.09748427672956,
      "grad_norm": 0.8966214656829834,
      "learning_rate": 3.713183955487487e-05,
      "loss": 0.1246,
      "num_input_tokens_seen": 10100240,
      "step": 15450
    },
    {
      "epoch": 8.100104821802935,
      "grad_norm": 2.261136531829834,
      "learning_rate": 3.7121840577853815e-05,
      "loss": 0.1923,
      "num_input_tokens_seen": 10103280,
      "step": 15455
    },
    {
      "epoch": 8.10272536687631,
      "grad_norm": 1.5878069400787354,
      "learning_rate": 3.7111839065082027e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 10106288,
      "step": 15460
    },
    {
      "epoch": 8.105345911949685,
      "grad_norm": 5.09915018081665,
      "learning_rate": 3.710183501865168e-05,
      "loss": 0.1462,
      "num_input_tokens_seen": 10110064,
      "step": 15465
    },
    {
      "epoch": 8.10796645702306,
      "grad_norm": 1.4789156913757324,
      "learning_rate": 3.7091828440655514e-05,
      "loss": 0.1322,
      "num_input_tokens_seen": 10113360,
      "step": 15470
    },
    {
      "epoch": 8.110587002096436,
      "grad_norm": 1.5528197288513184,
      "learning_rate": 3.708181933318678e-05,
      "loss": 0.1872,
      "num_input_tokens_seen": 10119760,
      "step": 15475
    },
    {
      "epoch": 8.11320754716981,
      "grad_norm": 2.8438873291015625,
      "learning_rate": 3.707180769833929e-05,
      "loss": 0.1077,
      "num_input_tokens_seen": 10126864,
      "step": 15480
    },
    {
      "epoch": 8.115828092243186,
      "grad_norm": 2.0285778045654297,
      "learning_rate": 3.706179353820734e-05,
      "loss": 0.1156,
      "num_input_tokens_seen": 10129776,
      "step": 15485
    },
    {
      "epoch": 8.118448637316561,
      "grad_norm": 7.281335830688477,
      "learning_rate": 3.7051776854885794e-05,
      "loss": 0.1479,
      "num_input_tokens_seen": 10132816,
      "step": 15490
    },
    {
      "epoch": 8.121069182389936,
      "grad_norm": 1.5642303228378296,
      "learning_rate": 3.7041757650470025e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 10136112,
      "step": 15495
    },
    {
      "epoch": 8.123689727463312,
      "grad_norm": 1.7675342559814453,
      "learning_rate": 3.703173592705594e-05,
      "loss": 0.1756,
      "num_input_tokens_seen": 10138896,
      "step": 15500
    },
    {
      "epoch": 8.126310272536688,
      "grad_norm": 2.028414011001587,
      "learning_rate": 3.702171168673996e-05,
      "loss": 0.2267,
      "num_input_tokens_seen": 10141456,
      "step": 15505
    },
    {
      "epoch": 8.128930817610064,
      "grad_norm": 2.4960951805114746,
      "learning_rate": 3.701168493161905e-05,
      "loss": 0.2103,
      "num_input_tokens_seen": 10144944,
      "step": 15510
    },
    {
      "epoch": 8.131551362683439,
      "grad_norm": 2.034813404083252,
      "learning_rate": 3.7001655663790676e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 10148016,
      "step": 15515
    },
    {
      "epoch": 8.134171907756814,
      "grad_norm": 2.0351130962371826,
      "learning_rate": 3.699162388535288e-05,
      "loss": 0.1946,
      "num_input_tokens_seen": 10152048,
      "step": 15520
    },
    {
      "epoch": 8.13679245283019,
      "grad_norm": 1.786028504371643,
      "learning_rate": 3.698158959840416e-05,
      "loss": 0.1642,
      "num_input_tokens_seen": 10155600,
      "step": 15525
    },
    {
      "epoch": 8.139412997903564,
      "grad_norm": 2.8249013423919678,
      "learning_rate": 3.6971552805043585e-05,
      "loss": 0.1904,
      "num_input_tokens_seen": 10158160,
      "step": 15530
    },
    {
      "epoch": 8.14203354297694,
      "grad_norm": 1.9258183240890503,
      "learning_rate": 3.6961513507370754e-05,
      "loss": 0.2171,
      "num_input_tokens_seen": 10161296,
      "step": 15535
    },
    {
      "epoch": 8.144654088050315,
      "grad_norm": 1.5748045444488525,
      "learning_rate": 3.695147170748576e-05,
      "loss": 0.1841,
      "num_input_tokens_seen": 10164112,
      "step": 15540
    },
    {
      "epoch": 8.14727463312369,
      "grad_norm": 2.17549729347229,
      "learning_rate": 3.694142740748923e-05,
      "loss": 0.169,
      "num_input_tokens_seen": 10167248,
      "step": 15545
    },
    {
      "epoch": 8.149895178197065,
      "grad_norm": 3.6717453002929688,
      "learning_rate": 3.693138060948232e-05,
      "loss": 0.1696,
      "num_input_tokens_seen": 10171088,
      "step": 15550
    },
    {
      "epoch": 8.15251572327044,
      "grad_norm": 1.5961706638336182,
      "learning_rate": 3.692133131556671e-05,
      "loss": 0.159,
      "num_input_tokens_seen": 10174192,
      "step": 15555
    },
    {
      "epoch": 8.155136268343815,
      "grad_norm": 2.9665825366973877,
      "learning_rate": 3.6911279527844596e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 10177264,
      "step": 15560
    },
    {
      "epoch": 8.15775681341719,
      "grad_norm": 2.5155534744262695,
      "learning_rate": 3.69012252484187e-05,
      "loss": 0.1533,
      "num_input_tokens_seen": 10179824,
      "step": 15565
    },
    {
      "epoch": 8.160377358490566,
      "grad_norm": 1.6122492551803589,
      "learning_rate": 3.6891168479392245e-05,
      "loss": 0.1686,
      "num_input_tokens_seen": 10183024,
      "step": 15570
    },
    {
      "epoch": 8.16299790356394,
      "grad_norm": 2.088686943054199,
      "learning_rate": 3.688110922286901e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 10186832,
      "step": 15575
    },
    {
      "epoch": 8.165618448637316,
      "grad_norm": 1.8939270973205566,
      "learning_rate": 3.687104748095328e-05,
      "loss": 0.1977,
      "num_input_tokens_seen": 10189360,
      "step": 15580
    },
    {
      "epoch": 8.168238993710691,
      "grad_norm": 1.4137846231460571,
      "learning_rate": 3.686098325574984e-05,
      "loss": 0.1765,
      "num_input_tokens_seen": 10193360,
      "step": 15585
    },
    {
      "epoch": 8.170859538784066,
      "grad_norm": 2.115147829055786,
      "learning_rate": 3.685091654936403e-05,
      "loss": 0.1945,
      "num_input_tokens_seen": 10197136,
      "step": 15590
    },
    {
      "epoch": 8.173480083857442,
      "grad_norm": 3.250844955444336,
      "learning_rate": 3.6840847363901675e-05,
      "loss": 0.179,
      "num_input_tokens_seen": 10200592,
      "step": 15595
    },
    {
      "epoch": 8.176100628930818,
      "grad_norm": 1.7474480867385864,
      "learning_rate": 3.683077570146914e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 10203088,
      "step": 15600
    },
    {
      "epoch": 8.178721174004194,
      "grad_norm": 2.0354275703430176,
      "learning_rate": 3.68207015641733e-05,
      "loss": 0.2922,
      "num_input_tokens_seen": 10205840,
      "step": 15605
    },
    {
      "epoch": 8.181341719077569,
      "grad_norm": 1.646694302558899,
      "learning_rate": 3.681062495412155e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 10209040,
      "step": 15610
    },
    {
      "epoch": 8.183962264150944,
      "grad_norm": 1.3941177129745483,
      "learning_rate": 3.68005458734218e-05,
      "loss": 0.1229,
      "num_input_tokens_seen": 10212048,
      "step": 15615
    },
    {
      "epoch": 8.18658280922432,
      "grad_norm": 4.961025238037109,
      "learning_rate": 3.679046432418248e-05,
      "loss": 0.1705,
      "num_input_tokens_seen": 10214896,
      "step": 15620
    },
    {
      "epoch": 8.189203354297694,
      "grad_norm": 3.468262195587158,
      "learning_rate": 3.6780380308512543e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 10217872,
      "step": 15625
    },
    {
      "epoch": 8.19182389937107,
      "grad_norm": 1.0938156843185425,
      "learning_rate": 3.677029382852144e-05,
      "loss": 0.1329,
      "num_input_tokens_seen": 10221488,
      "step": 15630
    },
    {
      "epoch": 8.194444444444445,
      "grad_norm": 2.149024248123169,
      "learning_rate": 3.676020488631915e-05,
      "loss": 0.092,
      "num_input_tokens_seen": 10224048,
      "step": 15635
    },
    {
      "epoch": 8.19706498951782,
      "grad_norm": 1.2770521640777588,
      "learning_rate": 3.675011348401617e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 10228368,
      "step": 15640
    },
    {
      "epoch": 8.199685534591195,
      "grad_norm": 1.2935594320297241,
      "learning_rate": 3.6740019623723494e-05,
      "loss": 0.1717,
      "num_input_tokens_seen": 10231728,
      "step": 15645
    },
    {
      "epoch": 8.20230607966457,
      "grad_norm": 2.2923502922058105,
      "learning_rate": 3.672992330755265e-05,
      "loss": 0.2215,
      "num_input_tokens_seen": 10234960,
      "step": 15650
    },
    {
      "epoch": 8.204926624737945,
      "grad_norm": 1.0599409341812134,
      "learning_rate": 3.671982453761568e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 10238448,
      "step": 15655
    },
    {
      "epoch": 8.20754716981132,
      "grad_norm": 1.9929866790771484,
      "learning_rate": 3.6709723316025114e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 10240912,
      "step": 15660
    },
    {
      "epoch": 8.210167714884696,
      "grad_norm": 1.3681344985961914,
      "learning_rate": 3.669961964489403e-05,
      "loss": 0.1964,
      "num_input_tokens_seen": 10244816,
      "step": 15665
    },
    {
      "epoch": 8.21278825995807,
      "grad_norm": 1.935120701789856,
      "learning_rate": 3.668951352633599e-05,
      "loss": 0.1649,
      "num_input_tokens_seen": 10247728,
      "step": 15670
    },
    {
      "epoch": 8.215408805031446,
      "grad_norm": 1.4875024557113647,
      "learning_rate": 3.6679404962465084e-05,
      "loss": 0.1508,
      "num_input_tokens_seen": 10251248,
      "step": 15675
    },
    {
      "epoch": 8.218029350104821,
      "grad_norm": 1.4369152784347534,
      "learning_rate": 3.66692939553959e-05,
      "loss": 0.1905,
      "num_input_tokens_seen": 10254160,
      "step": 15680
    },
    {
      "epoch": 8.220649895178196,
      "grad_norm": 1.8935291767120361,
      "learning_rate": 3.6659180507243554e-05,
      "loss": 0.1817,
      "num_input_tokens_seen": 10256944,
      "step": 15685
    },
    {
      "epoch": 8.223270440251572,
      "grad_norm": 2.243497610092163,
      "learning_rate": 3.664906462012366e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 10259888,
      "step": 15690
    },
    {
      "epoch": 8.225890985324948,
      "grad_norm": 2.1822493076324463,
      "learning_rate": 3.663894629615234e-05,
      "loss": 0.2123,
      "num_input_tokens_seen": 10262640,
      "step": 15695
    },
    {
      "epoch": 8.228511530398324,
      "grad_norm": 2.10434889793396,
      "learning_rate": 3.6628825537446246e-05,
      "loss": 0.1969,
      "num_input_tokens_seen": 10265424,
      "step": 15700
    },
    {
      "epoch": 8.231132075471699,
      "grad_norm": 1.4310365915298462,
      "learning_rate": 3.6618702346122525e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 10267888,
      "step": 15705
    },
    {
      "epoch": 8.233752620545074,
      "grad_norm": 1.217586874961853,
      "learning_rate": 3.6608576724298817e-05,
      "loss": 0.1369,
      "num_input_tokens_seen": 10270992,
      "step": 15710
    },
    {
      "epoch": 8.23637316561845,
      "grad_norm": 1.5071793794631958,
      "learning_rate": 3.65984486740933e-05,
      "loss": 0.2042,
      "num_input_tokens_seen": 10274128,
      "step": 15715
    },
    {
      "epoch": 8.238993710691824,
      "grad_norm": 1.590875506401062,
      "learning_rate": 3.6588318197624644e-05,
      "loss": 0.1531,
      "num_input_tokens_seen": 10278128,
      "step": 15720
    },
    {
      "epoch": 8.2416142557652,
      "grad_norm": 1.4870730638504028,
      "learning_rate": 3.6578185297012015e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 10280720,
      "step": 15725
    },
    {
      "epoch": 8.244234800838575,
      "grad_norm": 2.1415367126464844,
      "learning_rate": 3.656804997437512e-05,
      "loss": 0.1752,
      "num_input_tokens_seen": 10283280,
      "step": 15730
    },
    {
      "epoch": 8.24685534591195,
      "grad_norm": 1.5663930177688599,
      "learning_rate": 3.6557912231834155e-05,
      "loss": 0.1742,
      "num_input_tokens_seen": 10287600,
      "step": 15735
    },
    {
      "epoch": 8.249475890985325,
      "grad_norm": 2.371161937713623,
      "learning_rate": 3.65477720715098e-05,
      "loss": 0.1486,
      "num_input_tokens_seen": 10289808,
      "step": 15740
    },
    {
      "epoch": 8.2520964360587,
      "grad_norm": 5.498873710632324,
      "learning_rate": 3.653762949552328e-05,
      "loss": 0.2438,
      "num_input_tokens_seen": 10292656,
      "step": 15745
    },
    {
      "epoch": 8.254716981132075,
      "grad_norm": 1.6706598997116089,
      "learning_rate": 3.6527484505996295e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 10295664,
      "step": 15750
    },
    {
      "epoch": 8.25733752620545,
      "grad_norm": 1.4212027788162231,
      "learning_rate": 3.651733710505106e-05,
      "loss": 0.1442,
      "num_input_tokens_seen": 10299248,
      "step": 15755
    },
    {
      "epoch": 8.259958071278826,
      "grad_norm": 3.061708688735962,
      "learning_rate": 3.6507187294810305e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 10302832,
      "step": 15760
    },
    {
      "epoch": 8.2625786163522,
      "grad_norm": 2.101473808288574,
      "learning_rate": 3.649703507739724e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 10307888,
      "step": 15765
    },
    {
      "epoch": 8.265199161425576,
      "grad_norm": 3.977810859680176,
      "learning_rate": 3.6486880454935606e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 10311248,
      "step": 15770
    },
    {
      "epoch": 8.267819706498951,
      "grad_norm": 1.6291579008102417,
      "learning_rate": 3.647672342954963e-05,
      "loss": 0.1917,
      "num_input_tokens_seen": 10314320,
      "step": 15775
    },
    {
      "epoch": 8.270440251572326,
      "grad_norm": 1.8449169397354126,
      "learning_rate": 3.646656400336405e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 10317936,
      "step": 15780
    },
    {
      "epoch": 8.273060796645701,
      "grad_norm": 1.8379510641098022,
      "learning_rate": 3.6456402178504094e-05,
      "loss": 0.2004,
      "num_input_tokens_seen": 10321136,
      "step": 15785
    },
    {
      "epoch": 8.275681341719078,
      "grad_norm": 2.1854281425476074,
      "learning_rate": 3.6446237957095505e-05,
      "loss": 0.26,
      "num_input_tokens_seen": 10323760,
      "step": 15790
    },
    {
      "epoch": 8.278301886792454,
      "grad_norm": 2.288268804550171,
      "learning_rate": 3.643607134126452e-05,
      "loss": 0.1906,
      "num_input_tokens_seen": 10327088,
      "step": 15795
    },
    {
      "epoch": 8.280922431865829,
      "grad_norm": 1.803213357925415,
      "learning_rate": 3.642590233313789e-05,
      "loss": 0.2222,
      "num_input_tokens_seen": 10330448,
      "step": 15800
    },
    {
      "epoch": 8.283542976939204,
      "grad_norm": 2.2279560565948486,
      "learning_rate": 3.6415730934842827e-05,
      "loss": 0.1539,
      "num_input_tokens_seen": 10334672,
      "step": 15805
    },
    {
      "epoch": 8.286163522012579,
      "grad_norm": 2.4041247367858887,
      "learning_rate": 3.64055571485071e-05,
      "loss": 0.1654,
      "num_input_tokens_seen": 10337904,
      "step": 15810
    },
    {
      "epoch": 8.288784067085954,
      "grad_norm": 2.760643243789673,
      "learning_rate": 3.639538097625894e-05,
      "loss": 0.1842,
      "num_input_tokens_seen": 10341488,
      "step": 15815
    },
    {
      "epoch": 8.29140461215933,
      "grad_norm": 4.577783584594727,
      "learning_rate": 3.638520242022708e-05,
      "loss": 0.1596,
      "num_input_tokens_seen": 10344944,
      "step": 15820
    },
    {
      "epoch": 8.294025157232705,
      "grad_norm": 4.5982561111450195,
      "learning_rate": 3.637502148254077e-05,
      "loss": 0.2318,
      "num_input_tokens_seen": 10347248,
      "step": 15825
    },
    {
      "epoch": 8.29664570230608,
      "grad_norm": 2.878159523010254,
      "learning_rate": 3.636483816532974e-05,
      "loss": 0.2193,
      "num_input_tokens_seen": 10349712,
      "step": 15830
    },
    {
      "epoch": 8.299266247379455,
      "grad_norm": 1.1249170303344727,
      "learning_rate": 3.6354652470724217e-05,
      "loss": 0.1605,
      "num_input_tokens_seen": 10353936,
      "step": 15835
    },
    {
      "epoch": 8.30188679245283,
      "grad_norm": 16.506065368652344,
      "learning_rate": 3.634446440085494e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 10356432,
      "step": 15840
    },
    {
      "epoch": 8.304507337526205,
      "grad_norm": 2.6340296268463135,
      "learning_rate": 3.633427395785313e-05,
      "loss": 0.3262,
      "num_input_tokens_seen": 10360624,
      "step": 15845
    },
    {
      "epoch": 8.30712788259958,
      "grad_norm": 2.0983901023864746,
      "learning_rate": 3.6324081143850516e-05,
      "loss": 0.1757,
      "num_input_tokens_seen": 10364432,
      "step": 15850
    },
    {
      "epoch": 8.309748427672956,
      "grad_norm": 1.679207444190979,
      "learning_rate": 3.6313885960979326e-05,
      "loss": 0.1635,
      "num_input_tokens_seen": 10367184,
      "step": 15855
    },
    {
      "epoch": 8.31236897274633,
      "grad_norm": 2.1959197521209717,
      "learning_rate": 3.630368841137226e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 10370192,
      "step": 15860
    },
    {
      "epoch": 8.314989517819706,
      "grad_norm": 1.6489531993865967,
      "learning_rate": 3.6293488497162554e-05,
      "loss": 0.2809,
      "num_input_tokens_seen": 10373744,
      "step": 15865
    },
    {
      "epoch": 8.317610062893081,
      "grad_norm": 2.8991446495056152,
      "learning_rate": 3.628328622048388e-05,
      "loss": 0.2017,
      "num_input_tokens_seen": 10376400,
      "step": 15870
    },
    {
      "epoch": 8.320230607966456,
      "grad_norm": 1.3878971338272095,
      "learning_rate": 3.627308158347046e-05,
      "loss": 0.1409,
      "num_input_tokens_seen": 10379600,
      "step": 15875
    },
    {
      "epoch": 8.322851153039831,
      "grad_norm": 3.8514404296875,
      "learning_rate": 3.626287458825699e-05,
      "loss": 0.1935,
      "num_input_tokens_seen": 10383088,
      "step": 15880
    },
    {
      "epoch": 8.325471698113208,
      "grad_norm": 3.0509257316589355,
      "learning_rate": 3.625266523697864e-05,
      "loss": 0.2197,
      "num_input_tokens_seen": 10386288,
      "step": 15885
    },
    {
      "epoch": 8.328092243186584,
      "grad_norm": 2.883695363998413,
      "learning_rate": 3.62424535317711e-05,
      "loss": 0.2709,
      "num_input_tokens_seen": 10389744,
      "step": 15890
    },
    {
      "epoch": 8.330712788259959,
      "grad_norm": 3.3569769859313965,
      "learning_rate": 3.623223947477055e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 10392880,
      "step": 15895
    },
    {
      "epoch": 8.333333333333334,
      "grad_norm": 2.0037314891815186,
      "learning_rate": 3.6222023068113635e-05,
      "loss": 0.217,
      "num_input_tokens_seen": 10397168,
      "step": 15900
    },
    {
      "epoch": 8.335953878406709,
      "grad_norm": 1.861029863357544,
      "learning_rate": 3.621180431393754e-05,
      "loss": 0.1762,
      "num_input_tokens_seen": 10400784,
      "step": 15905
    },
    {
      "epoch": 8.338574423480084,
      "grad_norm": 1.3890000581741333,
      "learning_rate": 3.6201583214379874e-05,
      "loss": 0.1019,
      "num_input_tokens_seen": 10403568,
      "step": 15910
    },
    {
      "epoch": 8.34119496855346,
      "grad_norm": 2.9342026710510254,
      "learning_rate": 3.61913597715788e-05,
      "loss": 0.1452,
      "num_input_tokens_seen": 10406800,
      "step": 15915
    },
    {
      "epoch": 8.343815513626835,
      "grad_norm": 1.919296383857727,
      "learning_rate": 3.6181133987672944e-05,
      "loss": 0.1863,
      "num_input_tokens_seen": 10409744,
      "step": 15920
    },
    {
      "epoch": 8.34643605870021,
      "grad_norm": 2.3694539070129395,
      "learning_rate": 3.617090586480142e-05,
      "loss": 0.143,
      "num_input_tokens_seen": 10412880,
      "step": 15925
    },
    {
      "epoch": 8.349056603773585,
      "grad_norm": 4.457956790924072,
      "learning_rate": 3.6160675405103826e-05,
      "loss": 0.2206,
      "num_input_tokens_seen": 10416432,
      "step": 15930
    },
    {
      "epoch": 8.35167714884696,
      "grad_norm": 3.2294058799743652,
      "learning_rate": 3.615044261072027e-05,
      "loss": 0.1835,
      "num_input_tokens_seen": 10419824,
      "step": 15935
    },
    {
      "epoch": 8.354297693920335,
      "grad_norm": 1.9406319856643677,
      "learning_rate": 3.614020748379133e-05,
      "loss": 0.1766,
      "num_input_tokens_seen": 10422800,
      "step": 15940
    },
    {
      "epoch": 8.35691823899371,
      "grad_norm": 1.6730015277862549,
      "learning_rate": 3.6129970026458076e-05,
      "loss": 0.1481,
      "num_input_tokens_seen": 10425808,
      "step": 15945
    },
    {
      "epoch": 8.359538784067086,
      "grad_norm": 1.5724433660507202,
      "learning_rate": 3.611973024086207e-05,
      "loss": 0.1668,
      "num_input_tokens_seen": 10428944,
      "step": 15950
    },
    {
      "epoch": 8.36215932914046,
      "grad_norm": 2.571636199951172,
      "learning_rate": 3.610948812914536e-05,
      "loss": 0.2238,
      "num_input_tokens_seen": 10433648,
      "step": 15955
    },
    {
      "epoch": 8.364779874213836,
      "grad_norm": 2.0322887897491455,
      "learning_rate": 3.609924369345048e-05,
      "loss": 0.2008,
      "num_input_tokens_seen": 10436784,
      "step": 15960
    },
    {
      "epoch": 8.367400419287211,
      "grad_norm": 2.3323941230773926,
      "learning_rate": 3.6088996935920444e-05,
      "loss": 0.1557,
      "num_input_tokens_seen": 10439280,
      "step": 15965
    },
    {
      "epoch": 8.370020964360586,
      "grad_norm": 1.321123480796814,
      "learning_rate": 3.607874785869876e-05,
      "loss": 0.1544,
      "num_input_tokens_seen": 10441744,
      "step": 15970
    },
    {
      "epoch": 8.372641509433961,
      "grad_norm": 2.205382823944092,
      "learning_rate": 3.6068496463929416e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 10445424,
      "step": 15975
    },
    {
      "epoch": 8.375262054507338,
      "grad_norm": 3.2398011684417725,
      "learning_rate": 3.605824275375689e-05,
      "loss": 0.2553,
      "num_input_tokens_seen": 10448688,
      "step": 15980
    },
    {
      "epoch": 8.377882599580714,
      "grad_norm": 1.541135311126709,
      "learning_rate": 3.604798673032613e-05,
      "loss": 0.2068,
      "num_input_tokens_seen": 10453040,
      "step": 15985
    },
    {
      "epoch": 8.380503144654089,
      "grad_norm": 1.966154932975769,
      "learning_rate": 3.60377283957826e-05,
      "loss": 0.2188,
      "num_input_tokens_seen": 10457072,
      "step": 15990
    },
    {
      "epoch": 8.383123689727464,
      "grad_norm": 4.714531421661377,
      "learning_rate": 3.6027467752272203e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 10459280,
      "step": 15995
    },
    {
      "epoch": 8.385744234800839,
      "grad_norm": 4.011285781860352,
      "learning_rate": 3.601720480194137e-05,
      "loss": 0.1611,
      "num_input_tokens_seen": 10462192,
      "step": 16000
    },
    {
      "epoch": 8.388364779874214,
      "grad_norm": 1.252846121788025,
      "learning_rate": 3.600693954693698e-05,
      "loss": 0.2101,
      "num_input_tokens_seen": 10466672,
      "step": 16005
    },
    {
      "epoch": 8.39098532494759,
      "grad_norm": 1.8313052654266357,
      "learning_rate": 3.5996671989406414e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 10469776,
      "step": 16010
    },
    {
      "epoch": 8.393605870020965,
      "grad_norm": 3.386380910873413,
      "learning_rate": 3.598640213149752e-05,
      "loss": 0.139,
      "num_input_tokens_seen": 10472848,
      "step": 16015
    },
    {
      "epoch": 8.39622641509434,
      "grad_norm": 6.17066764831543,
      "learning_rate": 3.597612997535864e-05,
      "loss": 0.1795,
      "num_input_tokens_seen": 10476368,
      "step": 16020
    },
    {
      "epoch": 8.398846960167715,
      "grad_norm": 1.4201326370239258,
      "learning_rate": 3.596585552313859e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 10479472,
      "step": 16025
    },
    {
      "epoch": 8.40146750524109,
      "grad_norm": 5.880892276763916,
      "learning_rate": 3.595557877698666e-05,
      "loss": 0.1973,
      "num_input_tokens_seen": 10482000,
      "step": 16030
    },
    {
      "epoch": 8.404088050314465,
      "grad_norm": 12.605088233947754,
      "learning_rate": 3.594529973905266e-05,
      "loss": 0.2296,
      "num_input_tokens_seen": 10484592,
      "step": 16035
    },
    {
      "epoch": 8.40670859538784,
      "grad_norm": 5.127799987792969,
      "learning_rate": 3.593501841148681e-05,
      "loss": 0.1559,
      "num_input_tokens_seen": 10487216,
      "step": 16040
    },
    {
      "epoch": 8.409329140461216,
      "grad_norm": 1.4753143787384033,
      "learning_rate": 3.5924734796439864e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 10490512,
      "step": 16045
    },
    {
      "epoch": 8.41194968553459,
      "grad_norm": 3.0026094913482666,
      "learning_rate": 3.5914448896063044e-05,
      "loss": 0.142,
      "num_input_tokens_seen": 10493552,
      "step": 16050
    },
    {
      "epoch": 8.414570230607966,
      "grad_norm": 4.196080207824707,
      "learning_rate": 3.590416071250802e-05,
      "loss": 0.21,
      "num_input_tokens_seen": 10497264,
      "step": 16055
    },
    {
      "epoch": 8.417190775681341,
      "grad_norm": 2.146150588989258,
      "learning_rate": 3.589387024792699e-05,
      "loss": 0.1321,
      "num_input_tokens_seen": 10499984,
      "step": 16060
    },
    {
      "epoch": 8.419811320754716,
      "grad_norm": 2.40903639793396,
      "learning_rate": 3.5883577504472576e-05,
      "loss": 0.1633,
      "num_input_tokens_seen": 10503248,
      "step": 16065
    },
    {
      "epoch": 8.422431865828091,
      "grad_norm": 1.304602026939392,
      "learning_rate": 3.587328248429792e-05,
      "loss": 0.1773,
      "num_input_tokens_seen": 10507088,
      "step": 16070
    },
    {
      "epoch": 8.425052410901468,
      "grad_norm": 2.216035842895508,
      "learning_rate": 3.586298518955661e-05,
      "loss": 0.1679,
      "num_input_tokens_seen": 10510416,
      "step": 16075
    },
    {
      "epoch": 8.427672955974844,
      "grad_norm": 2.4233524799346924,
      "learning_rate": 3.5852685622402746e-05,
      "loss": 0.248,
      "num_input_tokens_seen": 10514384,
      "step": 16080
    },
    {
      "epoch": 8.430293501048219,
      "grad_norm": 1.286604642868042,
      "learning_rate": 3.584238378499085e-05,
      "loss": 0.2242,
      "num_input_tokens_seen": 10517872,
      "step": 16085
    },
    {
      "epoch": 8.432914046121594,
      "grad_norm": 2.9723894596099854,
      "learning_rate": 3.5832079679475974e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 10521200,
      "step": 16090
    },
    {
      "epoch": 8.435534591194969,
      "grad_norm": 1.9167107343673706,
      "learning_rate": 3.582177330801359e-05,
      "loss": 0.1634,
      "num_input_tokens_seen": 10523856,
      "step": 16095
    },
    {
      "epoch": 8.438155136268344,
      "grad_norm": 1.698254108428955,
      "learning_rate": 3.58114646727597e-05,
      "loss": 0.2558,
      "num_input_tokens_seen": 10527728,
      "step": 16100
    },
    {
      "epoch": 8.44077568134172,
      "grad_norm": 2.103081226348877,
      "learning_rate": 3.580115377587075e-05,
      "loss": 0.1868,
      "num_input_tokens_seen": 10531536,
      "step": 16105
    },
    {
      "epoch": 8.443396226415095,
      "grad_norm": 1.150317907333374,
      "learning_rate": 3.5790840619503654e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 10534352,
      "step": 16110
    },
    {
      "epoch": 8.44601677148847,
      "grad_norm": 2.6394031047821045,
      "learning_rate": 3.57805252058158e-05,
      "loss": 0.1976,
      "num_input_tokens_seen": 10537552,
      "step": 16115
    },
    {
      "epoch": 8.448637316561845,
      "grad_norm": 1.4212018251419067,
      "learning_rate": 3.577020753696507e-05,
      "loss": 0.1047,
      "num_input_tokens_seen": 10539952,
      "step": 16120
    },
    {
      "epoch": 8.45125786163522,
      "grad_norm": 1.9263732433319092,
      "learning_rate": 3.57598876151098e-05,
      "loss": 0.1771,
      "num_input_tokens_seen": 10543408,
      "step": 16125
    },
    {
      "epoch": 8.453878406708595,
      "grad_norm": 6.641373634338379,
      "learning_rate": 3.5749565442408796e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 10546352,
      "step": 16130
    },
    {
      "epoch": 8.45649895178197,
      "grad_norm": 1.5989567041397095,
      "learning_rate": 3.573924102102133e-05,
      "loss": 0.159,
      "num_input_tokens_seen": 10550352,
      "step": 16135
    },
    {
      "epoch": 8.459119496855346,
      "grad_norm": 2.0205113887786865,
      "learning_rate": 3.572891435310717e-05,
      "loss": 0.3028,
      "num_input_tokens_seen": 10554864,
      "step": 16140
    },
    {
      "epoch": 8.46174004192872,
      "grad_norm": 2.146071672439575,
      "learning_rate": 3.5718585440826527e-05,
      "loss": 0.1084,
      "num_input_tokens_seen": 10558960,
      "step": 16145
    },
    {
      "epoch": 8.464360587002096,
      "grad_norm": 2.1982216835021973,
      "learning_rate": 3.570825428634009e-05,
      "loss": 0.1815,
      "num_input_tokens_seen": 10563600,
      "step": 16150
    },
    {
      "epoch": 8.466981132075471,
      "grad_norm": 2.01692271232605,
      "learning_rate": 3.569792089180903e-05,
      "loss": 0.2009,
      "num_input_tokens_seen": 10566480,
      "step": 16155
    },
    {
      "epoch": 8.469601677148846,
      "grad_norm": 2.1346349716186523,
      "learning_rate": 3.568758525939497e-05,
      "loss": 0.1258,
      "num_input_tokens_seen": 10569616,
      "step": 16160
    },
    {
      "epoch": 8.472222222222221,
      "grad_norm": 3.8409082889556885,
      "learning_rate": 3.5677247391259995e-05,
      "loss": 0.2192,
      "num_input_tokens_seen": 10572336,
      "step": 16165
    },
    {
      "epoch": 8.474842767295598,
      "grad_norm": 3.3294737339019775,
      "learning_rate": 3.566690728956668e-05,
      "loss": 0.2217,
      "num_input_tokens_seen": 10574576,
      "step": 16170
    },
    {
      "epoch": 8.477463312368974,
      "grad_norm": 1.4701482057571411,
      "learning_rate": 3.5656564956478056e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 10578640,
      "step": 16175
    },
    {
      "epoch": 8.480083857442349,
      "grad_norm": 1.0046464204788208,
      "learning_rate": 3.5646220394157626e-05,
      "loss": 0.1434,
      "num_input_tokens_seen": 10581392,
      "step": 16180
    },
    {
      "epoch": 8.482704402515724,
      "grad_norm": 2.4220969676971436,
      "learning_rate": 3.563587360476933e-05,
      "loss": 0.2122,
      "num_input_tokens_seen": 10584144,
      "step": 16185
    },
    {
      "epoch": 8.485324947589099,
      "grad_norm": 6.724411964416504,
      "learning_rate": 3.562552459047763e-05,
      "loss": 0.2139,
      "num_input_tokens_seen": 10587600,
      "step": 16190
    },
    {
      "epoch": 8.487945492662474,
      "grad_norm": 3.6760311126708984,
      "learning_rate": 3.561517335344741e-05,
      "loss": 0.203,
      "num_input_tokens_seen": 10590832,
      "step": 16195
    },
    {
      "epoch": 8.49056603773585,
      "grad_norm": 2.994941473007202,
      "learning_rate": 3.560481989584402e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 10593712,
      "step": 16200
    },
    {
      "epoch": 8.493186582809225,
      "grad_norm": 1.7671982049942017,
      "learning_rate": 3.55944642198333e-05,
      "loss": 0.2081,
      "num_input_tokens_seen": 10596336,
      "step": 16205
    },
    {
      "epoch": 8.4958071278826,
      "grad_norm": 8.336376190185547,
      "learning_rate": 3.5584106327581536e-05,
      "loss": 0.2494,
      "num_input_tokens_seen": 10599536,
      "step": 16210
    },
    {
      "epoch": 8.498427672955975,
      "grad_norm": 2.0596446990966797,
      "learning_rate": 3.557374622125547e-05,
      "loss": 0.099,
      "num_input_tokens_seen": 10602576,
      "step": 16215
    },
    {
      "epoch": 8.50104821802935,
      "grad_norm": 2.011155843734741,
      "learning_rate": 3.556338390302234e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 10605616,
      "step": 16220
    },
    {
      "epoch": 8.503668763102725,
      "grad_norm": 2.566270351409912,
      "learning_rate": 3.55530193750498e-05,
      "loss": 0.2927,
      "num_input_tokens_seen": 10609040,
      "step": 16225
    },
    {
      "epoch": 8.5062893081761,
      "grad_norm": 2.0563690662384033,
      "learning_rate": 3.5542652639505996e-05,
      "loss": 0.2168,
      "num_input_tokens_seen": 10612304,
      "step": 16230
    },
    {
      "epoch": 8.508909853249476,
      "grad_norm": 1.2057305574417114,
      "learning_rate": 3.5532283698559556e-05,
      "loss": 0.1623,
      "num_input_tokens_seen": 10614928,
      "step": 16235
    },
    {
      "epoch": 8.51153039832285,
      "grad_norm": 2.6558005809783936,
      "learning_rate": 3.55219125543795e-05,
      "loss": 0.2031,
      "num_input_tokens_seen": 10618320,
      "step": 16240
    },
    {
      "epoch": 8.514150943396226,
      "grad_norm": 4.938521862030029,
      "learning_rate": 3.5511539209135396e-05,
      "loss": 0.2168,
      "num_input_tokens_seen": 10621296,
      "step": 16245
    },
    {
      "epoch": 8.516771488469601,
      "grad_norm": 1.7882487773895264,
      "learning_rate": 3.550116366499721e-05,
      "loss": 0.2212,
      "num_input_tokens_seen": 10623984,
      "step": 16250
    },
    {
      "epoch": 8.519392033542976,
      "grad_norm": 2.5758941173553467,
      "learning_rate": 3.549078592413538e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 10628560,
      "step": 16255
    },
    {
      "epoch": 8.522012578616351,
      "grad_norm": 2.6535634994506836,
      "learning_rate": 3.548040598872083e-05,
      "loss": 0.173,
      "num_input_tokens_seen": 10631856,
      "step": 16260
    },
    {
      "epoch": 8.524633123689728,
      "grad_norm": 1.6983355283737183,
      "learning_rate": 3.5470023860924905e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 10635696,
      "step": 16265
    },
    {
      "epoch": 8.527253668763104,
      "grad_norm": 2.1847310066223145,
      "learning_rate": 3.5459639542919444e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 10638384,
      "step": 16270
    },
    {
      "epoch": 8.529874213836479,
      "grad_norm": 2.928858518600464,
      "learning_rate": 3.5449253036876715e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 10641456,
      "step": 16275
    },
    {
      "epoch": 8.532494758909854,
      "grad_norm": 5.946885108947754,
      "learning_rate": 3.543886434496947e-05,
      "loss": 0.158,
      "num_input_tokens_seen": 10644336,
      "step": 16280
    },
    {
      "epoch": 8.535115303983229,
      "grad_norm": 1.6462256908416748,
      "learning_rate": 3.542847346937089e-05,
      "loss": 0.1861,
      "num_input_tokens_seen": 10647440,
      "step": 16285
    },
    {
      "epoch": 8.537735849056604,
      "grad_norm": 1.678225040435791,
      "learning_rate": 3.541808041225464e-05,
      "loss": 0.1671,
      "num_input_tokens_seen": 10649872,
      "step": 16290
    },
    {
      "epoch": 8.54035639412998,
      "grad_norm": 1.5174369812011719,
      "learning_rate": 3.540768517579481e-05,
      "loss": 0.1657,
      "num_input_tokens_seen": 10652976,
      "step": 16295
    },
    {
      "epoch": 8.542976939203355,
      "grad_norm": 1.6376596689224243,
      "learning_rate": 3.539728776216599e-05,
      "loss": 0.1113,
      "num_input_tokens_seen": 10656016,
      "step": 16300
    },
    {
      "epoch": 8.54559748427673,
      "grad_norm": 3.900141716003418,
      "learning_rate": 3.5386888173543185e-05,
      "loss": 0.2847,
      "num_input_tokens_seen": 10659536,
      "step": 16305
    },
    {
      "epoch": 8.548218029350105,
      "grad_norm": 2.771425485610962,
      "learning_rate": 3.5376486412101875e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 10662416,
      "step": 16310
    },
    {
      "epoch": 8.55083857442348,
      "grad_norm": 1.725378155708313,
      "learning_rate": 3.5366082480017994e-05,
      "loss": 0.1366,
      "num_input_tokens_seen": 10666320,
      "step": 16315
    },
    {
      "epoch": 8.553459119496855,
      "grad_norm": 2.021174430847168,
      "learning_rate": 3.535567637946792e-05,
      "loss": 0.2384,
      "num_input_tokens_seen": 10669136,
      "step": 16320
    },
    {
      "epoch": 8.55607966457023,
      "grad_norm": 1.843390703201294,
      "learning_rate": 3.534526811262848e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 10672240,
      "step": 16325
    },
    {
      "epoch": 8.558700209643606,
      "grad_norm": 8.352346420288086,
      "learning_rate": 3.533485768167698e-05,
      "loss": 0.2016,
      "num_input_tokens_seen": 10674992,
      "step": 16330
    },
    {
      "epoch": 8.56132075471698,
      "grad_norm": 2.015536069869995,
      "learning_rate": 3.532444508879117e-05,
      "loss": 0.1856,
      "num_input_tokens_seen": 10677616,
      "step": 16335
    },
    {
      "epoch": 8.563941299790356,
      "grad_norm": 3.5135128498077393,
      "learning_rate": 3.531403033614922e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 10681136,
      "step": 16340
    },
    {
      "epoch": 8.566561844863731,
      "grad_norm": 2.5123138427734375,
      "learning_rate": 3.530361342592981e-05,
      "loss": 0.2619,
      "num_input_tokens_seen": 10684368,
      "step": 16345
    },
    {
      "epoch": 8.569182389937106,
      "grad_norm": 1.6523761749267578,
      "learning_rate": 3.5293194360312006e-05,
      "loss": 0.2126,
      "num_input_tokens_seen": 10687632,
      "step": 16350
    },
    {
      "epoch": 8.571802935010481,
      "grad_norm": 1.0535985231399536,
      "learning_rate": 3.528277314147538e-05,
      "loss": 0.1925,
      "num_input_tokens_seen": 10692336,
      "step": 16355
    },
    {
      "epoch": 8.574423480083858,
      "grad_norm": 1.2760062217712402,
      "learning_rate": 3.5272349771599925e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 10696048,
      "step": 16360
    },
    {
      "epoch": 8.577044025157234,
      "grad_norm": 2.7044219970703125,
      "learning_rate": 3.526192425286609e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 10700304,
      "step": 16365
    },
    {
      "epoch": 8.579664570230609,
      "grad_norm": 5.564304351806641,
      "learning_rate": 3.525149658745476e-05,
      "loss": 0.1561,
      "num_input_tokens_seen": 10703120,
      "step": 16370
    },
    {
      "epoch": 8.582285115303984,
      "grad_norm": 1.8461999893188477,
      "learning_rate": 3.5241066777547315e-05,
      "loss": 0.1839,
      "num_input_tokens_seen": 10707536,
      "step": 16375
    },
    {
      "epoch": 8.584905660377359,
      "grad_norm": 1.7148022651672363,
      "learning_rate": 3.523063482532554e-05,
      "loss": 0.2248,
      "num_input_tokens_seen": 10710384,
      "step": 16380
    },
    {
      "epoch": 8.587526205450734,
      "grad_norm": 2.539025068283081,
      "learning_rate": 3.5220200732971665e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 10713840,
      "step": 16385
    },
    {
      "epoch": 8.59014675052411,
      "grad_norm": 2.2466862201690674,
      "learning_rate": 3.52097645026684e-05,
      "loss": 0.1668,
      "num_input_tokens_seen": 10716720,
      "step": 16390
    },
    {
      "epoch": 8.592767295597485,
      "grad_norm": 3.5349538326263428,
      "learning_rate": 3.519932613659889e-05,
      "loss": 0.2087,
      "num_input_tokens_seen": 10720720,
      "step": 16395
    },
    {
      "epoch": 8.59538784067086,
      "grad_norm": 1.9305959939956665,
      "learning_rate": 3.5188885636946696e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 10723952,
      "step": 16400
    },
    {
      "epoch": 8.598008385744235,
      "grad_norm": 2.066305160522461,
      "learning_rate": 3.5178443005895884e-05,
      "loss": 0.2379,
      "num_input_tokens_seen": 10726096,
      "step": 16405
    },
    {
      "epoch": 8.60062893081761,
      "grad_norm": 1.1203190088272095,
      "learning_rate": 3.51679982456309e-05,
      "loss": 0.1687,
      "num_input_tokens_seen": 10729456,
      "step": 16410
    },
    {
      "epoch": 8.603249475890985,
      "grad_norm": 1.7503165006637573,
      "learning_rate": 3.51575513583367e-05,
      "loss": 0.165,
      "num_input_tokens_seen": 10731824,
      "step": 16415
    },
    {
      "epoch": 8.60587002096436,
      "grad_norm": 1.0957598686218262,
      "learning_rate": 3.5147102346198635e-05,
      "loss": 0.1628,
      "num_input_tokens_seen": 10734864,
      "step": 16420
    },
    {
      "epoch": 8.608490566037736,
      "grad_norm": 1.806585431098938,
      "learning_rate": 3.5136651211402515e-05,
      "loss": 0.1585,
      "num_input_tokens_seen": 10737744,
      "step": 16425
    },
    {
      "epoch": 8.61111111111111,
      "grad_norm": 2.2320139408111572,
      "learning_rate": 3.5126197956134614e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 10740656,
      "step": 16430
    },
    {
      "epoch": 8.613731656184486,
      "grad_norm": 3.5707149505615234,
      "learning_rate": 3.5115742582581626e-05,
      "loss": 0.3036,
      "num_input_tokens_seen": 10744688,
      "step": 16435
    },
    {
      "epoch": 8.616352201257861,
      "grad_norm": 2.09541654586792,
      "learning_rate": 3.51052850929307e-05,
      "loss": 0.2278,
      "num_input_tokens_seen": 10747184,
      "step": 16440
    },
    {
      "epoch": 8.618972746331236,
      "grad_norm": 2.4037973880767822,
      "learning_rate": 3.509482548936941e-05,
      "loss": 0.2015,
      "num_input_tokens_seen": 10750320,
      "step": 16445
    },
    {
      "epoch": 8.621593291404611,
      "grad_norm": 2.1111879348754883,
      "learning_rate": 3.5084363774085804e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 10753328,
      "step": 16450
    },
    {
      "epoch": 8.624213836477988,
      "grad_norm": 2.484666109085083,
      "learning_rate": 3.507389994926834e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 10756208,
      "step": 16455
    },
    {
      "epoch": 8.626834381551364,
      "grad_norm": 2.28066086769104,
      "learning_rate": 3.5063434017105946e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 10759888,
      "step": 16460
    },
    {
      "epoch": 8.629454926624739,
      "grad_norm": 1.718326210975647,
      "learning_rate": 3.505296597978795e-05,
      "loss": 0.1364,
      "num_input_tokens_seen": 10763152,
      "step": 16465
    },
    {
      "epoch": 8.632075471698114,
      "grad_norm": 2.901092529296875,
      "learning_rate": 3.504249583950417e-05,
      "loss": 0.1895,
      "num_input_tokens_seen": 10767024,
      "step": 16470
    },
    {
      "epoch": 8.634696016771489,
      "grad_norm": 2.320711612701416,
      "learning_rate": 3.503202359844483e-05,
      "loss": 0.1488,
      "num_input_tokens_seen": 10770096,
      "step": 16475
    },
    {
      "epoch": 8.637316561844864,
      "grad_norm": 3.2429099082946777,
      "learning_rate": 3.50215492588006e-05,
      "loss": 0.1879,
      "num_input_tokens_seen": 10774576,
      "step": 16480
    },
    {
      "epoch": 8.63993710691824,
      "grad_norm": 4.67301082611084,
      "learning_rate": 3.5011072822762605e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 10777296,
      "step": 16485
    },
    {
      "epoch": 8.642557651991615,
      "grad_norm": 2.1317291259765625,
      "learning_rate": 3.500059429252238e-05,
      "loss": 0.177,
      "num_input_tokens_seen": 10779696,
      "step": 16490
    },
    {
      "epoch": 8.64517819706499,
      "grad_norm": 4.365866184234619,
      "learning_rate": 3.499011367027193e-05,
      "loss": 0.1877,
      "num_input_tokens_seen": 10783184,
      "step": 16495
    },
    {
      "epoch": 8.647798742138365,
      "grad_norm": 2.741454839706421,
      "learning_rate": 3.4979630958203665e-05,
      "loss": 0.3019,
      "num_input_tokens_seen": 10786512,
      "step": 16500
    },
    {
      "epoch": 8.65041928721174,
      "grad_norm": 1.8923002481460571,
      "learning_rate": 3.496914615851046e-05,
      "loss": 0.16,
      "num_input_tokens_seen": 10789872,
      "step": 16505
    },
    {
      "epoch": 8.653039832285115,
      "grad_norm": 1.815092921257019,
      "learning_rate": 3.495865927338562e-05,
      "loss": 0.1669,
      "num_input_tokens_seen": 10792944,
      "step": 16510
    },
    {
      "epoch": 8.65566037735849,
      "grad_norm": 2.68472957611084,
      "learning_rate": 3.494817030502286e-05,
      "loss": 0.1747,
      "num_input_tokens_seen": 10796624,
      "step": 16515
    },
    {
      "epoch": 8.658280922431866,
      "grad_norm": 2.161168336868286,
      "learning_rate": 3.493767925561638e-05,
      "loss": 0.1836,
      "num_input_tokens_seen": 10800144,
      "step": 16520
    },
    {
      "epoch": 8.66090146750524,
      "grad_norm": 2.371873140335083,
      "learning_rate": 3.492718612736077e-05,
      "loss": 0.139,
      "num_input_tokens_seen": 10803152,
      "step": 16525
    },
    {
      "epoch": 8.663522012578616,
      "grad_norm": 3.3897616863250732,
      "learning_rate": 3.491669092245108e-05,
      "loss": 0.2643,
      "num_input_tokens_seen": 10805552,
      "step": 16530
    },
    {
      "epoch": 8.666142557651991,
      "grad_norm": 3.0160906314849854,
      "learning_rate": 3.4906193643082784e-05,
      "loss": 0.1798,
      "num_input_tokens_seen": 10808880,
      "step": 16535
    },
    {
      "epoch": 8.668763102725366,
      "grad_norm": 3.113668441772461,
      "learning_rate": 3.4895694291451795e-05,
      "loss": 0.1739,
      "num_input_tokens_seen": 10812240,
      "step": 16540
    },
    {
      "epoch": 8.671383647798741,
      "grad_norm": 3.830559253692627,
      "learning_rate": 3.4885192869754444e-05,
      "loss": 0.1699,
      "num_input_tokens_seen": 10815600,
      "step": 16545
    },
    {
      "epoch": 8.674004192872118,
      "grad_norm": 2.2088847160339355,
      "learning_rate": 3.4874689380187537e-05,
      "loss": 0.159,
      "num_input_tokens_seen": 10819120,
      "step": 16550
    },
    {
      "epoch": 8.676624737945493,
      "grad_norm": 1.4125357866287231,
      "learning_rate": 3.486418382494826e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 10821744,
      "step": 16555
    },
    {
      "epoch": 8.679245283018869,
      "grad_norm": 2.4742324352264404,
      "learning_rate": 3.485367620623426e-05,
      "loss": 0.1827,
      "num_input_tokens_seen": 10824336,
      "step": 16560
    },
    {
      "epoch": 8.681865828092244,
      "grad_norm": 4.64440393447876,
      "learning_rate": 3.484316652624362e-05,
      "loss": 0.2374,
      "num_input_tokens_seen": 10826672,
      "step": 16565
    },
    {
      "epoch": 8.684486373165619,
      "grad_norm": 1.6354835033416748,
      "learning_rate": 3.483265478717484e-05,
      "loss": 0.149,
      "num_input_tokens_seen": 10830192,
      "step": 16570
    },
    {
      "epoch": 8.687106918238994,
      "grad_norm": 3.100144147872925,
      "learning_rate": 3.482214099122686e-05,
      "loss": 0.2754,
      "num_input_tokens_seen": 10834096,
      "step": 16575
    },
    {
      "epoch": 8.68972746331237,
      "grad_norm": 2.2235395908355713,
      "learning_rate": 3.481162514059903e-05,
      "loss": 0.1959,
      "num_input_tokens_seen": 10837360,
      "step": 16580
    },
    {
      "epoch": 8.692348008385745,
      "grad_norm": 1.9297282695770264,
      "learning_rate": 3.480110723749116e-05,
      "loss": 0.2569,
      "num_input_tokens_seen": 10841168,
      "step": 16585
    },
    {
      "epoch": 8.69496855345912,
      "grad_norm": 4.060883522033691,
      "learning_rate": 3.479058728410348e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 10844784,
      "step": 16590
    },
    {
      "epoch": 8.697589098532495,
      "grad_norm": 2.418910026550293,
      "learning_rate": 3.478006528263663e-05,
      "loss": 0.1881,
      "num_input_tokens_seen": 10847376,
      "step": 16595
    },
    {
      "epoch": 8.70020964360587,
      "grad_norm": 4.598176002502441,
      "learning_rate": 3.4769541235291704e-05,
      "loss": 0.1493,
      "num_input_tokens_seen": 10850000,
      "step": 16600
    },
    {
      "epoch": 8.702830188679245,
      "grad_norm": 2.3977725505828857,
      "learning_rate": 3.475901514427021e-05,
      "loss": 0.1242,
      "num_input_tokens_seen": 10852528,
      "step": 16605
    },
    {
      "epoch": 8.70545073375262,
      "grad_norm": 2.5803966522216797,
      "learning_rate": 3.4748487011774075e-05,
      "loss": 0.1504,
      "num_input_tokens_seen": 10855312,
      "step": 16610
    },
    {
      "epoch": 8.708071278825996,
      "grad_norm": 1.9526454210281372,
      "learning_rate": 3.473795684000569e-05,
      "loss": 0.3024,
      "num_input_tokens_seen": 10859376,
      "step": 16615
    },
    {
      "epoch": 8.71069182389937,
      "grad_norm": 2.120305299758911,
      "learning_rate": 3.4727424631167815e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 10862224,
      "step": 16620
    },
    {
      "epoch": 8.713312368972746,
      "grad_norm": 2.2840960025787354,
      "learning_rate": 3.471689038746369e-05,
      "loss": 0.1903,
      "num_input_tokens_seen": 10865104,
      "step": 16625
    },
    {
      "epoch": 8.715932914046121,
      "grad_norm": 2.9647891521453857,
      "learning_rate": 3.470635411109695e-05,
      "loss": 0.1487,
      "num_input_tokens_seen": 10867792,
      "step": 16630
    },
    {
      "epoch": 8.718553459119496,
      "grad_norm": 1.5780328512191772,
      "learning_rate": 3.469581580427167e-05,
      "loss": 0.1898,
      "num_input_tokens_seen": 10870768,
      "step": 16635
    },
    {
      "epoch": 8.721174004192871,
      "grad_norm": 1.8435559272766113,
      "learning_rate": 3.4685275469192335e-05,
      "loss": 0.1372,
      "num_input_tokens_seen": 10875056,
      "step": 16640
    },
    {
      "epoch": 8.723794549266248,
      "grad_norm": 2.862517833709717,
      "learning_rate": 3.467473310806387e-05,
      "loss": 0.2025,
      "num_input_tokens_seen": 10878288,
      "step": 16645
    },
    {
      "epoch": 8.726415094339622,
      "grad_norm": 10.452080726623535,
      "learning_rate": 3.466418872309161e-05,
      "loss": 0.2226,
      "num_input_tokens_seen": 10880720,
      "step": 16650
    },
    {
      "epoch": 8.729035639412999,
      "grad_norm": 1.1198738813400269,
      "learning_rate": 3.465364231648133e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 10884528,
      "step": 16655
    },
    {
      "epoch": 8.731656184486374,
      "grad_norm": 2.315084457397461,
      "learning_rate": 3.464309389043921e-05,
      "loss": 0.23,
      "num_input_tokens_seen": 10886960,
      "step": 16660
    },
    {
      "epoch": 8.734276729559749,
      "grad_norm": 1.535388469696045,
      "learning_rate": 3.4632543447171856e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 10889904,
      "step": 16665
    },
    {
      "epoch": 8.736897274633124,
      "grad_norm": 1.7160090208053589,
      "learning_rate": 3.4621990988886315e-05,
      "loss": 0.2261,
      "num_input_tokens_seen": 10893072,
      "step": 16670
    },
    {
      "epoch": 8.7395178197065,
      "grad_norm": 2.1790897846221924,
      "learning_rate": 3.461143651779003e-05,
      "loss": 0.2305,
      "num_input_tokens_seen": 10897392,
      "step": 16675
    },
    {
      "epoch": 8.742138364779874,
      "grad_norm": 1.447211742401123,
      "learning_rate": 3.4600880036090876e-05,
      "loss": 0.22,
      "num_input_tokens_seen": 10901008,
      "step": 16680
    },
    {
      "epoch": 8.74475890985325,
      "grad_norm": 1.6379258632659912,
      "learning_rate": 3.459032154599715e-05,
      "loss": 0.2071,
      "num_input_tokens_seen": 10904848,
      "step": 16685
    },
    {
      "epoch": 8.747379454926625,
      "grad_norm": 2.56962513923645,
      "learning_rate": 3.457976104971757e-05,
      "loss": 0.1792,
      "num_input_tokens_seen": 10908528,
      "step": 16690
    },
    {
      "epoch": 8.75,
      "grad_norm": 1.6696372032165527,
      "learning_rate": 3.4569198549461276e-05,
      "loss": 0.2333,
      "num_input_tokens_seen": 10912080,
      "step": 16695
    },
    {
      "epoch": 8.752620545073375,
      "grad_norm": 2.4494779109954834,
      "learning_rate": 3.455863404743781e-05,
      "loss": 0.2358,
      "num_input_tokens_seen": 10916528,
      "step": 16700
    },
    {
      "epoch": 8.75524109014675,
      "grad_norm": 2.072155714035034,
      "learning_rate": 3.454806754585716e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 10920688,
      "step": 16705
    },
    {
      "epoch": 8.757861635220126,
      "grad_norm": 2.331287145614624,
      "learning_rate": 3.453749904692971e-05,
      "loss": 0.1899,
      "num_input_tokens_seen": 10923600,
      "step": 16710
    },
    {
      "epoch": 8.7604821802935,
      "grad_norm": 2.049318552017212,
      "learning_rate": 3.452692855286626e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 10926544,
      "step": 16715
    },
    {
      "epoch": 8.763102725366876,
      "grad_norm": 2.871429204940796,
      "learning_rate": 3.4516356065878055e-05,
      "loss": 0.1716,
      "num_input_tokens_seen": 10929392,
      "step": 16720
    },
    {
      "epoch": 8.765723270440251,
      "grad_norm": 1.7704851627349854,
      "learning_rate": 3.450578158817671e-05,
      "loss": 0.1528,
      "num_input_tokens_seen": 10933456,
      "step": 16725
    },
    {
      "epoch": 8.768343815513626,
      "grad_norm": 4.184007167816162,
      "learning_rate": 3.4495205121974325e-05,
      "loss": 0.2361,
      "num_input_tokens_seen": 10936336,
      "step": 16730
    },
    {
      "epoch": 8.770964360587001,
      "grad_norm": 5.705991268157959,
      "learning_rate": 3.448462666948335e-05,
      "loss": 0.1979,
      "num_input_tokens_seen": 10939696,
      "step": 16735
    },
    {
      "epoch": 8.773584905660378,
      "grad_norm": 1.814212679862976,
      "learning_rate": 3.4474046232916674e-05,
      "loss": 0.1292,
      "num_input_tokens_seen": 10942192,
      "step": 16740
    },
    {
      "epoch": 8.776205450733752,
      "grad_norm": 1.7416234016418457,
      "learning_rate": 3.4463463814487616e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 10944624,
      "step": 16745
    },
    {
      "epoch": 8.778825995807129,
      "grad_norm": 1.722973108291626,
      "learning_rate": 3.44528794164099e-05,
      "loss": 0.1389,
      "num_input_tokens_seen": 10947952,
      "step": 16750
    },
    {
      "epoch": 8.781446540880504,
      "grad_norm": 2.2070324420928955,
      "learning_rate": 3.4442293040897646e-05,
      "loss": 0.1568,
      "num_input_tokens_seen": 10951440,
      "step": 16755
    },
    {
      "epoch": 8.784067085953879,
      "grad_norm": 1.6845296621322632,
      "learning_rate": 3.443170469016542e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 10954224,
      "step": 16760
    },
    {
      "epoch": 8.786687631027254,
      "grad_norm": 2.5984761714935303,
      "learning_rate": 3.442111436642817e-05,
      "loss": 0.1811,
      "num_input_tokens_seen": 10957392,
      "step": 16765
    },
    {
      "epoch": 8.78930817610063,
      "grad_norm": 1.3434154987335205,
      "learning_rate": 3.441052207190127e-05,
      "loss": 0.235,
      "num_input_tokens_seen": 10961136,
      "step": 16770
    },
    {
      "epoch": 8.791928721174004,
      "grad_norm": 1.807708501815796,
      "learning_rate": 3.4399927808800526e-05,
      "loss": 0.199,
      "num_input_tokens_seen": 10964272,
      "step": 16775
    },
    {
      "epoch": 8.79454926624738,
      "grad_norm": 2.8077616691589355,
      "learning_rate": 3.438933157934211e-05,
      "loss": 0.1107,
      "num_input_tokens_seen": 10967344,
      "step": 16780
    },
    {
      "epoch": 8.797169811320755,
      "grad_norm": 2.8890395164489746,
      "learning_rate": 3.4378733385742653e-05,
      "loss": 0.1867,
      "num_input_tokens_seen": 10970800,
      "step": 16785
    },
    {
      "epoch": 8.79979035639413,
      "grad_norm": 1.3472661972045898,
      "learning_rate": 3.436813323021918e-05,
      "loss": 0.1527,
      "num_input_tokens_seen": 10974544,
      "step": 16790
    },
    {
      "epoch": 8.802410901467505,
      "grad_norm": 3.2723548412323,
      "learning_rate": 3.43575311149891e-05,
      "loss": 0.202,
      "num_input_tokens_seen": 10976784,
      "step": 16795
    },
    {
      "epoch": 8.80503144654088,
      "grad_norm": 1.945166826248169,
      "learning_rate": 3.434692704227028e-05,
      "loss": 0.2263,
      "num_input_tokens_seen": 10980400,
      "step": 16800
    },
    {
      "epoch": 8.807651991614255,
      "grad_norm": 1.9365262985229492,
      "learning_rate": 3.4336321014280944e-05,
      "loss": 0.1349,
      "num_input_tokens_seen": 10986160,
      "step": 16805
    },
    {
      "epoch": 8.81027253668763,
      "grad_norm": 1.535451054573059,
      "learning_rate": 3.432571303323977e-05,
      "loss": 0.1785,
      "num_input_tokens_seen": 10989584,
      "step": 16810
    },
    {
      "epoch": 8.812893081761006,
      "grad_norm": 2.766631841659546,
      "learning_rate": 3.431510310136583e-05,
      "loss": 0.2367,
      "num_input_tokens_seen": 10992464,
      "step": 16815
    },
    {
      "epoch": 8.815513626834381,
      "grad_norm": 1.8980069160461426,
      "learning_rate": 3.4304491220878585e-05,
      "loss": 0.2234,
      "num_input_tokens_seen": 10995472,
      "step": 16820
    },
    {
      "epoch": 8.818134171907756,
      "grad_norm": 3.5462265014648438,
      "learning_rate": 3.429387739399793e-05,
      "loss": 0.1646,
      "num_input_tokens_seen": 10999088,
      "step": 16825
    },
    {
      "epoch": 8.820754716981131,
      "grad_norm": 2.8981595039367676,
      "learning_rate": 3.428326162294416e-05,
      "loss": 0.1696,
      "num_input_tokens_seen": 11002000,
      "step": 16830
    },
    {
      "epoch": 8.823375262054507,
      "grad_norm": 2.8151748180389404,
      "learning_rate": 3.427264390993795e-05,
      "loss": 0.2176,
      "num_input_tokens_seen": 11005360,
      "step": 16835
    },
    {
      "epoch": 8.825995807127882,
      "grad_norm": 1.7899035215377808,
      "learning_rate": 3.426202425720044e-05,
      "loss": 0.1495,
      "num_input_tokens_seen": 11008272,
      "step": 16840
    },
    {
      "epoch": 8.828616352201259,
      "grad_norm": 3.095613479614258,
      "learning_rate": 3.425140266695312e-05,
      "loss": 0.1844,
      "num_input_tokens_seen": 11011760,
      "step": 16845
    },
    {
      "epoch": 8.831236897274634,
      "grad_norm": 1.9137204885482788,
      "learning_rate": 3.4240779141417905e-05,
      "loss": 0.1786,
      "num_input_tokens_seen": 11014736,
      "step": 16850
    },
    {
      "epoch": 8.833857442348009,
      "grad_norm": 1.947016954421997,
      "learning_rate": 3.423015368281711e-05,
      "loss": 0.1819,
      "num_input_tokens_seen": 11018128,
      "step": 16855
    },
    {
      "epoch": 8.836477987421384,
      "grad_norm": 1.1913211345672607,
      "learning_rate": 3.421952629337348e-05,
      "loss": 0.1328,
      "num_input_tokens_seen": 11022032,
      "step": 16860
    },
    {
      "epoch": 8.83909853249476,
      "grad_norm": 2.6249685287475586,
      "learning_rate": 3.420889697531013e-05,
      "loss": 0.152,
      "num_input_tokens_seen": 11024944,
      "step": 16865
    },
    {
      "epoch": 8.841719077568134,
      "grad_norm": 2.0252199172973633,
      "learning_rate": 3.419826573085058e-05,
      "loss": 0.1727,
      "num_input_tokens_seen": 11028272,
      "step": 16870
    },
    {
      "epoch": 8.84433962264151,
      "grad_norm": 5.949840068817139,
      "learning_rate": 3.418763256221879e-05,
      "loss": 0.1695,
      "num_input_tokens_seen": 11031088,
      "step": 16875
    },
    {
      "epoch": 8.846960167714885,
      "grad_norm": 5.931471347808838,
      "learning_rate": 3.4176997471639085e-05,
      "loss": 0.2177,
      "num_input_tokens_seen": 11033456,
      "step": 16880
    },
    {
      "epoch": 8.84958071278826,
      "grad_norm": 1.7035247087478638,
      "learning_rate": 3.4166360461336206e-05,
      "loss": 0.2381,
      "num_input_tokens_seen": 11036976,
      "step": 16885
    },
    {
      "epoch": 8.852201257861635,
      "grad_norm": 3.3474841117858887,
      "learning_rate": 3.41557215335353e-05,
      "loss": 0.1812,
      "num_input_tokens_seen": 11039472,
      "step": 16890
    },
    {
      "epoch": 8.85482180293501,
      "grad_norm": 2.401527166366577,
      "learning_rate": 3.4145080690461887e-05,
      "loss": 0.2005,
      "num_input_tokens_seen": 11042064,
      "step": 16895
    },
    {
      "epoch": 8.857442348008385,
      "grad_norm": 2.3557581901550293,
      "learning_rate": 3.4134437934341936e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 11045232,
      "step": 16900
    },
    {
      "epoch": 8.86006289308176,
      "grad_norm": 2.1589889526367188,
      "learning_rate": 3.4123793267401784e-05,
      "loss": 0.1847,
      "num_input_tokens_seen": 11048784,
      "step": 16905
    },
    {
      "epoch": 8.862683438155136,
      "grad_norm": 1.3479530811309814,
      "learning_rate": 3.411314669186816e-05,
      "loss": 0.2113,
      "num_input_tokens_seen": 11052528,
      "step": 16910
    },
    {
      "epoch": 8.865303983228511,
      "grad_norm": 4.544378757476807,
      "learning_rate": 3.410249820996823e-05,
      "loss": 0.227,
      "num_input_tokens_seen": 11055824,
      "step": 16915
    },
    {
      "epoch": 8.867924528301886,
      "grad_norm": 1.7607927322387695,
      "learning_rate": 3.4091847823929513e-05,
      "loss": 0.1316,
      "num_input_tokens_seen": 11059504,
      "step": 16920
    },
    {
      "epoch": 8.870545073375261,
      "grad_norm": 1.9552778005599976,
      "learning_rate": 3.4081195535979946e-05,
      "loss": 0.142,
      "num_input_tokens_seen": 11062544,
      "step": 16925
    },
    {
      "epoch": 8.873165618448636,
      "grad_norm": 2.4308266639709473,
      "learning_rate": 3.40705413483479e-05,
      "loss": 0.1961,
      "num_input_tokens_seen": 11065584,
      "step": 16930
    },
    {
      "epoch": 8.875786163522012,
      "grad_norm": 2.478379249572754,
      "learning_rate": 3.405988526326207e-05,
      "loss": 0.1175,
      "num_input_tokens_seen": 11068336,
      "step": 16935
    },
    {
      "epoch": 8.878406708595389,
      "grad_norm": 2.307340621948242,
      "learning_rate": 3.40492272829516e-05,
      "loss": 0.1554,
      "num_input_tokens_seen": 11071184,
      "step": 16940
    },
    {
      "epoch": 8.881027253668764,
      "grad_norm": 1.4906883239746094,
      "learning_rate": 3.403856740964604e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 11074320,
      "step": 16945
    },
    {
      "epoch": 8.883647798742139,
      "grad_norm": 2.452709913253784,
      "learning_rate": 3.4027905645575275e-05,
      "loss": 0.147,
      "num_input_tokens_seen": 11077168,
      "step": 16950
    },
    {
      "epoch": 8.886268343815514,
      "grad_norm": 4.010046482086182,
      "learning_rate": 3.401724199296966e-05,
      "loss": 0.1873,
      "num_input_tokens_seen": 11080080,
      "step": 16955
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 2.1737112998962402,
      "learning_rate": 3.400657645405989e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 11083408,
      "step": 16960
    },
    {
      "epoch": 8.891509433962264,
      "grad_norm": 2.6236302852630615,
      "learning_rate": 3.3995909031077065e-05,
      "loss": 0.2243,
      "num_input_tokens_seen": 11087184,
      "step": 16965
    },
    {
      "epoch": 8.89412997903564,
      "grad_norm": 1.3126825094223022,
      "learning_rate": 3.398523972625272e-05,
      "loss": 0.0949,
      "num_input_tokens_seen": 11090864,
      "step": 16970
    },
    {
      "epoch": 8.896750524109015,
      "grad_norm": 1.3036116361618042,
      "learning_rate": 3.3974568541818723e-05,
      "loss": 0.1172,
      "num_input_tokens_seen": 11094288,
      "step": 16975
    },
    {
      "epoch": 8.89937106918239,
      "grad_norm": 1.9983091354370117,
      "learning_rate": 3.396389548000738e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 11096656,
      "step": 16980
    },
    {
      "epoch": 8.901991614255765,
      "grad_norm": 1.4303326606750488,
      "learning_rate": 3.395322054305137e-05,
      "loss": 0.1859,
      "num_input_tokens_seen": 11100144,
      "step": 16985
    },
    {
      "epoch": 8.90461215932914,
      "grad_norm": 1.481019377708435,
      "learning_rate": 3.394254373318376e-05,
      "loss": 0.1686,
      "num_input_tokens_seen": 11103440,
      "step": 16990
    },
    {
      "epoch": 8.907232704402515,
      "grad_norm": 2.4690909385681152,
      "learning_rate": 3.3931865052638026e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 11107056,
      "step": 16995
    },
    {
      "epoch": 8.90985324947589,
      "grad_norm": 3.5592164993286133,
      "learning_rate": 3.392118450364802e-05,
      "loss": 0.1567,
      "num_input_tokens_seen": 11110640,
      "step": 17000
    },
    {
      "epoch": 8.912473794549266,
      "grad_norm": 0.37388864159584045,
      "learning_rate": 3.391050208844799e-05,
      "loss": 0.124,
      "num_input_tokens_seen": 11117008,
      "step": 17005
    },
    {
      "epoch": 8.915094339622641,
      "grad_norm": 1.830063819885254,
      "learning_rate": 3.389981780927259e-05,
      "loss": 0.1377,
      "num_input_tokens_seen": 11120176,
      "step": 17010
    },
    {
      "epoch": 8.917714884696016,
      "grad_norm": 2.270888328552246,
      "learning_rate": 3.3889131668356834e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 11123504,
      "step": 17015
    },
    {
      "epoch": 8.920335429769391,
      "grad_norm": 2.542530059814453,
      "learning_rate": 3.387844366793614e-05,
      "loss": 0.2625,
      "num_input_tokens_seen": 11126000,
      "step": 17020
    },
    {
      "epoch": 8.922955974842766,
      "grad_norm": 1.5016415119171143,
      "learning_rate": 3.3867753810246326e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 11129584,
      "step": 17025
    },
    {
      "epoch": 8.925576519916142,
      "grad_norm": 2.5978891849517822,
      "learning_rate": 3.3857062097523586e-05,
      "loss": 0.2012,
      "num_input_tokens_seen": 11132240,
      "step": 17030
    },
    {
      "epoch": 8.928197064989519,
      "grad_norm": 2.4791510105133057,
      "learning_rate": 3.38463685320045e-05,
      "loss": 0.1822,
      "num_input_tokens_seen": 11135120,
      "step": 17035
    },
    {
      "epoch": 8.930817610062894,
      "grad_norm": 2.6651010513305664,
      "learning_rate": 3.3835673115926045e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 11137968,
      "step": 17040
    },
    {
      "epoch": 8.933438155136269,
      "grad_norm": 1.4959603548049927,
      "learning_rate": 3.382497585152557e-05,
      "loss": 0.1638,
      "num_input_tokens_seen": 11140368,
      "step": 17045
    },
    {
      "epoch": 8.936058700209644,
      "grad_norm": 3.3373076915740967,
      "learning_rate": 3.3814276741040835e-05,
      "loss": 0.2064,
      "num_input_tokens_seen": 11145168,
      "step": 17050
    },
    {
      "epoch": 8.93867924528302,
      "grad_norm": 1.723082184791565,
      "learning_rate": 3.380357578670996e-05,
      "loss": 0.1919,
      "num_input_tokens_seen": 11149008,
      "step": 17055
    },
    {
      "epoch": 8.941299790356394,
      "grad_norm": 2.414355993270874,
      "learning_rate": 3.379287299077146e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 11156336,
      "step": 17060
    },
    {
      "epoch": 8.94392033542977,
      "grad_norm": 3.884629011154175,
      "learning_rate": 3.3782168355464263e-05,
      "loss": 0.2454,
      "num_input_tokens_seen": 11159152,
      "step": 17065
    },
    {
      "epoch": 8.946540880503145,
      "grad_norm": 2.680210590362549,
      "learning_rate": 3.377146188302762e-05,
      "loss": 0.1854,
      "num_input_tokens_seen": 11162096,
      "step": 17070
    },
    {
      "epoch": 8.94916142557652,
      "grad_norm": 3.329530715942383,
      "learning_rate": 3.376075357570124e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 11165200,
      "step": 17075
    },
    {
      "epoch": 8.951781970649895,
      "grad_norm": 3.101600408554077,
      "learning_rate": 3.3750043435725144e-05,
      "loss": 0.2237,
      "num_input_tokens_seen": 11168368,
      "step": 17080
    },
    {
      "epoch": 8.95440251572327,
      "grad_norm": 1.7697007656097412,
      "learning_rate": 3.373933146533979e-05,
      "loss": 0.1986,
      "num_input_tokens_seen": 11172048,
      "step": 17085
    },
    {
      "epoch": 8.957023060796645,
      "grad_norm": 1.9723565578460693,
      "learning_rate": 3.3728617666786e-05,
      "loss": 0.1746,
      "num_input_tokens_seen": 11175600,
      "step": 17090
    },
    {
      "epoch": 8.95964360587002,
      "grad_norm": 4.108666896820068,
      "learning_rate": 3.371790204230497e-05,
      "loss": 0.188,
      "num_input_tokens_seen": 11178576,
      "step": 17095
    },
    {
      "epoch": 8.962264150943396,
      "grad_norm": 2.1938579082489014,
      "learning_rate": 3.3707184594138295e-05,
      "loss": 0.1999,
      "num_input_tokens_seen": 11181168,
      "step": 17100
    },
    {
      "epoch": 8.964884696016771,
      "grad_norm": 2.151735782623291,
      "learning_rate": 3.369646532452795e-05,
      "loss": 0.1465,
      "num_input_tokens_seen": 11184400,
      "step": 17105
    },
    {
      "epoch": 8.967505241090146,
      "grad_norm": 1.8322619199752808,
      "learning_rate": 3.3685744235716254e-05,
      "loss": 0.2313,
      "num_input_tokens_seen": 11187536,
      "step": 17110
    },
    {
      "epoch": 8.970125786163521,
      "grad_norm": 2.117065668106079,
      "learning_rate": 3.367502132994598e-05,
      "loss": 0.1938,
      "num_input_tokens_seen": 11191568,
      "step": 17115
    },
    {
      "epoch": 8.972746331236896,
      "grad_norm": 2.6390719413757324,
      "learning_rate": 3.366429660946019e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 11195024,
      "step": 17120
    },
    {
      "epoch": 8.975366876310272,
      "grad_norm": 2.823573112487793,
      "learning_rate": 3.365357007650241e-05,
      "loss": 0.2183,
      "num_input_tokens_seen": 11198480,
      "step": 17125
    },
    {
      "epoch": 8.977987421383649,
      "grad_norm": 2.2533247470855713,
      "learning_rate": 3.364284173331649e-05,
      "loss": 0.2013,
      "num_input_tokens_seen": 11201712,
      "step": 17130
    },
    {
      "epoch": 8.980607966457024,
      "grad_norm": 1.6889317035675049,
      "learning_rate": 3.3632111582146686e-05,
      "loss": 0.138,
      "num_input_tokens_seen": 11204720,
      "step": 17135
    },
    {
      "epoch": 8.983228511530399,
      "grad_norm": 1.8561409711837769,
      "learning_rate": 3.362137962523761e-05,
      "loss": 0.1985,
      "num_input_tokens_seen": 11207440,
      "step": 17140
    },
    {
      "epoch": 8.985849056603774,
      "grad_norm": 1.9891055822372437,
      "learning_rate": 3.361064586483429e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 11210384,
      "step": 17145
    },
    {
      "epoch": 8.98846960167715,
      "grad_norm": 2.2130699157714844,
      "learning_rate": 3.359991030318208e-05,
      "loss": 0.1865,
      "num_input_tokens_seen": 11213392,
      "step": 17150
    },
    {
      "epoch": 8.991090146750524,
      "grad_norm": 2.100940465927124,
      "learning_rate": 3.358917294252675e-05,
      "loss": 0.225,
      "num_input_tokens_seen": 11216016,
      "step": 17155
    },
    {
      "epoch": 8.9937106918239,
      "grad_norm": 1.940007209777832,
      "learning_rate": 3.357843378511443e-05,
      "loss": 0.1796,
      "num_input_tokens_seen": 11218160,
      "step": 17160
    },
    {
      "epoch": 8.996331236897275,
      "grad_norm": 3.269497871398926,
      "learning_rate": 3.3567692833191635e-05,
      "loss": 0.3303,
      "num_input_tokens_seen": 11222000,
      "step": 17165
    },
    {
      "epoch": 8.99895178197065,
      "grad_norm": 1.70021390914917,
      "learning_rate": 3.355695008900525e-05,
      "loss": 0.1853,
      "num_input_tokens_seen": 11224816,
      "step": 17170
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.7038738131523132,
      "eval_runtime": 11.6566,
      "eval_samples_per_second": 72.749,
      "eval_steps_per_second": 18.187,
      "num_input_tokens_seen": 11225416,
      "step": 17172
    },
    {
      "epoch": 9.001572327044025,
      "grad_norm": 2.2338738441467285,
      "learning_rate": 3.354620555480251e-05,
      "loss": 0.189,
      "num_input_tokens_seen": 11227112,
      "step": 17175
    },
    {
      "epoch": 9.0041928721174,
      "grad_norm": 1.177148461341858,
      "learning_rate": 3.353545923283108e-05,
      "loss": 0.1216,
      "num_input_tokens_seen": 11229544,
      "step": 17180
    },
    {
      "epoch": 9.006813417190775,
      "grad_norm": 1.572739839553833,
      "learning_rate": 3.352471112533895e-05,
      "loss": 0.1724,
      "num_input_tokens_seen": 11233640,
      "step": 17185
    },
    {
      "epoch": 9.00943396226415,
      "grad_norm": 1.9590831995010376,
      "learning_rate": 3.351396123457451e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 11236552,
      "step": 17190
    },
    {
      "epoch": 9.012054507337526,
      "grad_norm": 18.072154998779297,
      "learning_rate": 3.350320956278652e-05,
      "loss": 0.1698,
      "num_input_tokens_seen": 11240008,
      "step": 17195
    },
    {
      "epoch": 9.014675052410901,
      "grad_norm": 2.457444667816162,
      "learning_rate": 3.349245611222409e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 11243176,
      "step": 17200
    },
    {
      "epoch": 9.017295597484276,
      "grad_norm": 1.6989799737930298,
      "learning_rate": 3.348170088513673e-05,
      "loss": 0.1206,
      "num_input_tokens_seen": 11246568,
      "step": 17205
    },
    {
      "epoch": 9.019916142557651,
      "grad_norm": 1.202537178993225,
      "learning_rate": 3.347094388377431e-05,
      "loss": 0.1037,
      "num_input_tokens_seen": 11251080,
      "step": 17210
    },
    {
      "epoch": 9.022536687631026,
      "grad_norm": 5.189813137054443,
      "learning_rate": 3.3460185110387076e-05,
      "loss": 0.1647,
      "num_input_tokens_seen": 11253256,
      "step": 17215
    },
    {
      "epoch": 9.025157232704403,
      "grad_norm": 1.8758859634399414,
      "learning_rate": 3.344942456722563e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 11256936,
      "step": 17220
    },
    {
      "epoch": 9.027777777777779,
      "grad_norm": 2.937217950820923,
      "learning_rate": 3.343866225654096e-05,
      "loss": 0.106,
      "num_input_tokens_seen": 11259784,
      "step": 17225
    },
    {
      "epoch": 9.030398322851154,
      "grad_norm": 1.619920253753662,
      "learning_rate": 3.342789818058443e-05,
      "loss": 0.1303,
      "num_input_tokens_seen": 11262728,
      "step": 17230
    },
    {
      "epoch": 9.033018867924529,
      "grad_norm": 2.1665797233581543,
      "learning_rate": 3.341713234160774e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 11265640,
      "step": 17235
    },
    {
      "epoch": 9.035639412997904,
      "grad_norm": 2.516555070877075,
      "learning_rate": 3.3406364741862986e-05,
      "loss": 0.1256,
      "num_input_tokens_seen": 11270344,
      "step": 17240
    },
    {
      "epoch": 9.03825995807128,
      "grad_norm": 8.573737144470215,
      "learning_rate": 3.3395595383602644e-05,
      "loss": 0.1388,
      "num_input_tokens_seen": 11272584,
      "step": 17245
    },
    {
      "epoch": 9.040880503144654,
      "grad_norm": 1.997445821762085,
      "learning_rate": 3.338482426907953e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 11276872,
      "step": 17250
    },
    {
      "epoch": 9.04350104821803,
      "grad_norm": 2.196409225463867,
      "learning_rate": 3.337405140054683e-05,
      "loss": 0.138,
      "num_input_tokens_seen": 11279688,
      "step": 17255
    },
    {
      "epoch": 9.046121593291405,
      "grad_norm": 2.5843729972839355,
      "learning_rate": 3.336327678025813e-05,
      "loss": 0.1134,
      "num_input_tokens_seen": 11282888,
      "step": 17260
    },
    {
      "epoch": 9.04874213836478,
      "grad_norm": 2.9832940101623535,
      "learning_rate": 3.3352500410467324e-05,
      "loss": 0.2121,
      "num_input_tokens_seen": 11285928,
      "step": 17265
    },
    {
      "epoch": 9.051362683438155,
      "grad_norm": 4.462443828582764,
      "learning_rate": 3.3341722293428726e-05,
      "loss": 0.1313,
      "num_input_tokens_seen": 11289288,
      "step": 17270
    },
    {
      "epoch": 9.05398322851153,
      "grad_norm": 1.3532663583755493,
      "learning_rate": 3.3330942431397e-05,
      "loss": 0.1305,
      "num_input_tokens_seen": 11295976,
      "step": 17275
    },
    {
      "epoch": 9.056603773584905,
      "grad_norm": 1.3924949169158936,
      "learning_rate": 3.332016082662715e-05,
      "loss": 0.134,
      "num_input_tokens_seen": 11299816,
      "step": 17280
    },
    {
      "epoch": 9.05922431865828,
      "grad_norm": 1.4854735136032104,
      "learning_rate": 3.3309377481374584e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 11303176,
      "step": 17285
    },
    {
      "epoch": 9.061844863731656,
      "grad_norm": 2.4068551063537598,
      "learning_rate": 3.3298592397895046e-05,
      "loss": 0.0857,
      "num_input_tokens_seen": 11306376,
      "step": 17290
    },
    {
      "epoch": 9.064465408805031,
      "grad_norm": 1.3161211013793945,
      "learning_rate": 3.3287805578444645e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 11310696,
      "step": 17295
    },
    {
      "epoch": 9.067085953878406,
      "grad_norm": 2.8004491329193115,
      "learning_rate": 3.327701702527989e-05,
      "loss": 0.1175,
      "num_input_tokens_seen": 11314888,
      "step": 17300
    },
    {
      "epoch": 9.069706498951781,
      "grad_norm": 3.0941050052642822,
      "learning_rate": 3.3266226740657584e-05,
      "loss": 0.1268,
      "num_input_tokens_seen": 11317576,
      "step": 17305
    },
    {
      "epoch": 9.072327044025156,
      "grad_norm": 2.9195139408111572,
      "learning_rate": 3.3255434726834954e-05,
      "loss": 0.1574,
      "num_input_tokens_seen": 11320840,
      "step": 17310
    },
    {
      "epoch": 9.074947589098532,
      "grad_norm": 1.7299765348434448,
      "learning_rate": 3.324464098606957e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 11324040,
      "step": 17315
    },
    {
      "epoch": 9.077568134171909,
      "grad_norm": 3.0371012687683105,
      "learning_rate": 3.323384552061934e-05,
      "loss": 0.1495,
      "num_input_tokens_seen": 11327784,
      "step": 17320
    },
    {
      "epoch": 9.080188679245284,
      "grad_norm": 1.195084571838379,
      "learning_rate": 3.322304833274257e-05,
      "loss": 0.1123,
      "num_input_tokens_seen": 11331304,
      "step": 17325
    },
    {
      "epoch": 9.082809224318659,
      "grad_norm": 1.7770240306854248,
      "learning_rate": 3.32122494246979e-05,
      "loss": 0.1465,
      "num_input_tokens_seen": 11334312,
      "step": 17330
    },
    {
      "epoch": 9.085429769392034,
      "grad_norm": 1.1110436916351318,
      "learning_rate": 3.320144879874434e-05,
      "loss": 0.1615,
      "num_input_tokens_seen": 11339144,
      "step": 17335
    },
    {
      "epoch": 9.08805031446541,
      "grad_norm": 7.528820037841797,
      "learning_rate": 3.319064645714126e-05,
      "loss": 0.1125,
      "num_input_tokens_seen": 11342664,
      "step": 17340
    },
    {
      "epoch": 9.090670859538784,
      "grad_norm": 3.210130214691162,
      "learning_rate": 3.3179842402148386e-05,
      "loss": 0.1315,
      "num_input_tokens_seen": 11345768,
      "step": 17345
    },
    {
      "epoch": 9.09329140461216,
      "grad_norm": 2.6637299060821533,
      "learning_rate": 3.316903663602581e-05,
      "loss": 0.1887,
      "num_input_tokens_seen": 11349448,
      "step": 17350
    },
    {
      "epoch": 9.095911949685535,
      "grad_norm": 3.3259549140930176,
      "learning_rate": 3.315822916103395e-05,
      "loss": 0.1109,
      "num_input_tokens_seen": 11352744,
      "step": 17355
    },
    {
      "epoch": 9.09853249475891,
      "grad_norm": 5.056654930114746,
      "learning_rate": 3.3147419979433635e-05,
      "loss": 0.1153,
      "num_input_tokens_seen": 11354920,
      "step": 17360
    },
    {
      "epoch": 9.101153039832285,
      "grad_norm": 2.0628910064697266,
      "learning_rate": 3.313660909348601e-05,
      "loss": 0.1151,
      "num_input_tokens_seen": 11359240,
      "step": 17365
    },
    {
      "epoch": 9.10377358490566,
      "grad_norm": 2.4469048976898193,
      "learning_rate": 3.312579650545259e-05,
      "loss": 0.1934,
      "num_input_tokens_seen": 11362056,
      "step": 17370
    },
    {
      "epoch": 9.106394129979035,
      "grad_norm": 1.8330423831939697,
      "learning_rate": 3.311498221759524e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 11364648,
      "step": 17375
    },
    {
      "epoch": 9.10901467505241,
      "grad_norm": 1.3366847038269043,
      "learning_rate": 3.310416623217619e-05,
      "loss": 0.091,
      "num_input_tokens_seen": 11368168,
      "step": 17380
    },
    {
      "epoch": 9.111635220125786,
      "grad_norm": 1.2568466663360596,
      "learning_rate": 3.309334855145803e-05,
      "loss": 0.1208,
      "num_input_tokens_seen": 11370856,
      "step": 17385
    },
    {
      "epoch": 9.114255765199161,
      "grad_norm": 2.838200807571411,
      "learning_rate": 3.3082529177703684e-05,
      "loss": 0.1291,
      "num_input_tokens_seen": 11373768,
      "step": 17390
    },
    {
      "epoch": 9.116876310272536,
      "grad_norm": 2.4314253330230713,
      "learning_rate": 3.3071708113176435e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 11377192,
      "step": 17395
    },
    {
      "epoch": 9.119496855345911,
      "grad_norm": 1.5259675979614258,
      "learning_rate": 3.306088536013995e-05,
      "loss": 0.1216,
      "num_input_tokens_seen": 11379912,
      "step": 17400
    },
    {
      "epoch": 9.122117400419286,
      "grad_norm": 3.894653081893921,
      "learning_rate": 3.3050060920858206e-05,
      "loss": 0.1284,
      "num_input_tokens_seen": 11383496,
      "step": 17405
    },
    {
      "epoch": 9.124737945492662,
      "grad_norm": 2.2668497562408447,
      "learning_rate": 3.3039234797595544e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 11386504,
      "step": 17410
    },
    {
      "epoch": 9.127358490566039,
      "grad_norm": 2.2497451305389404,
      "learning_rate": 3.3028406992616695e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 11388776,
      "step": 17415
    },
    {
      "epoch": 9.129979035639414,
      "grad_norm": 1.8428285121917725,
      "learning_rate": 3.301757750818669e-05,
      "loss": 0.1451,
      "num_input_tokens_seen": 11391976,
      "step": 17420
    },
    {
      "epoch": 9.132599580712789,
      "grad_norm": 1.4892754554748535,
      "learning_rate": 3.300674634657094e-05,
      "loss": 0.1286,
      "num_input_tokens_seen": 11394856,
      "step": 17425
    },
    {
      "epoch": 9.135220125786164,
      "grad_norm": 1.6246970891952515,
      "learning_rate": 3.2995913510035206e-05,
      "loss": 0.1359,
      "num_input_tokens_seen": 11397896,
      "step": 17430
    },
    {
      "epoch": 9.13784067085954,
      "grad_norm": 1.4834110736846924,
      "learning_rate": 3.298507900084558e-05,
      "loss": 0.1286,
      "num_input_tokens_seen": 11401096,
      "step": 17435
    },
    {
      "epoch": 9.140461215932914,
      "grad_norm": 1.6637704372406006,
      "learning_rate": 3.2974242821268524e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 11404616,
      "step": 17440
    },
    {
      "epoch": 9.14308176100629,
      "grad_norm": 2.5238535404205322,
      "learning_rate": 3.2963404973570855e-05,
      "loss": 0.1155,
      "num_input_tokens_seen": 11407848,
      "step": 17445
    },
    {
      "epoch": 9.145702306079665,
      "grad_norm": 6.1634039878845215,
      "learning_rate": 3.2952565460019715e-05,
      "loss": 0.1184,
      "num_input_tokens_seen": 11410760,
      "step": 17450
    },
    {
      "epoch": 9.14832285115304,
      "grad_norm": 1.7738007307052612,
      "learning_rate": 3.29417242828826e-05,
      "loss": 0.1498,
      "num_input_tokens_seen": 11413864,
      "step": 17455
    },
    {
      "epoch": 9.150943396226415,
      "grad_norm": 1.9741714000701904,
      "learning_rate": 3.2930881444427374e-05,
      "loss": 0.1315,
      "num_input_tokens_seen": 11416808,
      "step": 17460
    },
    {
      "epoch": 9.15356394129979,
      "grad_norm": 4.785383701324463,
      "learning_rate": 3.292003694692223e-05,
      "loss": 0.215,
      "num_input_tokens_seen": 11419944,
      "step": 17465
    },
    {
      "epoch": 9.156184486373165,
      "grad_norm": 3.1454403400421143,
      "learning_rate": 3.290919079263572e-05,
      "loss": 0.1225,
      "num_input_tokens_seen": 11422728,
      "step": 17470
    },
    {
      "epoch": 9.15880503144654,
      "grad_norm": 2.959354877471924,
      "learning_rate": 3.2898342983836715e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 11430984,
      "step": 17475
    },
    {
      "epoch": 9.161425576519916,
      "grad_norm": 2.156912326812744,
      "learning_rate": 3.288749352279447e-05,
      "loss": 0.129,
      "num_input_tokens_seen": 11433672,
      "step": 17480
    },
    {
      "epoch": 9.164046121593291,
      "grad_norm": 3.940140724182129,
      "learning_rate": 3.2876642411778573e-05,
      "loss": 0.1494,
      "num_input_tokens_seen": 11436168,
      "step": 17485
    },
    {
      "epoch": 9.166666666666666,
      "grad_norm": 1.7346556186676025,
      "learning_rate": 3.286578965305894e-05,
      "loss": 0.1051,
      "num_input_tokens_seen": 11439528,
      "step": 17490
    },
    {
      "epoch": 9.169287211740041,
      "grad_norm": 4.399982929229736,
      "learning_rate": 3.285493524890585e-05,
      "loss": 0.1594,
      "num_input_tokens_seen": 11442312,
      "step": 17495
    },
    {
      "epoch": 9.171907756813416,
      "grad_norm": 2.664268970489502,
      "learning_rate": 3.284407920158992e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 11445416,
      "step": 17500
    },
    {
      "epoch": 9.174528301886792,
      "grad_norm": 4.306344509124756,
      "learning_rate": 3.283322151338211e-05,
      "loss": 0.1921,
      "num_input_tokens_seen": 11448712,
      "step": 17505
    },
    {
      "epoch": 9.177148846960169,
      "grad_norm": 1.8206071853637695,
      "learning_rate": 3.282236218655373e-05,
      "loss": 0.123,
      "num_input_tokens_seen": 11452104,
      "step": 17510
    },
    {
      "epoch": 9.179769392033544,
      "grad_norm": 1.7672371864318848,
      "learning_rate": 3.281150122337642e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 11456136,
      "step": 17515
    },
    {
      "epoch": 9.182389937106919,
      "grad_norm": 3.188779592514038,
      "learning_rate": 3.280063862612216e-05,
      "loss": 0.1398,
      "num_input_tokens_seen": 11459272,
      "step": 17520
    },
    {
      "epoch": 9.185010482180294,
      "grad_norm": 2.4913694858551025,
      "learning_rate": 3.2789774397063314e-05,
      "loss": 0.1297,
      "num_input_tokens_seen": 11463208,
      "step": 17525
    },
    {
      "epoch": 9.18763102725367,
      "grad_norm": 3.6152782440185547,
      "learning_rate": 3.277890853847253e-05,
      "loss": 0.1438,
      "num_input_tokens_seen": 11466472,
      "step": 17530
    },
    {
      "epoch": 9.190251572327044,
      "grad_norm": 3.341240167617798,
      "learning_rate": 3.2768041052622824e-05,
      "loss": 0.1205,
      "num_input_tokens_seen": 11469192,
      "step": 17535
    },
    {
      "epoch": 9.19287211740042,
      "grad_norm": 1.921576738357544,
      "learning_rate": 3.275717194178755e-05,
      "loss": 0.1251,
      "num_input_tokens_seen": 11472232,
      "step": 17540
    },
    {
      "epoch": 9.195492662473795,
      "grad_norm": 3.1709303855895996,
      "learning_rate": 3.27463012082404e-05,
      "loss": 0.1529,
      "num_input_tokens_seen": 11476040,
      "step": 17545
    },
    {
      "epoch": 9.19811320754717,
      "grad_norm": 3.651822566986084,
      "learning_rate": 3.2735428854255416e-05,
      "loss": 0.133,
      "num_input_tokens_seen": 11478728,
      "step": 17550
    },
    {
      "epoch": 9.200733752620545,
      "grad_norm": 3.283104658126831,
      "learning_rate": 3.272455488210697e-05,
      "loss": 0.1426,
      "num_input_tokens_seen": 11481448,
      "step": 17555
    },
    {
      "epoch": 9.20335429769392,
      "grad_norm": 2.595454454421997,
      "learning_rate": 3.271367929406977e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 11484424,
      "step": 17560
    },
    {
      "epoch": 9.205974842767295,
      "grad_norm": 1.2565178871154785,
      "learning_rate": 3.270280209241886e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 11487560,
      "step": 17565
    },
    {
      "epoch": 9.20859538784067,
      "grad_norm": 6.564395904541016,
      "learning_rate": 3.2691923279429624e-05,
      "loss": 0.1161,
      "num_input_tokens_seen": 11490632,
      "step": 17570
    },
    {
      "epoch": 9.211215932914046,
      "grad_norm": 2.3749492168426514,
      "learning_rate": 3.2681042857377804e-05,
      "loss": 0.144,
      "num_input_tokens_seen": 11493032,
      "step": 17575
    },
    {
      "epoch": 9.213836477987421,
      "grad_norm": 2.2453320026397705,
      "learning_rate": 3.2670160828539435e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 11496200,
      "step": 17580
    },
    {
      "epoch": 9.216457023060796,
      "grad_norm": 2.246446371078491,
      "learning_rate": 3.2659277195190934e-05,
      "loss": 0.1497,
      "num_input_tokens_seen": 11498696,
      "step": 17585
    },
    {
      "epoch": 9.219077568134171,
      "grad_norm": 2.3321011066436768,
      "learning_rate": 3.2648391959609026e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 11502312,
      "step": 17590
    },
    {
      "epoch": 9.221698113207546,
      "grad_norm": 3.2055814266204834,
      "learning_rate": 3.263750512407078e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 11505160,
      "step": 17595
    },
    {
      "epoch": 9.224318658280922,
      "grad_norm": 1.2597577571868896,
      "learning_rate": 3.262661669085359e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 11508456,
      "step": 17600
    },
    {
      "epoch": 9.226939203354299,
      "grad_norm": 2.2859137058258057,
      "learning_rate": 3.2615726662235205e-05,
      "loss": 0.1511,
      "num_input_tokens_seen": 11511816,
      "step": 17605
    },
    {
      "epoch": 9.229559748427674,
      "grad_norm": 2.235961675643921,
      "learning_rate": 3.2604835040493676e-05,
      "loss": 0.1269,
      "num_input_tokens_seen": 11514088,
      "step": 17610
    },
    {
      "epoch": 9.232180293501049,
      "grad_norm": 3.1833367347717285,
      "learning_rate": 3.259394182790742e-05,
      "loss": 0.1135,
      "num_input_tokens_seen": 11517096,
      "step": 17615
    },
    {
      "epoch": 9.234800838574424,
      "grad_norm": 2.129333257675171,
      "learning_rate": 3.2583047026755174e-05,
      "loss": 0.1156,
      "num_input_tokens_seen": 11520168,
      "step": 17620
    },
    {
      "epoch": 9.2374213836478,
      "grad_norm": 2.791353225708008,
      "learning_rate": 3.2572150639316e-05,
      "loss": 0.2102,
      "num_input_tokens_seen": 11523944,
      "step": 17625
    },
    {
      "epoch": 9.240041928721174,
      "grad_norm": 1.4496586322784424,
      "learning_rate": 3.256125266786931e-05,
      "loss": 0.1774,
      "num_input_tokens_seen": 11527720,
      "step": 17630
    },
    {
      "epoch": 9.24266247379455,
      "grad_norm": 4.43626594543457,
      "learning_rate": 3.255035311469481e-05,
      "loss": 0.2222,
      "num_input_tokens_seen": 11531304,
      "step": 17635
    },
    {
      "epoch": 9.245283018867925,
      "grad_norm": 2.834303140640259,
      "learning_rate": 3.2539451982072586e-05,
      "loss": 0.131,
      "num_input_tokens_seen": 11535176,
      "step": 17640
    },
    {
      "epoch": 9.2479035639413,
      "grad_norm": 2.1399502754211426,
      "learning_rate": 3.2528549272283016e-05,
      "loss": 0.2077,
      "num_input_tokens_seen": 11539016,
      "step": 17645
    },
    {
      "epoch": 9.250524109014675,
      "grad_norm": 1.4981833696365356,
      "learning_rate": 3.251764498760683e-05,
      "loss": 0.0902,
      "num_input_tokens_seen": 11543752,
      "step": 17650
    },
    {
      "epoch": 9.25314465408805,
      "grad_norm": 3.466740369796753,
      "learning_rate": 3.2506739130325084e-05,
      "loss": 0.1315,
      "num_input_tokens_seen": 11546664,
      "step": 17655
    },
    {
      "epoch": 9.255765199161425,
      "grad_norm": 3.0858025550842285,
      "learning_rate": 3.2495831702719146e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 11549032,
      "step": 17660
    },
    {
      "epoch": 9.2583857442348,
      "grad_norm": 1.4145745038986206,
      "learning_rate": 3.248492270707074e-05,
      "loss": 0.1996,
      "num_input_tokens_seen": 11552328,
      "step": 17665
    },
    {
      "epoch": 9.261006289308176,
      "grad_norm": 3.718473196029663,
      "learning_rate": 3.247401214566189e-05,
      "loss": 0.228,
      "num_input_tokens_seen": 11555464,
      "step": 17670
    },
    {
      "epoch": 9.26362683438155,
      "grad_norm": 5.535244464874268,
      "learning_rate": 3.2463100020774965e-05,
      "loss": 0.1407,
      "num_input_tokens_seen": 11558088,
      "step": 17675
    },
    {
      "epoch": 9.266247379454926,
      "grad_norm": 3.249528169631958,
      "learning_rate": 3.245218633469267e-05,
      "loss": 0.1612,
      "num_input_tokens_seen": 11560872,
      "step": 17680
    },
    {
      "epoch": 9.268867924528301,
      "grad_norm": 3.2836616039276123,
      "learning_rate": 3.244127108969801e-05,
      "loss": 0.1098,
      "num_input_tokens_seen": 11564264,
      "step": 17685
    },
    {
      "epoch": 9.271488469601676,
      "grad_norm": 1.7928117513656616,
      "learning_rate": 3.2430354288074314e-05,
      "loss": 0.1224,
      "num_input_tokens_seen": 11567656,
      "step": 17690
    },
    {
      "epoch": 9.274109014675052,
      "grad_norm": 2.516010284423828,
      "learning_rate": 3.241943593210529e-05,
      "loss": 0.1228,
      "num_input_tokens_seen": 11571176,
      "step": 17695
    },
    {
      "epoch": 9.276729559748428,
      "grad_norm": 2.2416727542877197,
      "learning_rate": 3.2408516024074916e-05,
      "loss": 0.1603,
      "num_input_tokens_seen": 11574216,
      "step": 17700
    },
    {
      "epoch": 9.279350104821804,
      "grad_norm": 2.0629308223724365,
      "learning_rate": 3.2397594566267505e-05,
      "loss": 0.1012,
      "num_input_tokens_seen": 11577608,
      "step": 17705
    },
    {
      "epoch": 9.281970649895179,
      "grad_norm": 2.069634437561035,
      "learning_rate": 3.23866715609677e-05,
      "loss": 0.1428,
      "num_input_tokens_seen": 11581000,
      "step": 17710
    },
    {
      "epoch": 9.284591194968554,
      "grad_norm": 1.516592025756836,
      "learning_rate": 3.237574701046049e-05,
      "loss": 0.152,
      "num_input_tokens_seen": 11584840,
      "step": 17715
    },
    {
      "epoch": 9.28721174004193,
      "grad_norm": 2.777665376663208,
      "learning_rate": 3.236482091703115e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 11587912,
      "step": 17720
    },
    {
      "epoch": 9.289832285115304,
      "grad_norm": 2.8675687313079834,
      "learning_rate": 3.235389328296528e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 11591752,
      "step": 17725
    },
    {
      "epoch": 9.29245283018868,
      "grad_norm": 2.6604907512664795,
      "learning_rate": 3.2342964110548845e-05,
      "loss": 0.1821,
      "num_input_tokens_seen": 11594248,
      "step": 17730
    },
    {
      "epoch": 9.295073375262055,
      "grad_norm": 3.396599054336548,
      "learning_rate": 3.2332033402068093e-05,
      "loss": 0.1308,
      "num_input_tokens_seen": 11597448,
      "step": 17735
    },
    {
      "epoch": 9.29769392033543,
      "grad_norm": 1.4642943143844604,
      "learning_rate": 3.2321101159809605e-05,
      "loss": 0.2033,
      "num_input_tokens_seen": 11600360,
      "step": 17740
    },
    {
      "epoch": 9.300314465408805,
      "grad_norm": 2.9682931900024414,
      "learning_rate": 3.231016738606027e-05,
      "loss": 0.1808,
      "num_input_tokens_seen": 11603528,
      "step": 17745
    },
    {
      "epoch": 9.30293501048218,
      "grad_norm": 2.5125303268432617,
      "learning_rate": 3.2299232083107313e-05,
      "loss": 0.1657,
      "num_input_tokens_seen": 11606600,
      "step": 17750
    },
    {
      "epoch": 9.305555555555555,
      "grad_norm": 1.7588684558868408,
      "learning_rate": 3.228829525323829e-05,
      "loss": 0.1195,
      "num_input_tokens_seen": 11609064,
      "step": 17755
    },
    {
      "epoch": 9.30817610062893,
      "grad_norm": 1.4553123712539673,
      "learning_rate": 3.227735689874104e-05,
      "loss": 0.126,
      "num_input_tokens_seen": 11612328,
      "step": 17760
    },
    {
      "epoch": 9.310796645702306,
      "grad_norm": 2.49920916557312,
      "learning_rate": 3.226641702190375e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 11615912,
      "step": 17765
    },
    {
      "epoch": 9.31341719077568,
      "grad_norm": 1.4535348415374756,
      "learning_rate": 3.225547562501493e-05,
      "loss": 0.1793,
      "num_input_tokens_seen": 11618888,
      "step": 17770
    },
    {
      "epoch": 9.316037735849056,
      "grad_norm": 2.8179666996002197,
      "learning_rate": 3.2244532710363375e-05,
      "loss": 0.2876,
      "num_input_tokens_seen": 11621544,
      "step": 17775
    },
    {
      "epoch": 9.318658280922431,
      "grad_norm": 1.937740445137024,
      "learning_rate": 3.2233588280238224e-05,
      "loss": 0.1222,
      "num_input_tokens_seen": 11624040,
      "step": 17780
    },
    {
      "epoch": 9.321278825995806,
      "grad_norm": 2.4351654052734375,
      "learning_rate": 3.2222642336928944e-05,
      "loss": 0.1346,
      "num_input_tokens_seen": 11626728,
      "step": 17785
    },
    {
      "epoch": 9.323899371069182,
      "grad_norm": 5.756466865539551,
      "learning_rate": 3.2211694882725275e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 11629832,
      "step": 17790
    },
    {
      "epoch": 9.326519916142558,
      "grad_norm": 1.7997008562088013,
      "learning_rate": 3.220074591991731e-05,
      "loss": 0.0953,
      "num_input_tokens_seen": 11633320,
      "step": 17795
    },
    {
      "epoch": 9.329140461215934,
      "grad_norm": 2.3008224964141846,
      "learning_rate": 3.218979545079546e-05,
      "loss": 0.1293,
      "num_input_tokens_seen": 11636104,
      "step": 17800
    },
    {
      "epoch": 9.331761006289309,
      "grad_norm": 2.3399860858917236,
      "learning_rate": 3.217884347765042e-05,
      "loss": 0.1713,
      "num_input_tokens_seen": 11639752,
      "step": 17805
    },
    {
      "epoch": 9.334381551362684,
      "grad_norm": 1.3285224437713623,
      "learning_rate": 3.216789000277323e-05,
      "loss": 0.1563,
      "num_input_tokens_seen": 11642664,
      "step": 17810
    },
    {
      "epoch": 9.33700209643606,
      "grad_norm": 2.8557095527648926,
      "learning_rate": 3.2156935028455224e-05,
      "loss": 0.1475,
      "num_input_tokens_seen": 11646024,
      "step": 17815
    },
    {
      "epoch": 9.339622641509434,
      "grad_norm": 1.821001410484314,
      "learning_rate": 3.2145978556988056e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 11649192,
      "step": 17820
    },
    {
      "epoch": 9.34224318658281,
      "grad_norm": 2.535188674926758,
      "learning_rate": 3.213502059066371e-05,
      "loss": 0.1875,
      "num_input_tokens_seen": 11653192,
      "step": 17825
    },
    {
      "epoch": 9.344863731656185,
      "grad_norm": 3.956536293029785,
      "learning_rate": 3.212406113177445e-05,
      "loss": 0.1266,
      "num_input_tokens_seen": 11657224,
      "step": 17830
    },
    {
      "epoch": 9.34748427672956,
      "grad_norm": 7.599135398864746,
      "learning_rate": 3.211310018261287e-05,
      "loss": 0.2119,
      "num_input_tokens_seen": 11660328,
      "step": 17835
    },
    {
      "epoch": 9.350104821802935,
      "grad_norm": 3.1743743419647217,
      "learning_rate": 3.21021377454719e-05,
      "loss": 0.1513,
      "num_input_tokens_seen": 11663400,
      "step": 17840
    },
    {
      "epoch": 9.35272536687631,
      "grad_norm": 3.0937955379486084,
      "learning_rate": 3.209117382264473e-05,
      "loss": 0.1675,
      "num_input_tokens_seen": 11665992,
      "step": 17845
    },
    {
      "epoch": 9.355345911949685,
      "grad_norm": 1.415640950202942,
      "learning_rate": 3.208020841642489e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 11669000,
      "step": 17850
    },
    {
      "epoch": 9.35796645702306,
      "grad_norm": 1.4867298603057861,
      "learning_rate": 3.206924152910623e-05,
      "loss": 0.1178,
      "num_input_tokens_seen": 11671976,
      "step": 17855
    },
    {
      "epoch": 9.360587002096436,
      "grad_norm": 1.6895273923873901,
      "learning_rate": 3.205827316298289e-05,
      "loss": 0.1357,
      "num_input_tokens_seen": 11674984,
      "step": 17860
    },
    {
      "epoch": 9.36320754716981,
      "grad_norm": 2.1543784141540527,
      "learning_rate": 3.204730332034934e-05,
      "loss": 0.1368,
      "num_input_tokens_seen": 11677864,
      "step": 17865
    },
    {
      "epoch": 9.365828092243186,
      "grad_norm": 1.2016891241073608,
      "learning_rate": 3.2036332003500324e-05,
      "loss": 0.1423,
      "num_input_tokens_seen": 11681128,
      "step": 17870
    },
    {
      "epoch": 9.368448637316561,
      "grad_norm": 3.175535202026367,
      "learning_rate": 3.2025359214730924e-05,
      "loss": 0.1164,
      "num_input_tokens_seen": 11683880,
      "step": 17875
    },
    {
      "epoch": 9.371069182389936,
      "grad_norm": 2.3337509632110596,
      "learning_rate": 3.201438495633654e-05,
      "loss": 0.11,
      "num_input_tokens_seen": 11687496,
      "step": 17880
    },
    {
      "epoch": 9.373689727463312,
      "grad_norm": 5.457139492034912,
      "learning_rate": 3.200340923061284e-05,
      "loss": 0.1738,
      "num_input_tokens_seen": 11690696,
      "step": 17885
    },
    {
      "epoch": 9.376310272536688,
      "grad_norm": 2.5966482162475586,
      "learning_rate": 3.199243203985583e-05,
      "loss": 0.1655,
      "num_input_tokens_seen": 11693544,
      "step": 17890
    },
    {
      "epoch": 9.378930817610064,
      "grad_norm": 4.445423126220703,
      "learning_rate": 3.1981453386361804e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 11696840,
      "step": 17895
    },
    {
      "epoch": 9.381551362683439,
      "grad_norm": 1.872765064239502,
      "learning_rate": 3.1970473272427386e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 11700264,
      "step": 17900
    },
    {
      "epoch": 9.384171907756814,
      "grad_norm": 2.053901433944702,
      "learning_rate": 3.195949170034948e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 11703496,
      "step": 17905
    },
    {
      "epoch": 9.38679245283019,
      "grad_norm": 1.4511748552322388,
      "learning_rate": 3.19485086724253e-05,
      "loss": 0.1541,
      "num_input_tokens_seen": 11706664,
      "step": 17910
    },
    {
      "epoch": 9.389412997903564,
      "grad_norm": 6.750287055969238,
      "learning_rate": 3.193752419095239e-05,
      "loss": 0.1396,
      "num_input_tokens_seen": 11709480,
      "step": 17915
    },
    {
      "epoch": 9.39203354297694,
      "grad_norm": 2.117400884628296,
      "learning_rate": 3.192653825822857e-05,
      "loss": 0.1068,
      "num_input_tokens_seen": 11712712,
      "step": 17920
    },
    {
      "epoch": 9.394654088050315,
      "grad_norm": 2.0844497680664062,
      "learning_rate": 3.1915550876551964e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 11715464,
      "step": 17925
    },
    {
      "epoch": 9.39727463312369,
      "grad_norm": 2.0771729946136475,
      "learning_rate": 3.190456204822101e-05,
      "loss": 0.1186,
      "num_input_tokens_seen": 11718120,
      "step": 17930
    },
    {
      "epoch": 9.399895178197065,
      "grad_norm": 1.3135913610458374,
      "learning_rate": 3.1893571775534445e-05,
      "loss": 0.1492,
      "num_input_tokens_seen": 11721512,
      "step": 17935
    },
    {
      "epoch": 9.40251572327044,
      "grad_norm": 1.6364752054214478,
      "learning_rate": 3.1882580060791315e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 11724008,
      "step": 17940
    },
    {
      "epoch": 9.405136268343815,
      "grad_norm": 1.7194823026657104,
      "learning_rate": 3.187158690629095e-05,
      "loss": 0.1258,
      "num_input_tokens_seen": 11726696,
      "step": 17945
    },
    {
      "epoch": 9.40775681341719,
      "grad_norm": 1.5406972169876099,
      "learning_rate": 3.1860592314333e-05,
      "loss": 0.1119,
      "num_input_tokens_seen": 11730408,
      "step": 17950
    },
    {
      "epoch": 9.410377358490566,
      "grad_norm": 2.11665415763855,
      "learning_rate": 3.184959628721742e-05,
      "loss": 0.1512,
      "num_input_tokens_seen": 11733320,
      "step": 17955
    },
    {
      "epoch": 9.41299790356394,
      "grad_norm": 2.2650272846221924,
      "learning_rate": 3.183859882724443e-05,
      "loss": 0.1583,
      "num_input_tokens_seen": 11738472,
      "step": 17960
    },
    {
      "epoch": 9.415618448637316,
      "grad_norm": 2.5178825855255127,
      "learning_rate": 3.1827599936714584e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 11741032,
      "step": 17965
    },
    {
      "epoch": 9.418238993710691,
      "grad_norm": 4.529767036437988,
      "learning_rate": 3.1816599617928736e-05,
      "loss": 0.1048,
      "num_input_tokens_seen": 11744328,
      "step": 17970
    },
    {
      "epoch": 9.420859538784066,
      "grad_norm": 1.7914179563522339,
      "learning_rate": 3.1805597873188e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 11747976,
      "step": 17975
    },
    {
      "epoch": 9.423480083857442,
      "grad_norm": 4.750181674957275,
      "learning_rate": 3.179459470479384e-05,
      "loss": 0.2139,
      "num_input_tokens_seen": 11751144,
      "step": 17980
    },
    {
      "epoch": 9.426100628930818,
      "grad_norm": 1.9450030326843262,
      "learning_rate": 3.1783590115047983e-05,
      "loss": 0.1482,
      "num_input_tokens_seen": 11754312,
      "step": 17985
    },
    {
      "epoch": 9.428721174004194,
      "grad_norm": 1.3842140436172485,
      "learning_rate": 3.177258410625246e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 11757896,
      "step": 17990
    },
    {
      "epoch": 9.431341719077569,
      "grad_norm": 2.2226366996765137,
      "learning_rate": 3.176157668070962e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 11760840,
      "step": 17995
    },
    {
      "epoch": 9.433962264150944,
      "grad_norm": 2.14851975440979,
      "learning_rate": 3.175056784072207e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 11764136,
      "step": 18000
    },
    {
      "epoch": 9.43658280922432,
      "grad_norm": 0.528634250164032,
      "learning_rate": 3.173955758859274e-05,
      "loss": 0.1614,
      "num_input_tokens_seen": 11766696,
      "step": 18005
    },
    {
      "epoch": 9.439203354297694,
      "grad_norm": 2.6327743530273438,
      "learning_rate": 3.172854592662486e-05,
      "loss": 0.1517,
      "num_input_tokens_seen": 11769160,
      "step": 18010
    },
    {
      "epoch": 9.44182389937107,
      "grad_norm": 1.9767807722091675,
      "learning_rate": 3.1717532857121926e-05,
      "loss": 0.1614,
      "num_input_tokens_seen": 11772360,
      "step": 18015
    },
    {
      "epoch": 9.444444444444445,
      "grad_norm": 2.4775943756103516,
      "learning_rate": 3.170651838238776e-05,
      "loss": 0.166,
      "num_input_tokens_seen": 11774984,
      "step": 18020
    },
    {
      "epoch": 9.44706498951782,
      "grad_norm": 1.912765383720398,
      "learning_rate": 3.169550250472646e-05,
      "loss": 0.1204,
      "num_input_tokens_seen": 11777928,
      "step": 18025
    },
    {
      "epoch": 9.449685534591195,
      "grad_norm": 2.2946908473968506,
      "learning_rate": 3.168448522644243e-05,
      "loss": 0.1006,
      "num_input_tokens_seen": 11781096,
      "step": 18030
    },
    {
      "epoch": 9.45230607966457,
      "grad_norm": 3.6780059337615967,
      "learning_rate": 3.167346654984034e-05,
      "loss": 0.1166,
      "num_input_tokens_seen": 11784104,
      "step": 18035
    },
    {
      "epoch": 9.454926624737945,
      "grad_norm": 2.383603572845459,
      "learning_rate": 3.166244647722519e-05,
      "loss": 0.1339,
      "num_input_tokens_seen": 11787304,
      "step": 18040
    },
    {
      "epoch": 9.45754716981132,
      "grad_norm": 4.49743127822876,
      "learning_rate": 3.1651425010902244e-05,
      "loss": 0.0923,
      "num_input_tokens_seen": 11790024,
      "step": 18045
    },
    {
      "epoch": 9.460167714884696,
      "grad_norm": 2.186427593231201,
      "learning_rate": 3.164040215317706e-05,
      "loss": 0.0963,
      "num_input_tokens_seen": 11792840,
      "step": 18050
    },
    {
      "epoch": 9.46278825995807,
      "grad_norm": 2.1602652072906494,
      "learning_rate": 3.162937790635552e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 11795720,
      "step": 18055
    },
    {
      "epoch": 9.465408805031446,
      "grad_norm": 2.4199140071868896,
      "learning_rate": 3.1618352272743746e-05,
      "loss": 0.1026,
      "num_input_tokens_seen": 11798472,
      "step": 18060
    },
    {
      "epoch": 9.468029350104821,
      "grad_norm": 2.585249185562134,
      "learning_rate": 3.160732525464818e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 11801800,
      "step": 18065
    },
    {
      "epoch": 9.470649895178196,
      "grad_norm": 4.474305629730225,
      "learning_rate": 3.1596296854375554e-05,
      "loss": 0.1305,
      "num_input_tokens_seen": 11805160,
      "step": 18070
    },
    {
      "epoch": 9.473270440251572,
      "grad_norm": 2.0725347995758057,
      "learning_rate": 3.1585267074232884e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 11811816,
      "step": 18075
    },
    {
      "epoch": 9.475890985324948,
      "grad_norm": 0.9825284481048584,
      "learning_rate": 3.157423591652746e-05,
      "loss": 0.1828,
      "num_input_tokens_seen": 11816040,
      "step": 18080
    },
    {
      "epoch": 9.478511530398324,
      "grad_norm": 2.233828544616699,
      "learning_rate": 3.1563203383566894e-05,
      "loss": 0.1844,
      "num_input_tokens_seen": 11818920,
      "step": 18085
    },
    {
      "epoch": 9.481132075471699,
      "grad_norm": 2.238248586654663,
      "learning_rate": 3.155216947765906e-05,
      "loss": 0.1604,
      "num_input_tokens_seen": 11821736,
      "step": 18090
    },
    {
      "epoch": 9.483752620545074,
      "grad_norm": 5.282876014709473,
      "learning_rate": 3.154113420111211e-05,
      "loss": 0.1485,
      "num_input_tokens_seen": 11825256,
      "step": 18095
    },
    {
      "epoch": 9.48637316561845,
      "grad_norm": 3.2656493186950684,
      "learning_rate": 3.1530097556234526e-05,
      "loss": 0.1719,
      "num_input_tokens_seen": 11829416,
      "step": 18100
    },
    {
      "epoch": 9.488993710691824,
      "grad_norm": 1.738970398902893,
      "learning_rate": 3.151905954533502e-05,
      "loss": 0.2023,
      "num_input_tokens_seen": 11832200,
      "step": 18105
    },
    {
      "epoch": 9.4916142557652,
      "grad_norm": 2.8072166442871094,
      "learning_rate": 3.150802017072263e-05,
      "loss": 0.1455,
      "num_input_tokens_seen": 11835592,
      "step": 18110
    },
    {
      "epoch": 9.494234800838575,
      "grad_norm": 1.5875645875930786,
      "learning_rate": 3.149697943470668e-05,
      "loss": 0.1407,
      "num_input_tokens_seen": 11839400,
      "step": 18115
    },
    {
      "epoch": 9.49685534591195,
      "grad_norm": 4.230952262878418,
      "learning_rate": 3.1485937339596736e-05,
      "loss": 0.1714,
      "num_input_tokens_seen": 11842216,
      "step": 18120
    },
    {
      "epoch": 9.499475890985325,
      "grad_norm": 1.8907196521759033,
      "learning_rate": 3.147489388770271e-05,
      "loss": 0.0891,
      "num_input_tokens_seen": 11844936,
      "step": 18125
    },
    {
      "epoch": 9.5020964360587,
      "grad_norm": 2.824403762817383,
      "learning_rate": 3.146384908133475e-05,
      "loss": 0.1586,
      "num_input_tokens_seen": 11847912,
      "step": 18130
    },
    {
      "epoch": 9.504716981132075,
      "grad_norm": 1.4081388711929321,
      "learning_rate": 3.14528029228033e-05,
      "loss": 0.0933,
      "num_input_tokens_seen": 11853192,
      "step": 18135
    },
    {
      "epoch": 9.50733752620545,
      "grad_norm": 5.150257110595703,
      "learning_rate": 3.144175541441911e-05,
      "loss": 0.1408,
      "num_input_tokens_seen": 11856808,
      "step": 18140
    },
    {
      "epoch": 9.509958071278826,
      "grad_norm": 3.6133670806884766,
      "learning_rate": 3.1430706558493166e-05,
      "loss": 0.3099,
      "num_input_tokens_seen": 11860360,
      "step": 18145
    },
    {
      "epoch": 9.5125786163522,
      "grad_norm": 2.995764970779419,
      "learning_rate": 3.141965635733677e-05,
      "loss": 0.0871,
      "num_input_tokens_seen": 11863048,
      "step": 18150
    },
    {
      "epoch": 9.515199161425576,
      "grad_norm": 2.32356333732605,
      "learning_rate": 3.140860481326152e-05,
      "loss": 0.1372,
      "num_input_tokens_seen": 11867016,
      "step": 18155
    },
    {
      "epoch": 9.517819706498951,
      "grad_norm": 2.4020235538482666,
      "learning_rate": 3.139755192857925e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 11870728,
      "step": 18160
    },
    {
      "epoch": 9.520440251572326,
      "grad_norm": 1.2615834474563599,
      "learning_rate": 3.1386497705602094e-05,
      "loss": 0.1329,
      "num_input_tokens_seen": 11873928,
      "step": 18165
    },
    {
      "epoch": 9.523060796645701,
      "grad_norm": 4.121642112731934,
      "learning_rate": 3.137544214664249e-05,
      "loss": 0.1103,
      "num_input_tokens_seen": 11876968,
      "step": 18170
    },
    {
      "epoch": 9.525681341719078,
      "grad_norm": 1.455438256263733,
      "learning_rate": 3.136438525401312e-05,
      "loss": 0.1264,
      "num_input_tokens_seen": 11880648,
      "step": 18175
    },
    {
      "epoch": 9.528301886792454,
      "grad_norm": 3.1063430309295654,
      "learning_rate": 3.135332703002696e-05,
      "loss": 0.1661,
      "num_input_tokens_seen": 11883560,
      "step": 18180
    },
    {
      "epoch": 9.530922431865829,
      "grad_norm": 2.5786588191986084,
      "learning_rate": 3.1342267476997276e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 11888296,
      "step": 18185
    },
    {
      "epoch": 9.533542976939204,
      "grad_norm": 2.253990650177002,
      "learning_rate": 3.1331206597237584e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 11892872,
      "step": 18190
    },
    {
      "epoch": 9.536163522012579,
      "grad_norm": 2.0128870010375977,
      "learning_rate": 3.1320144393061715e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 11895944,
      "step": 18195
    },
    {
      "epoch": 9.538784067085954,
      "grad_norm": 1.0637145042419434,
      "learning_rate": 3.130908086678373e-05,
      "loss": 0.0899,
      "num_input_tokens_seen": 11898696,
      "step": 18200
    },
    {
      "epoch": 9.54140461215933,
      "grad_norm": 0.6679860949516296,
      "learning_rate": 3.1298016020718014e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 11904744,
      "step": 18205
    },
    {
      "epoch": 9.544025157232705,
      "grad_norm": 3.1019833087921143,
      "learning_rate": 3.128694985717919e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 11907656,
      "step": 18210
    },
    {
      "epoch": 9.54664570230608,
      "grad_norm": 1.853473424911499,
      "learning_rate": 3.1275882378482185e-05,
      "loss": 0.1901,
      "num_input_tokens_seen": 11910504,
      "step": 18215
    },
    {
      "epoch": 9.549266247379455,
      "grad_norm": 4.573065280914307,
      "learning_rate": 3.1264813586942185e-05,
      "loss": 0.1127,
      "num_input_tokens_seen": 11913192,
      "step": 18220
    },
    {
      "epoch": 9.55188679245283,
      "grad_norm": 2.3029873371124268,
      "learning_rate": 3.125374348487466e-05,
      "loss": 0.1874,
      "num_input_tokens_seen": 11916680,
      "step": 18225
    },
    {
      "epoch": 9.554507337526205,
      "grad_norm": 4.228594779968262,
      "learning_rate": 3.1242672074595334e-05,
      "loss": 0.1071,
      "num_input_tokens_seen": 11919656,
      "step": 18230
    },
    {
      "epoch": 9.55712788259958,
      "grad_norm": 2.116464376449585,
      "learning_rate": 3.123159935842024e-05,
      "loss": 0.1394,
      "num_input_tokens_seen": 11922632,
      "step": 18235
    },
    {
      "epoch": 9.559748427672956,
      "grad_norm": 1.7544292211532593,
      "learning_rate": 3.122052533866565e-05,
      "loss": 0.1373,
      "num_input_tokens_seen": 11926280,
      "step": 18240
    },
    {
      "epoch": 9.56236897274633,
      "grad_norm": 0.7470082640647888,
      "learning_rate": 3.120945001764813e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 11929768,
      "step": 18245
    },
    {
      "epoch": 9.564989517819706,
      "grad_norm": 2.9041500091552734,
      "learning_rate": 3.11983733976845e-05,
      "loss": 0.1483,
      "num_input_tokens_seen": 11933480,
      "step": 18250
    },
    {
      "epoch": 9.567610062893081,
      "grad_norm": 7.824368953704834,
      "learning_rate": 3.1187295481091876e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 11936200,
      "step": 18255
    },
    {
      "epoch": 9.570230607966456,
      "grad_norm": 3.3277060985565186,
      "learning_rate": 3.1176216270187625e-05,
      "loss": 0.1652,
      "num_input_tokens_seen": 11938472,
      "step": 18260
    },
    {
      "epoch": 9.572851153039831,
      "grad_norm": 2.3754444122314453,
      "learning_rate": 3.116513576728939e-05,
      "loss": 0.1748,
      "num_input_tokens_seen": 11942280,
      "step": 18265
    },
    {
      "epoch": 9.575471698113208,
      "grad_norm": 3.2439982891082764,
      "learning_rate": 3.1154053974715085e-05,
      "loss": 0.1138,
      "num_input_tokens_seen": 11945800,
      "step": 18270
    },
    {
      "epoch": 9.578092243186584,
      "grad_norm": 0.7429378628730774,
      "learning_rate": 3.1142970894782906e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 11948200,
      "step": 18275
    },
    {
      "epoch": 9.580712788259959,
      "grad_norm": 3.857999563217163,
      "learning_rate": 3.1131886529811296e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 11951656,
      "step": 18280
    },
    {
      "epoch": 9.583333333333334,
      "grad_norm": 1.3860764503479004,
      "learning_rate": 3.1120800882118985e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 11955208,
      "step": 18285
    },
    {
      "epoch": 9.585953878406709,
      "grad_norm": 2.1036691665649414,
      "learning_rate": 3.1109713954024945e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 11957768,
      "step": 18290
    },
    {
      "epoch": 9.588574423480084,
      "grad_norm": 2.211710214614868,
      "learning_rate": 3.109862574784846e-05,
      "loss": 0.1312,
      "num_input_tokens_seen": 11960552,
      "step": 18295
    },
    {
      "epoch": 9.59119496855346,
      "grad_norm": 3.3217813968658447,
      "learning_rate": 3.108753626590905e-05,
      "loss": 0.1704,
      "num_input_tokens_seen": 11964552,
      "step": 18300
    },
    {
      "epoch": 9.593815513626835,
      "grad_norm": 2.7707254886627197,
      "learning_rate": 3.1076445510526486e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 11966824,
      "step": 18305
    },
    {
      "epoch": 9.59643605870021,
      "grad_norm": 3.325366497039795,
      "learning_rate": 3.106535348402086e-05,
      "loss": 0.121,
      "num_input_tokens_seen": 11969832,
      "step": 18310
    },
    {
      "epoch": 9.599056603773585,
      "grad_norm": 2.5148983001708984,
      "learning_rate": 3.105426018871247e-05,
      "loss": 0.1043,
      "num_input_tokens_seen": 11972424,
      "step": 18315
    },
    {
      "epoch": 9.60167714884696,
      "grad_norm": 3.679203510284424,
      "learning_rate": 3.104316562692193e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 11976552,
      "step": 18320
    },
    {
      "epoch": 9.604297693920335,
      "grad_norm": 1.5314819812774658,
      "learning_rate": 3.103206980097009e-05,
      "loss": 0.1383,
      "num_input_tokens_seen": 11979720,
      "step": 18325
    },
    {
      "epoch": 9.60691823899371,
      "grad_norm": 2.7887072563171387,
      "learning_rate": 3.102097271317805e-05,
      "loss": 0.1691,
      "num_input_tokens_seen": 11982536,
      "step": 18330
    },
    {
      "epoch": 9.609538784067086,
      "grad_norm": 1.7140896320343018,
      "learning_rate": 3.100987436586721e-05,
      "loss": 0.1006,
      "num_input_tokens_seen": 11985384,
      "step": 18335
    },
    {
      "epoch": 9.61215932914046,
      "grad_norm": 12.912583351135254,
      "learning_rate": 3.099877476135923e-05,
      "loss": 0.1247,
      "num_input_tokens_seen": 11988712,
      "step": 18340
    },
    {
      "epoch": 9.614779874213836,
      "grad_norm": 2.3603756427764893,
      "learning_rate": 3.0987673901976e-05,
      "loss": 0.1384,
      "num_input_tokens_seen": 11993352,
      "step": 18345
    },
    {
      "epoch": 9.617400419287211,
      "grad_norm": 2.471151351928711,
      "learning_rate": 3.0976571790039697e-05,
      "loss": 0.1325,
      "num_input_tokens_seen": 11995848,
      "step": 18350
    },
    {
      "epoch": 9.620020964360586,
      "grad_norm": 1.6779924631118774,
      "learning_rate": 3.096546842787277e-05,
      "loss": 0.1783,
      "num_input_tokens_seen": 11999336,
      "step": 18355
    },
    {
      "epoch": 9.622641509433961,
      "grad_norm": 1.4760494232177734,
      "learning_rate": 3.095436381779789e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 12003080,
      "step": 18360
    },
    {
      "epoch": 9.625262054507338,
      "grad_norm": 0.9329490661621094,
      "learning_rate": 3.0943257962138045e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 12007240,
      "step": 18365
    },
    {
      "epoch": 9.627882599580714,
      "grad_norm": 1.4276460409164429,
      "learning_rate": 3.093215086321644e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 12009928,
      "step": 18370
    },
    {
      "epoch": 9.630503144654089,
      "grad_norm": 2.572695016860962,
      "learning_rate": 3.092104252335654e-05,
      "loss": 0.1292,
      "num_input_tokens_seen": 12013384,
      "step": 18375
    },
    {
      "epoch": 9.633123689727464,
      "grad_norm": 3.9356555938720703,
      "learning_rate": 3.09099329448821e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 12016520,
      "step": 18380
    },
    {
      "epoch": 9.635744234800839,
      "grad_norm": 2.20704984664917,
      "learning_rate": 3.089882213011711e-05,
      "loss": 0.1175,
      "num_input_tokens_seen": 12020040,
      "step": 18385
    },
    {
      "epoch": 9.638364779874214,
      "grad_norm": 2.3202648162841797,
      "learning_rate": 3.088771008138583e-05,
      "loss": 0.1309,
      "num_input_tokens_seen": 12023464,
      "step": 18390
    },
    {
      "epoch": 9.64098532494759,
      "grad_norm": 2.4974210262298584,
      "learning_rate": 3.087659680101278e-05,
      "loss": 0.085,
      "num_input_tokens_seen": 12026664,
      "step": 18395
    },
    {
      "epoch": 9.643605870020965,
      "grad_norm": 1.9821707010269165,
      "learning_rate": 3.086548229132271e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 12030440,
      "step": 18400
    },
    {
      "epoch": 9.64622641509434,
      "grad_norm": 1.592589259147644,
      "learning_rate": 3.085436655464067e-05,
      "loss": 0.1376,
      "num_input_tokens_seen": 12032872,
      "step": 18405
    },
    {
      "epoch": 9.648846960167715,
      "grad_norm": 6.146122455596924,
      "learning_rate": 3.084324959329193e-05,
      "loss": 0.1019,
      "num_input_tokens_seen": 12037864,
      "step": 18410
    },
    {
      "epoch": 9.65146750524109,
      "grad_norm": 3.2387473583221436,
      "learning_rate": 3.083213140960204e-05,
      "loss": 0.1017,
      "num_input_tokens_seen": 12040904,
      "step": 18415
    },
    {
      "epoch": 9.654088050314465,
      "grad_norm": 1.196893334388733,
      "learning_rate": 3.08210120058968e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 12044296,
      "step": 18420
    },
    {
      "epoch": 9.65670859538784,
      "grad_norm": 1.944940447807312,
      "learning_rate": 3.080989138450225e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 12047208,
      "step": 18425
    },
    {
      "epoch": 9.659329140461216,
      "grad_norm": 1.9742192029953003,
      "learning_rate": 3.0798769547744713e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 12050696,
      "step": 18430
    },
    {
      "epoch": 9.66194968553459,
      "grad_norm": 1.6697101593017578,
      "learning_rate": 3.0787646497950736e-05,
      "loss": 0.1526,
      "num_input_tokens_seen": 12054536,
      "step": 18435
    },
    {
      "epoch": 9.664570230607966,
      "grad_norm": 3.6499717235565186,
      "learning_rate": 3.0776522237447146e-05,
      "loss": 0.141,
      "num_input_tokens_seen": 12056776,
      "step": 18440
    },
    {
      "epoch": 9.667190775681341,
      "grad_norm": 2.2932288646698,
      "learning_rate": 3.076539676856101e-05,
      "loss": 0.1248,
      "num_input_tokens_seen": 12059720,
      "step": 18445
    },
    {
      "epoch": 9.669811320754716,
      "grad_norm": 2.448474407196045,
      "learning_rate": 3.0754270093619634e-05,
      "loss": 0.1535,
      "num_input_tokens_seen": 12063304,
      "step": 18450
    },
    {
      "epoch": 9.672431865828091,
      "grad_norm": 3.7703261375427246,
      "learning_rate": 3.074314221495061e-05,
      "loss": 0.146,
      "num_input_tokens_seen": 12067592,
      "step": 18455
    },
    {
      "epoch": 9.675052410901468,
      "grad_norm": 1.5054594278335571,
      "learning_rate": 3.0732013134881756e-05,
      "loss": 0.1312,
      "num_input_tokens_seen": 12070856,
      "step": 18460
    },
    {
      "epoch": 9.677672955974844,
      "grad_norm": 1.8001750707626343,
      "learning_rate": 3.072088285574114e-05,
      "loss": 0.1559,
      "num_input_tokens_seen": 12074280,
      "step": 18465
    },
    {
      "epoch": 9.680293501048219,
      "grad_norm": 2.4805142879486084,
      "learning_rate": 3.070975137985711e-05,
      "loss": 0.1173,
      "num_input_tokens_seen": 12077288,
      "step": 18470
    },
    {
      "epoch": 9.682914046121594,
      "grad_norm": 1.818504810333252,
      "learning_rate": 3.0698618709558226e-05,
      "loss": 0.11,
      "num_input_tokens_seen": 12082056,
      "step": 18475
    },
    {
      "epoch": 9.685534591194969,
      "grad_norm": 3.6359264850616455,
      "learning_rate": 3.068748484717332e-05,
      "loss": 0.1602,
      "num_input_tokens_seen": 12085128,
      "step": 18480
    },
    {
      "epoch": 9.688155136268344,
      "grad_norm": 2.313922643661499,
      "learning_rate": 3.067634979503147e-05,
      "loss": 0.1549,
      "num_input_tokens_seen": 12089032,
      "step": 18485
    },
    {
      "epoch": 9.69077568134172,
      "grad_norm": 2.4140584468841553,
      "learning_rate": 3.0665213555462e-05,
      "loss": 0.1253,
      "num_input_tokens_seen": 12091496,
      "step": 18490
    },
    {
      "epoch": 9.693396226415095,
      "grad_norm": 5.79524564743042,
      "learning_rate": 3.0654076130794494e-05,
      "loss": 0.1857,
      "num_input_tokens_seen": 12095208,
      "step": 18495
    },
    {
      "epoch": 9.69601677148847,
      "grad_norm": 2.1162729263305664,
      "learning_rate": 3.064293752335876e-05,
      "loss": 0.1962,
      "num_input_tokens_seen": 12099176,
      "step": 18500
    },
    {
      "epoch": 9.698637316561845,
      "grad_norm": 2.1825833320617676,
      "learning_rate": 3.063179773548487e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 12102408,
      "step": 18505
    },
    {
      "epoch": 9.70125786163522,
      "grad_norm": 2.724219799041748,
      "learning_rate": 3.0620656769503164e-05,
      "loss": 0.1688,
      "num_input_tokens_seen": 12105704,
      "step": 18510
    },
    {
      "epoch": 9.703878406708595,
      "grad_norm": 1.799387812614441,
      "learning_rate": 3.060951462774417e-05,
      "loss": 0.2508,
      "num_input_tokens_seen": 12108520,
      "step": 18515
    },
    {
      "epoch": 9.70649895178197,
      "grad_norm": 2.1366348266601562,
      "learning_rate": 3.0598371312538714e-05,
      "loss": 0.1423,
      "num_input_tokens_seen": 12111912,
      "step": 18520
    },
    {
      "epoch": 9.709119496855346,
      "grad_norm": 4.76451301574707,
      "learning_rate": 3.058722682621785e-05,
      "loss": 0.1377,
      "num_input_tokens_seen": 12116040,
      "step": 18525
    },
    {
      "epoch": 9.71174004192872,
      "grad_norm": 1.686957836151123,
      "learning_rate": 3.0576081171112875e-05,
      "loss": 0.0957,
      "num_input_tokens_seen": 12118632,
      "step": 18530
    },
    {
      "epoch": 9.714360587002096,
      "grad_norm": 1.0431219339370728,
      "learning_rate": 3.056493434955534e-05,
      "loss": 0.0863,
      "num_input_tokens_seen": 12121192,
      "step": 18535
    },
    {
      "epoch": 9.716981132075471,
      "grad_norm": 2.9822633266448975,
      "learning_rate": 3.0553786363877027e-05,
      "loss": 0.1092,
      "num_input_tokens_seen": 12123976,
      "step": 18540
    },
    {
      "epoch": 9.719601677148846,
      "grad_norm": 1.6214367151260376,
      "learning_rate": 3.054263721640996e-05,
      "loss": 0.2022,
      "num_input_tokens_seen": 12127816,
      "step": 18545
    },
    {
      "epoch": 9.722222222222221,
      "grad_norm": 1.9078859090805054,
      "learning_rate": 3.053148690948642e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 12131848,
      "step": 18550
    },
    {
      "epoch": 9.724842767295598,
      "grad_norm": 3.9665119647979736,
      "learning_rate": 3.0520335445438935e-05,
      "loss": 0.1839,
      "num_input_tokens_seen": 12135432,
      "step": 18555
    },
    {
      "epoch": 9.727463312368974,
      "grad_norm": 2.9246106147766113,
      "learning_rate": 3.0509182826600246e-05,
      "loss": 0.1794,
      "num_input_tokens_seen": 12138376,
      "step": 18560
    },
    {
      "epoch": 9.730083857442349,
      "grad_norm": 2.15832257270813,
      "learning_rate": 3.049802905530337e-05,
      "loss": 0.1606,
      "num_input_tokens_seen": 12141480,
      "step": 18565
    },
    {
      "epoch": 9.732704402515724,
      "grad_norm": 2.1895639896392822,
      "learning_rate": 3.0486874133881533e-05,
      "loss": 0.1121,
      "num_input_tokens_seen": 12144168,
      "step": 18570
    },
    {
      "epoch": 9.735324947589099,
      "grad_norm": 2.155874013900757,
      "learning_rate": 3.047571806466823e-05,
      "loss": 0.1636,
      "num_input_tokens_seen": 12148168,
      "step": 18575
    },
    {
      "epoch": 9.737945492662474,
      "grad_norm": 1.2359836101531982,
      "learning_rate": 3.0464560849997177e-05,
      "loss": 0.0958,
      "num_input_tokens_seen": 12150696,
      "step": 18580
    },
    {
      "epoch": 9.74056603773585,
      "grad_norm": 2.6199824810028076,
      "learning_rate": 3.0453402492202325e-05,
      "loss": 0.1142,
      "num_input_tokens_seen": 12153928,
      "step": 18585
    },
    {
      "epoch": 9.743186582809225,
      "grad_norm": 2.4939658641815186,
      "learning_rate": 3.0442242993617904e-05,
      "loss": 0.1409,
      "num_input_tokens_seen": 12157384,
      "step": 18590
    },
    {
      "epoch": 9.7458071278826,
      "grad_norm": 2.288037061691284,
      "learning_rate": 3.0431082356578334e-05,
      "loss": 0.1372,
      "num_input_tokens_seen": 12160200,
      "step": 18595
    },
    {
      "epoch": 9.748427672955975,
      "grad_norm": 2.0418262481689453,
      "learning_rate": 3.0419920583418294e-05,
      "loss": 0.1755,
      "num_input_tokens_seen": 12162504,
      "step": 18600
    },
    {
      "epoch": 9.75104821802935,
      "grad_norm": 2.3984572887420654,
      "learning_rate": 3.0408757676472694e-05,
      "loss": 0.1454,
      "num_input_tokens_seen": 12165288,
      "step": 18605
    },
    {
      "epoch": 9.753668763102725,
      "grad_norm": 5.002269744873047,
      "learning_rate": 3.0397593638076705e-05,
      "loss": 0.1148,
      "num_input_tokens_seen": 12167912,
      "step": 18610
    },
    {
      "epoch": 9.7562893081761,
      "grad_norm": 16.567359924316406,
      "learning_rate": 3.0386428470565702e-05,
      "loss": 0.2116,
      "num_input_tokens_seen": 12171624,
      "step": 18615
    },
    {
      "epoch": 9.758909853249476,
      "grad_norm": 4.45750093460083,
      "learning_rate": 3.0375262176275308e-05,
      "loss": 0.1276,
      "num_input_tokens_seen": 12174632,
      "step": 18620
    },
    {
      "epoch": 9.76153039832285,
      "grad_norm": 5.2537455558776855,
      "learning_rate": 3.03640947575414e-05,
      "loss": 0.2024,
      "num_input_tokens_seen": 12177896,
      "step": 18625
    },
    {
      "epoch": 9.764150943396226,
      "grad_norm": 7.090062618255615,
      "learning_rate": 3.0352926216700067e-05,
      "loss": 0.1751,
      "num_input_tokens_seen": 12181256,
      "step": 18630
    },
    {
      "epoch": 9.766771488469601,
      "grad_norm": 3.8556110858917236,
      "learning_rate": 3.034175655608763e-05,
      "loss": 0.1604,
      "num_input_tokens_seen": 12183848,
      "step": 18635
    },
    {
      "epoch": 9.769392033542976,
      "grad_norm": 1.9736456871032715,
      "learning_rate": 3.0330585778040677e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 12187208,
      "step": 18640
    },
    {
      "epoch": 9.772012578616351,
      "grad_norm": 2.008349895477295,
      "learning_rate": 3.031941388489598e-05,
      "loss": 0.1375,
      "num_input_tokens_seen": 12190568,
      "step": 18645
    },
    {
      "epoch": 9.774633123689728,
      "grad_norm": 2.1562998294830322,
      "learning_rate": 3.0308240878990584e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 12193224,
      "step": 18650
    },
    {
      "epoch": 9.777253668763104,
      "grad_norm": 5.2356157302856445,
      "learning_rate": 3.0297066762661773e-05,
      "loss": 0.1094,
      "num_input_tokens_seen": 12195816,
      "step": 18655
    },
    {
      "epoch": 9.779874213836479,
      "grad_norm": 4.539966583251953,
      "learning_rate": 3.0285891538247015e-05,
      "loss": 0.154,
      "num_input_tokens_seen": 12198856,
      "step": 18660
    },
    {
      "epoch": 9.782494758909854,
      "grad_norm": 2.0953338146209717,
      "learning_rate": 3.0274715208084053e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 12201992,
      "step": 18665
    },
    {
      "epoch": 9.785115303983229,
      "grad_norm": 1.5738404989242554,
      "learning_rate": 3.0263537774510847e-05,
      "loss": 0.1176,
      "num_input_tokens_seen": 12205608,
      "step": 18670
    },
    {
      "epoch": 9.787735849056604,
      "grad_norm": 3.0021111965179443,
      "learning_rate": 3.0252359239865585e-05,
      "loss": 0.1824,
      "num_input_tokens_seen": 12209352,
      "step": 18675
    },
    {
      "epoch": 9.79035639412998,
      "grad_norm": 2.942917823791504,
      "learning_rate": 3.0241179606486704e-05,
      "loss": 0.1228,
      "num_input_tokens_seen": 12213768,
      "step": 18680
    },
    {
      "epoch": 9.792976939203355,
      "grad_norm": 3.1001341342926025,
      "learning_rate": 3.0229998876712827e-05,
      "loss": 0.1102,
      "num_input_tokens_seen": 12216776,
      "step": 18685
    },
    {
      "epoch": 9.79559748427673,
      "grad_norm": 1.6897404193878174,
      "learning_rate": 3.0218817052882858e-05,
      "loss": 0.122,
      "num_input_tokens_seen": 12221000,
      "step": 18690
    },
    {
      "epoch": 9.798218029350105,
      "grad_norm": 4.362632751464844,
      "learning_rate": 3.020763413733591e-05,
      "loss": 0.1452,
      "num_input_tokens_seen": 12224200,
      "step": 18695
    },
    {
      "epoch": 9.80083857442348,
      "grad_norm": 2.451937437057495,
      "learning_rate": 3.0196450132411302e-05,
      "loss": 0.1132,
      "num_input_tokens_seen": 12226920,
      "step": 18700
    },
    {
      "epoch": 9.803459119496855,
      "grad_norm": 2.3663928508758545,
      "learning_rate": 3.0185265040448616e-05,
      "loss": 0.1723,
      "num_input_tokens_seen": 12229960,
      "step": 18705
    },
    {
      "epoch": 9.80607966457023,
      "grad_norm": 1.9920318126678467,
      "learning_rate": 3.017407886378764e-05,
      "loss": 0.1361,
      "num_input_tokens_seen": 12232040,
      "step": 18710
    },
    {
      "epoch": 9.808700209643606,
      "grad_norm": 1.1453500986099243,
      "learning_rate": 3.016289160476839e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 12234472,
      "step": 18715
    },
    {
      "epoch": 9.81132075471698,
      "grad_norm": 5.5663371086120605,
      "learning_rate": 3.0151703265731125e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 12237448,
      "step": 18720
    },
    {
      "epoch": 9.813941299790356,
      "grad_norm": 1.785191535949707,
      "learning_rate": 3.0140513849016305e-05,
      "loss": 0.1779,
      "num_input_tokens_seen": 12240264,
      "step": 18725
    },
    {
      "epoch": 9.816561844863731,
      "grad_norm": 2.0088376998901367,
      "learning_rate": 3.0129323356964627e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 12243176,
      "step": 18730
    },
    {
      "epoch": 9.819182389937106,
      "grad_norm": 4.891030311584473,
      "learning_rate": 3.0118131791917032e-05,
      "loss": 0.1089,
      "num_input_tokens_seen": 12245864,
      "step": 18735
    },
    {
      "epoch": 9.821802935010481,
      "grad_norm": 2.807803153991699,
      "learning_rate": 3.010693915621466e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 12248712,
      "step": 18740
    },
    {
      "epoch": 9.824423480083858,
      "grad_norm": 2.1496129035949707,
      "learning_rate": 3.009574545219887e-05,
      "loss": 0.1592,
      "num_input_tokens_seen": 12251144,
      "step": 18745
    },
    {
      "epoch": 9.827044025157234,
      "grad_norm": 1.9344474077224731,
      "learning_rate": 3.0084550682211272e-05,
      "loss": 0.136,
      "num_input_tokens_seen": 12255336,
      "step": 18750
    },
    {
      "epoch": 9.829664570230609,
      "grad_norm": 5.541412830352783,
      "learning_rate": 3.0073354848593677e-05,
      "loss": 0.1122,
      "num_input_tokens_seen": 12257672,
      "step": 18755
    },
    {
      "epoch": 9.832285115303984,
      "grad_norm": 3.8696072101593018,
      "learning_rate": 3.0062157953688125e-05,
      "loss": 0.1414,
      "num_input_tokens_seen": 12261032,
      "step": 18760
    },
    {
      "epoch": 9.834905660377359,
      "grad_norm": 2.3443188667297363,
      "learning_rate": 3.0050959999836887e-05,
      "loss": 0.1444,
      "num_input_tokens_seen": 12264488,
      "step": 18765
    },
    {
      "epoch": 9.837526205450734,
      "grad_norm": 2.6955935955047607,
      "learning_rate": 3.003976098938244e-05,
      "loss": 0.1401,
      "num_input_tokens_seen": 12267432,
      "step": 18770
    },
    {
      "epoch": 9.84014675052411,
      "grad_norm": 2.2477664947509766,
      "learning_rate": 3.0028560924667487e-05,
      "loss": 0.0991,
      "num_input_tokens_seen": 12270888,
      "step": 18775
    },
    {
      "epoch": 9.842767295597485,
      "grad_norm": 4.122145652770996,
      "learning_rate": 3.001735980803496e-05,
      "loss": 0.125,
      "num_input_tokens_seen": 12273896,
      "step": 18780
    },
    {
      "epoch": 9.84538784067086,
      "grad_norm": 4.80640983581543,
      "learning_rate": 3.0006157641828004e-05,
      "loss": 0.148,
      "num_input_tokens_seen": 12276328,
      "step": 18785
    },
    {
      "epoch": 9.848008385744235,
      "grad_norm": 2.2000198364257812,
      "learning_rate": 2.9994954428389977e-05,
      "loss": 0.1103,
      "num_input_tokens_seen": 12279016,
      "step": 18790
    },
    {
      "epoch": 9.85062893081761,
      "grad_norm": 0.9106708765029907,
      "learning_rate": 2.9983750170064472e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 12281512,
      "step": 18795
    },
    {
      "epoch": 9.853249475890985,
      "grad_norm": 2.174830913543701,
      "learning_rate": 2.9972544869195292e-05,
      "loss": 0.1848,
      "num_input_tokens_seen": 12285160,
      "step": 18800
    },
    {
      "epoch": 9.85587002096436,
      "grad_norm": 1.3771566152572632,
      "learning_rate": 2.9961338528126447e-05,
      "loss": 0.1892,
      "num_input_tokens_seen": 12288616,
      "step": 18805
    },
    {
      "epoch": 9.858490566037736,
      "grad_norm": 2.703800916671753,
      "learning_rate": 2.9950131149202192e-05,
      "loss": 0.2446,
      "num_input_tokens_seen": 12292680,
      "step": 18810
    },
    {
      "epoch": 9.86111111111111,
      "grad_norm": 1.6331137418746948,
      "learning_rate": 2.9938922734766973e-05,
      "loss": 0.1552,
      "num_input_tokens_seen": 12297160,
      "step": 18815
    },
    {
      "epoch": 9.863731656184486,
      "grad_norm": 1.4520777463912964,
      "learning_rate": 2.992771328716546e-05,
      "loss": 0.1787,
      "num_input_tokens_seen": 12300520,
      "step": 18820
    },
    {
      "epoch": 9.866352201257861,
      "grad_norm": 1.404599666595459,
      "learning_rate": 2.991650280874255e-05,
      "loss": 0.1864,
      "num_input_tokens_seen": 12303496,
      "step": 18825
    },
    {
      "epoch": 9.868972746331236,
      "grad_norm": 2.6481738090515137,
      "learning_rate": 2.990529130184334e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 12306568,
      "step": 18830
    },
    {
      "epoch": 9.871593291404611,
      "grad_norm": 1.6502755880355835,
      "learning_rate": 2.989407876881316e-05,
      "loss": 0.172,
      "num_input_tokens_seen": 12309160,
      "step": 18835
    },
    {
      "epoch": 9.874213836477988,
      "grad_norm": 1.7088278532028198,
      "learning_rate": 2.988286521199753e-05,
      "loss": 0.1463,
      "num_input_tokens_seen": 12312776,
      "step": 18840
    },
    {
      "epoch": 9.876834381551364,
      "grad_norm": 2.90170955657959,
      "learning_rate": 2.9871650633742198e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 12316104,
      "step": 18845
    },
    {
      "epoch": 9.879454926624739,
      "grad_norm": 2.4007012844085693,
      "learning_rate": 2.9860435036393142e-05,
      "loss": 0.1678,
      "num_input_tokens_seen": 12318568,
      "step": 18850
    },
    {
      "epoch": 9.882075471698114,
      "grad_norm": 1.6911659240722656,
      "learning_rate": 2.9849218422296527e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 12321736,
      "step": 18855
    },
    {
      "epoch": 9.884696016771489,
      "grad_norm": 1.6483945846557617,
      "learning_rate": 2.983800079379873e-05,
      "loss": 0.1778,
      "num_input_tokens_seen": 12325032,
      "step": 18860
    },
    {
      "epoch": 9.887316561844864,
      "grad_norm": 2.2468063831329346,
      "learning_rate": 2.982678215324638e-05,
      "loss": 0.2098,
      "num_input_tokens_seen": 12329160,
      "step": 18865
    },
    {
      "epoch": 9.88993710691824,
      "grad_norm": 2.012514114379883,
      "learning_rate": 2.9815562502986256e-05,
      "loss": 0.3122,
      "num_input_tokens_seen": 12332392,
      "step": 18870
    },
    {
      "epoch": 9.892557651991615,
      "grad_norm": 2.175074815750122,
      "learning_rate": 2.9804341845365398e-05,
      "loss": 0.162,
      "num_input_tokens_seen": 12336456,
      "step": 18875
    },
    {
      "epoch": 9.89517819706499,
      "grad_norm": 1.7467129230499268,
      "learning_rate": 2.9793120182731037e-05,
      "loss": 0.1436,
      "num_input_tokens_seen": 12339080,
      "step": 18880
    },
    {
      "epoch": 9.897798742138365,
      "grad_norm": 2.446732521057129,
      "learning_rate": 2.9781897517430617e-05,
      "loss": 0.1102,
      "num_input_tokens_seen": 12341768,
      "step": 18885
    },
    {
      "epoch": 9.90041928721174,
      "grad_norm": 1.7392455339431763,
      "learning_rate": 2.977067385181179e-05,
      "loss": 0.1368,
      "num_input_tokens_seen": 12344936,
      "step": 18890
    },
    {
      "epoch": 9.903039832285115,
      "grad_norm": 2.402162790298462,
      "learning_rate": 2.9759449188222426e-05,
      "loss": 0.1194,
      "num_input_tokens_seen": 12349416,
      "step": 18895
    },
    {
      "epoch": 9.90566037735849,
      "grad_norm": 2.3639354705810547,
      "learning_rate": 2.9748223529010582e-05,
      "loss": 0.1248,
      "num_input_tokens_seen": 12352360,
      "step": 18900
    },
    {
      "epoch": 9.908280922431866,
      "grad_norm": 2.067545175552368,
      "learning_rate": 2.9736996876524546e-05,
      "loss": 0.1852,
      "num_input_tokens_seen": 12355688,
      "step": 18905
    },
    {
      "epoch": 9.91090146750524,
      "grad_norm": 1.1813477277755737,
      "learning_rate": 2.9725769233112816e-05,
      "loss": 0.1561,
      "num_input_tokens_seen": 12359560,
      "step": 18910
    },
    {
      "epoch": 9.913522012578616,
      "grad_norm": 3.0779764652252197,
      "learning_rate": 2.9714540601124074e-05,
      "loss": 0.2172,
      "num_input_tokens_seen": 12363656,
      "step": 18915
    },
    {
      "epoch": 9.916142557651991,
      "grad_norm": 1.5874422788619995,
      "learning_rate": 2.9703310982907218e-05,
      "loss": 0.1262,
      "num_input_tokens_seen": 12367688,
      "step": 18920
    },
    {
      "epoch": 9.918763102725366,
      "grad_norm": 2.851893424987793,
      "learning_rate": 2.9692080380811365e-05,
      "loss": 0.1381,
      "num_input_tokens_seen": 12371400,
      "step": 18925
    },
    {
      "epoch": 9.921383647798741,
      "grad_norm": 2.3207552433013916,
      "learning_rate": 2.9680848797185828e-05,
      "loss": 0.1267,
      "num_input_tokens_seen": 12374152,
      "step": 18930
    },
    {
      "epoch": 9.924004192872118,
      "grad_norm": 2.5917842388153076,
      "learning_rate": 2.966961623438012e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 12378472,
      "step": 18935
    },
    {
      "epoch": 9.926624737945493,
      "grad_norm": 3.027552843093872,
      "learning_rate": 2.9658382694743973e-05,
      "loss": 0.1465,
      "num_input_tokens_seen": 12381480,
      "step": 18940
    },
    {
      "epoch": 9.929245283018869,
      "grad_norm": 1.3025680780410767,
      "learning_rate": 2.9647148180627317e-05,
      "loss": 0.164,
      "num_input_tokens_seen": 12384648,
      "step": 18945
    },
    {
      "epoch": 9.931865828092244,
      "grad_norm": 1.3616704940795898,
      "learning_rate": 2.9635912694380275e-05,
      "loss": 0.104,
      "num_input_tokens_seen": 12387432,
      "step": 18950
    },
    {
      "epoch": 9.934486373165619,
      "grad_norm": 2.4375264644622803,
      "learning_rate": 2.962467623835319e-05,
      "loss": 0.142,
      "num_input_tokens_seen": 12390984,
      "step": 18955
    },
    {
      "epoch": 9.937106918238994,
      "grad_norm": 2.795006275177002,
      "learning_rate": 2.961343881489659e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 12394760,
      "step": 18960
    },
    {
      "epoch": 9.93972746331237,
      "grad_norm": 5.37154483795166,
      "learning_rate": 2.9602200426361226e-05,
      "loss": 0.2078,
      "num_input_tokens_seen": 12397768,
      "step": 18965
    },
    {
      "epoch": 9.942348008385745,
      "grad_norm": 2.01718807220459,
      "learning_rate": 2.9590961075098038e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 12400968,
      "step": 18970
    },
    {
      "epoch": 9.94496855345912,
      "grad_norm": 2.4187912940979004,
      "learning_rate": 2.957972076345817e-05,
      "loss": 0.1607,
      "num_input_tokens_seen": 12406664,
      "step": 18975
    },
    {
      "epoch": 9.947589098532495,
      "grad_norm": 2.2251088619232178,
      "learning_rate": 2.9568479493792967e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 12409448,
      "step": 18980
    },
    {
      "epoch": 9.95020964360587,
      "grad_norm": 1.616584300994873,
      "learning_rate": 2.955723726845397e-05,
      "loss": 0.0589,
      "num_input_tokens_seen": 12412744,
      "step": 18985
    },
    {
      "epoch": 9.952830188679245,
      "grad_norm": 2.8177764415740967,
      "learning_rate": 2.9545994089792928e-05,
      "loss": 0.2658,
      "num_input_tokens_seen": 12415656,
      "step": 18990
    },
    {
      "epoch": 9.95545073375262,
      "grad_norm": 2.9741339683532715,
      "learning_rate": 2.9534749960161795e-05,
      "loss": 0.1745,
      "num_input_tokens_seen": 12418472,
      "step": 18995
    },
    {
      "epoch": 9.958071278825996,
      "grad_norm": 2.6695799827575684,
      "learning_rate": 2.9523504881912694e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 12421768,
      "step": 19000
    },
    {
      "epoch": 9.96069182389937,
      "grad_norm": 2.199805498123169,
      "learning_rate": 2.9512258857397978e-05,
      "loss": 0.18,
      "num_input_tokens_seen": 12425800,
      "step": 19005
    },
    {
      "epoch": 9.963312368972746,
      "grad_norm": 2.8579838275909424,
      "learning_rate": 2.9501011888970203e-05,
      "loss": 0.1664,
      "num_input_tokens_seen": 12429288,
      "step": 19010
    },
    {
      "epoch": 9.965932914046121,
      "grad_norm": 2.915940284729004,
      "learning_rate": 2.9489763978982075e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 12433160,
      "step": 19015
    },
    {
      "epoch": 9.968553459119496,
      "grad_norm": 1.0635759830474854,
      "learning_rate": 2.9478515129786556e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 12435944,
      "step": 19020
    },
    {
      "epoch": 9.971174004192871,
      "grad_norm": 1.8614065647125244,
      "learning_rate": 2.946726534373676e-05,
      "loss": 0.1157,
      "num_input_tokens_seen": 12438216,
      "step": 19025
    },
    {
      "epoch": 9.973794549266248,
      "grad_norm": 1.7913899421691895,
      "learning_rate": 2.945601462318601e-05,
      "loss": 0.1129,
      "num_input_tokens_seen": 12440904,
      "step": 19030
    },
    {
      "epoch": 9.976415094339622,
      "grad_norm": 2.12516713142395,
      "learning_rate": 2.944476297048786e-05,
      "loss": 0.1182,
      "num_input_tokens_seen": 12443752,
      "step": 19035
    },
    {
      "epoch": 9.979035639412999,
      "grad_norm": 9.38184642791748,
      "learning_rate": 2.9433510387995987e-05,
      "loss": 0.2191,
      "num_input_tokens_seen": 12446152,
      "step": 19040
    },
    {
      "epoch": 9.981656184486374,
      "grad_norm": 4.53554105758667,
      "learning_rate": 2.9422256878064325e-05,
      "loss": 0.097,
      "num_input_tokens_seen": 12449160,
      "step": 19045
    },
    {
      "epoch": 9.984276729559749,
      "grad_norm": 2.291579246520996,
      "learning_rate": 2.9411002443046982e-05,
      "loss": 0.1956,
      "num_input_tokens_seen": 12453064,
      "step": 19050
    },
    {
      "epoch": 9.986897274633124,
      "grad_norm": 9.331277847290039,
      "learning_rate": 2.9399747085298245e-05,
      "loss": 0.1409,
      "num_input_tokens_seen": 12455944,
      "step": 19055
    },
    {
      "epoch": 9.9895178197065,
      "grad_norm": 3.947296380996704,
      "learning_rate": 2.9388490807172614e-05,
      "loss": 0.1692,
      "num_input_tokens_seen": 12459528,
      "step": 19060
    },
    {
      "epoch": 9.992138364779874,
      "grad_norm": 1.8039560317993164,
      "learning_rate": 2.9377233611024775e-05,
      "loss": 0.1755,
      "num_input_tokens_seen": 12463048,
      "step": 19065
    },
    {
      "epoch": 9.99475890985325,
      "grad_norm": 1.917494773864746,
      "learning_rate": 2.9365975499209596e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 12467240,
      "step": 19070
    },
    {
      "epoch": 9.997379454926625,
      "grad_norm": 2.1162006855010986,
      "learning_rate": 2.9354716474082154e-05,
      "loss": 0.286,
      "num_input_tokens_seen": 12470216,
      "step": 19075
    },
    {
      "epoch": 10.0,
      "grad_norm": 5.538087844848633,
      "learning_rate": 2.9343456537997714e-05,
      "loss": 0.2099,
      "num_input_tokens_seen": 12472912,
      "step": 19080
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.8089353442192078,
      "eval_runtime": 11.6381,
      "eval_samples_per_second": 72.864,
      "eval_steps_per_second": 18.216,
      "num_input_tokens_seen": 12472912,
      "step": 19080
    },
    {
      "epoch": 10.002620545073375,
      "grad_norm": 1.3531988859176636,
      "learning_rate": 2.9332195693311714e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 12475568,
      "step": 19085
    },
    {
      "epoch": 10.00524109014675,
      "grad_norm": 1.522802472114563,
      "learning_rate": 2.9320933942379803e-05,
      "loss": 0.0448,
      "num_input_tokens_seen": 12478128,
      "step": 19090
    },
    {
      "epoch": 10.007861635220126,
      "grad_norm": 1.2846890687942505,
      "learning_rate": 2.9309671287557805e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 12481520,
      "step": 19095
    },
    {
      "epoch": 10.0104821802935,
      "grad_norm": 1.635995864868164,
      "learning_rate": 2.9298407731201737e-05,
      "loss": 0.1406,
      "num_input_tokens_seen": 12485264,
      "step": 19100
    },
    {
      "epoch": 10.013102725366876,
      "grad_norm": 1.5886133909225464,
      "learning_rate": 2.928714327566781e-05,
      "loss": 0.0762,
      "num_input_tokens_seen": 12488464,
      "step": 19105
    },
    {
      "epoch": 10.015723270440251,
      "grad_norm": 2.9472031593322754,
      "learning_rate": 2.9275877923312427e-05,
      "loss": 0.0922,
      "num_input_tokens_seen": 12491504,
      "step": 19110
    },
    {
      "epoch": 10.018343815513626,
      "grad_norm": 4.0031232833862305,
      "learning_rate": 2.9264611676492165e-05,
      "loss": 0.2096,
      "num_input_tokens_seen": 12494576,
      "step": 19115
    },
    {
      "epoch": 10.020964360587001,
      "grad_norm": 1.625637173652649,
      "learning_rate": 2.9253344537563782e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 12497648,
      "step": 19120
    },
    {
      "epoch": 10.023584905660377,
      "grad_norm": 2.628143072128296,
      "learning_rate": 2.924207650888426e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 12500208,
      "step": 19125
    },
    {
      "epoch": 10.026205450733753,
      "grad_norm": 4.2709527015686035,
      "learning_rate": 2.9230807592810723e-05,
      "loss": 0.1185,
      "num_input_tokens_seen": 12503248,
      "step": 19130
    },
    {
      "epoch": 10.028825995807129,
      "grad_norm": 1.5374776124954224,
      "learning_rate": 2.92195377917005e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 12508080,
      "step": 19135
    },
    {
      "epoch": 10.031446540880504,
      "grad_norm": 1.3583366870880127,
      "learning_rate": 2.920826710791112e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 12511856,
      "step": 19140
    },
    {
      "epoch": 10.034067085953879,
      "grad_norm": 2.7152819633483887,
      "learning_rate": 2.919699554380026e-05,
      "loss": 0.1033,
      "num_input_tokens_seen": 12515920,
      "step": 19145
    },
    {
      "epoch": 10.036687631027254,
      "grad_norm": 2.9736881256103516,
      "learning_rate": 2.9185723101725825e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 12519024,
      "step": 19150
    },
    {
      "epoch": 10.03930817610063,
      "grad_norm": 2.516216278076172,
      "learning_rate": 2.9174449784045867e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 12521840,
      "step": 19155
    },
    {
      "epoch": 10.041928721174004,
      "grad_norm": 1.8522863388061523,
      "learning_rate": 2.916317559311863e-05,
      "loss": 0.1147,
      "num_input_tokens_seen": 12525008,
      "step": 19160
    },
    {
      "epoch": 10.04454926624738,
      "grad_norm": 1.9251779317855835,
      "learning_rate": 2.9151900531302566e-05,
      "loss": 0.0917,
      "num_input_tokens_seen": 12528048,
      "step": 19165
    },
    {
      "epoch": 10.047169811320755,
      "grad_norm": 2.5607075691223145,
      "learning_rate": 2.9140624600956272e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 12531536,
      "step": 19170
    },
    {
      "epoch": 10.04979035639413,
      "grad_norm": 2.5065720081329346,
      "learning_rate": 2.912934780443855e-05,
      "loss": 0.1256,
      "num_input_tokens_seen": 12535248,
      "step": 19175
    },
    {
      "epoch": 10.052410901467505,
      "grad_norm": 1.3678336143493652,
      "learning_rate": 2.9118070144108388e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 12538128,
      "step": 19180
    },
    {
      "epoch": 10.05503144654088,
      "grad_norm": 1.2766627073287964,
      "learning_rate": 2.9106791622324924e-05,
      "loss": 0.0928,
      "num_input_tokens_seen": 12541136,
      "step": 19185
    },
    {
      "epoch": 10.057651991614255,
      "grad_norm": 1.8068407773971558,
      "learning_rate": 2.909551224144751e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 12544272,
      "step": 19190
    },
    {
      "epoch": 10.06027253668763,
      "grad_norm": 2.4681813716888428,
      "learning_rate": 2.9084232003835653e-05,
      "loss": 0.1009,
      "num_input_tokens_seen": 12547184,
      "step": 19195
    },
    {
      "epoch": 10.062893081761006,
      "grad_norm": 3.1708905696868896,
      "learning_rate": 2.9072950911849058e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 12550416,
      "step": 19200
    },
    {
      "epoch": 10.065513626834381,
      "grad_norm": 3.650263547897339,
      "learning_rate": 2.90616689678476e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 12552688,
      "step": 19205
    },
    {
      "epoch": 10.068134171907756,
      "grad_norm": 1.4742413759231567,
      "learning_rate": 2.9050386174191337e-05,
      "loss": 0.1141,
      "num_input_tokens_seen": 12555536,
      "step": 19210
    },
    {
      "epoch": 10.070754716981131,
      "grad_norm": 1.7405692338943481,
      "learning_rate": 2.903910253324048e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 12558672,
      "step": 19215
    },
    {
      "epoch": 10.073375262054507,
      "grad_norm": 3.185971260070801,
      "learning_rate": 2.9027818047355477e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 12562192,
      "step": 19220
    },
    {
      "epoch": 10.075995807127883,
      "grad_norm": 1.845445990562439,
      "learning_rate": 2.9016532718896873e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 12565584,
      "step": 19225
    },
    {
      "epoch": 10.078616352201259,
      "grad_norm": 2.520451784133911,
      "learning_rate": 2.900524655022545e-05,
      "loss": 0.1088,
      "num_input_tokens_seen": 12568720,
      "step": 19230
    },
    {
      "epoch": 10.081236897274634,
      "grad_norm": 2.2738335132598877,
      "learning_rate": 2.899395954370215e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 12572432,
      "step": 19235
    },
    {
      "epoch": 10.083857442348009,
      "grad_norm": 2.3220200538635254,
      "learning_rate": 2.898267170168807e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 12575632,
      "step": 19240
    },
    {
      "epoch": 10.086477987421384,
      "grad_norm": 1.3607420921325684,
      "learning_rate": 2.8971383026544507e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 12579888,
      "step": 19245
    },
    {
      "epoch": 10.08909853249476,
      "grad_norm": 1.4177732467651367,
      "learning_rate": 2.896009352063293e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 12582416,
      "step": 19250
    },
    {
      "epoch": 10.091719077568134,
      "grad_norm": 2.248706340789795,
      "learning_rate": 2.894880318631496e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 12585360,
      "step": 19255
    },
    {
      "epoch": 10.09433962264151,
      "grad_norm": 1.4659414291381836,
      "learning_rate": 2.8937512025952417e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 12589456,
      "step": 19260
    },
    {
      "epoch": 10.096960167714885,
      "grad_norm": 2.09352707862854,
      "learning_rate": 2.8926220041907283e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 12592752,
      "step": 19265
    },
    {
      "epoch": 10.09958071278826,
      "grad_norm": 2.1759021282196045,
      "learning_rate": 2.8914927236541706e-05,
      "loss": 0.1553,
      "num_input_tokens_seen": 12596304,
      "step": 19270
    },
    {
      "epoch": 10.102201257861635,
      "grad_norm": 4.975203990936279,
      "learning_rate": 2.8903633612218012e-05,
      "loss": 0.151,
      "num_input_tokens_seen": 12600048,
      "step": 19275
    },
    {
      "epoch": 10.10482180293501,
      "grad_norm": 2.622962236404419,
      "learning_rate": 2.8892339171298703e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 12602992,
      "step": 19280
    },
    {
      "epoch": 10.107442348008385,
      "grad_norm": 4.872739791870117,
      "learning_rate": 2.888104391614645e-05,
      "loss": 0.1083,
      "num_input_tokens_seen": 12606960,
      "step": 19285
    },
    {
      "epoch": 10.11006289308176,
      "grad_norm": 2.1887080669403076,
      "learning_rate": 2.886974784912408e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 12609584,
      "step": 19290
    },
    {
      "epoch": 10.112683438155136,
      "grad_norm": 1.201395869255066,
      "learning_rate": 2.8858450972594612e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 12612368,
      "step": 19295
    },
    {
      "epoch": 10.115303983228511,
      "grad_norm": 2.22279953956604,
      "learning_rate": 2.8847153288921226e-05,
      "loss": 0.0949,
      "num_input_tokens_seen": 12615472,
      "step": 19300
    },
    {
      "epoch": 10.117924528301886,
      "grad_norm": 2.4032864570617676,
      "learning_rate": 2.883585480046725e-05,
      "loss": 0.1524,
      "num_input_tokens_seen": 12621456,
      "step": 19305
    },
    {
      "epoch": 10.120545073375261,
      "grad_norm": 3.6406822204589844,
      "learning_rate": 2.8824555509596223e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 12624624,
      "step": 19310
    },
    {
      "epoch": 10.123165618448636,
      "grad_norm": 6.122007846832275,
      "learning_rate": 2.8813255418671815e-05,
      "loss": 0.1139,
      "num_input_tokens_seen": 12628112,
      "step": 19315
    },
    {
      "epoch": 10.125786163522013,
      "grad_norm": 2.598405361175537,
      "learning_rate": 2.8801954530057874e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 12631536,
      "step": 19320
    },
    {
      "epoch": 10.128406708595389,
      "grad_norm": 1.8082233667373657,
      "learning_rate": 2.879065284611842e-05,
      "loss": 0.0623,
      "num_input_tokens_seen": 12634704,
      "step": 19325
    },
    {
      "epoch": 10.131027253668764,
      "grad_norm": 1.4816266298294067,
      "learning_rate": 2.8779350369217634e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 12638640,
      "step": 19330
    },
    {
      "epoch": 10.133647798742139,
      "grad_norm": 1.524580717086792,
      "learning_rate": 2.876804710171987e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 12641392,
      "step": 19335
    },
    {
      "epoch": 10.136268343815514,
      "grad_norm": 2.7718183994293213,
      "learning_rate": 2.8756743045989643e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 12648656,
      "step": 19340
    },
    {
      "epoch": 10.13888888888889,
      "grad_norm": 5.063328266143799,
      "learning_rate": 2.8745438204391623e-05,
      "loss": 0.0836,
      "num_input_tokens_seen": 12651760,
      "step": 19345
    },
    {
      "epoch": 10.141509433962264,
      "grad_norm": 1.7269881963729858,
      "learning_rate": 2.8734132579290675e-05,
      "loss": 0.1116,
      "num_input_tokens_seen": 12655120,
      "step": 19350
    },
    {
      "epoch": 10.14412997903564,
      "grad_norm": 6.602084636688232,
      "learning_rate": 2.8722826173051782e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 12657936,
      "step": 19355
    },
    {
      "epoch": 10.146750524109015,
      "grad_norm": 3.1622653007507324,
      "learning_rate": 2.871151898804012e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 12661648,
      "step": 19360
    },
    {
      "epoch": 10.14937106918239,
      "grad_norm": 1.4414055347442627,
      "learning_rate": 2.8700211026621043e-05,
      "loss": 0.1272,
      "num_input_tokens_seen": 12665328,
      "step": 19365
    },
    {
      "epoch": 10.151991614255765,
      "grad_norm": 2.8807296752929688,
      "learning_rate": 2.8688902291160025e-05,
      "loss": 0.1179,
      "num_input_tokens_seen": 12668464,
      "step": 19370
    },
    {
      "epoch": 10.15461215932914,
      "grad_norm": 2.1605117321014404,
      "learning_rate": 2.867759278402273e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 12672016,
      "step": 19375
    },
    {
      "epoch": 10.157232704402515,
      "grad_norm": 2.741445302963257,
      "learning_rate": 2.8666282507574994e-05,
      "loss": 0.1064,
      "num_input_tokens_seen": 12675312,
      "step": 19380
    },
    {
      "epoch": 10.15985324947589,
      "grad_norm": 3.361107587814331,
      "learning_rate": 2.8654971464182773e-05,
      "loss": 0.1502,
      "num_input_tokens_seen": 12677872,
      "step": 19385
    },
    {
      "epoch": 10.162473794549266,
      "grad_norm": 4.165562152862549,
      "learning_rate": 2.8643659656212222e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 12684112,
      "step": 19390
    },
    {
      "epoch": 10.165094339622641,
      "grad_norm": 3.4293878078460693,
      "learning_rate": 2.8632347086029642e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 12686736,
      "step": 19395
    },
    {
      "epoch": 10.167714884696016,
      "grad_norm": 1.1216944456100464,
      "learning_rate": 2.8621033756001486e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 12690352,
      "step": 19400
    },
    {
      "epoch": 10.170335429769391,
      "grad_norm": 1.4984623193740845,
      "learning_rate": 2.8609719668494394e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 12693264,
      "step": 19405
    },
    {
      "epoch": 10.172955974842766,
      "grad_norm": 2.1224188804626465,
      "learning_rate": 2.8598404825875113e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 12695888,
      "step": 19410
    },
    {
      "epoch": 10.175576519916143,
      "grad_norm": 3.5969607830047607,
      "learning_rate": 2.8587089230510604e-05,
      "loss": 0.0692,
      "num_input_tokens_seen": 12699216,
      "step": 19415
    },
    {
      "epoch": 10.178197064989519,
      "grad_norm": 1.4036011695861816,
      "learning_rate": 2.857577288476795e-05,
      "loss": 0.0819,
      "num_input_tokens_seen": 12703344,
      "step": 19420
    },
    {
      "epoch": 10.180817610062894,
      "grad_norm": 1.5616176128387451,
      "learning_rate": 2.85644557910144e-05,
      "loss": 0.0856,
      "num_input_tokens_seen": 12706480,
      "step": 19425
    },
    {
      "epoch": 10.183438155136269,
      "grad_norm": 1.5703669786453247,
      "learning_rate": 2.8553137951617366e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 12709744,
      "step": 19430
    },
    {
      "epoch": 10.186058700209644,
      "grad_norm": 2.062670946121216,
      "learning_rate": 2.854181936894441e-05,
      "loss": 0.0999,
      "num_input_tokens_seen": 12712816,
      "step": 19435
    },
    {
      "epoch": 10.18867924528302,
      "grad_norm": 3.670243263244629,
      "learning_rate": 2.8530500045363247e-05,
      "loss": 0.0831,
      "num_input_tokens_seen": 12716304,
      "step": 19440
    },
    {
      "epoch": 10.191299790356394,
      "grad_norm": 1.8187823295593262,
      "learning_rate": 2.851917998324175e-05,
      "loss": 0.0841,
      "num_input_tokens_seen": 12719216,
      "step": 19445
    },
    {
      "epoch": 10.19392033542977,
      "grad_norm": 2.205277442932129,
      "learning_rate": 2.8507859184947957e-05,
      "loss": 0.1252,
      "num_input_tokens_seen": 12723312,
      "step": 19450
    },
    {
      "epoch": 10.196540880503145,
      "grad_norm": 1.1264617443084717,
      "learning_rate": 2.8496537652850043e-05,
      "loss": 0.087,
      "num_input_tokens_seen": 12726320,
      "step": 19455
    },
    {
      "epoch": 10.19916142557652,
      "grad_norm": 3.382911205291748,
      "learning_rate": 2.8485215389316333e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 12730128,
      "step": 19460
    },
    {
      "epoch": 10.201781970649895,
      "grad_norm": 1.4371131658554077,
      "learning_rate": 2.8473892396715328e-05,
      "loss": 0.192,
      "num_input_tokens_seen": 12735024,
      "step": 19465
    },
    {
      "epoch": 10.20440251572327,
      "grad_norm": 1.0755422115325928,
      "learning_rate": 2.846256867741567e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 12737872,
      "step": 19470
    },
    {
      "epoch": 10.207023060796645,
      "grad_norm": 5.221160888671875,
      "learning_rate": 2.845124423378614e-05,
      "loss": 0.1424,
      "num_input_tokens_seen": 12740400,
      "step": 19475
    },
    {
      "epoch": 10.20964360587002,
      "grad_norm": 1.6818718910217285,
      "learning_rate": 2.8439919068195697e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 12743664,
      "step": 19480
    },
    {
      "epoch": 10.212264150943396,
      "grad_norm": 4.45837926864624,
      "learning_rate": 2.842859318301343e-05,
      "loss": 0.1525,
      "num_input_tokens_seen": 12746512,
      "step": 19485
    },
    {
      "epoch": 10.214884696016771,
      "grad_norm": 2.060892105102539,
      "learning_rate": 2.841726658060858e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 12750160,
      "step": 19490
    },
    {
      "epoch": 10.217505241090146,
      "grad_norm": 4.157636642456055,
      "learning_rate": 2.8405939263350555e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 12753520,
      "step": 19495
    },
    {
      "epoch": 10.220125786163521,
      "grad_norm": 2.604443311691284,
      "learning_rate": 2.8394611233608886e-05,
      "loss": 0.1172,
      "num_input_tokens_seen": 12756624,
      "step": 19500
    },
    {
      "epoch": 10.222746331236896,
      "grad_norm": 3.1190831661224365,
      "learning_rate": 2.8383282493753283e-05,
      "loss": 0.0897,
      "num_input_tokens_seen": 12759248,
      "step": 19505
    },
    {
      "epoch": 10.225366876310272,
      "grad_norm": 3.033435821533203,
      "learning_rate": 2.8371953046153575e-05,
      "loss": 0.0915,
      "num_input_tokens_seen": 12762384,
      "step": 19510
    },
    {
      "epoch": 10.227987421383649,
      "grad_norm": 1.660704493522644,
      "learning_rate": 2.8360622893179757e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 12766224,
      "step": 19515
    },
    {
      "epoch": 10.230607966457024,
      "grad_norm": 3.085876941680908,
      "learning_rate": 2.834929203720197e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 12768912,
      "step": 19520
    },
    {
      "epoch": 10.233228511530399,
      "grad_norm": 1.5904204845428467,
      "learning_rate": 2.83379604805905e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 12772336,
      "step": 19525
    },
    {
      "epoch": 10.235849056603774,
      "grad_norm": 15.078893661499023,
      "learning_rate": 2.832662822571578e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 12775216,
      "step": 19530
    },
    {
      "epoch": 10.23846960167715,
      "grad_norm": 2.1156370639801025,
      "learning_rate": 2.8315295274948395e-05,
      "loss": 0.0867,
      "num_input_tokens_seen": 12778512,
      "step": 19535
    },
    {
      "epoch": 10.241090146750524,
      "grad_norm": 2.5630359649658203,
      "learning_rate": 2.8303961630659047e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 12782576,
      "step": 19540
    },
    {
      "epoch": 10.2437106918239,
      "grad_norm": 4.88059139251709,
      "learning_rate": 2.8292627295218627e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 12785584,
      "step": 19545
    },
    {
      "epoch": 10.246331236897275,
      "grad_norm": 4.231548309326172,
      "learning_rate": 2.828129227099815e-05,
      "loss": 0.0842,
      "num_input_tokens_seen": 12788432,
      "step": 19550
    },
    {
      "epoch": 10.24895178197065,
      "grad_norm": 1.8474881649017334,
      "learning_rate": 2.826995656036875e-05,
      "loss": 0.1756,
      "num_input_tokens_seen": 12791312,
      "step": 19555
    },
    {
      "epoch": 10.251572327044025,
      "grad_norm": 2.4480276107788086,
      "learning_rate": 2.8258620165701747e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 12795120,
      "step": 19560
    },
    {
      "epoch": 10.2541928721174,
      "grad_norm": 1.1495671272277832,
      "learning_rate": 2.8247283089368586e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 12798032,
      "step": 19565
    },
    {
      "epoch": 10.256813417190775,
      "grad_norm": 2.876976728439331,
      "learning_rate": 2.8235945333740842e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 12800688,
      "step": 19570
    },
    {
      "epoch": 10.25943396226415,
      "grad_norm": 2.383733034133911,
      "learning_rate": 2.822460690119026e-05,
      "loss": 0.12,
      "num_input_tokens_seen": 12804240,
      "step": 19575
    },
    {
      "epoch": 10.262054507337526,
      "grad_norm": 2.4047324657440186,
      "learning_rate": 2.82132677940887e-05,
      "loss": 0.171,
      "num_input_tokens_seen": 12807216,
      "step": 19580
    },
    {
      "epoch": 10.264675052410901,
      "grad_norm": 1.4891396760940552,
      "learning_rate": 2.820192801480817e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 12810352,
      "step": 19585
    },
    {
      "epoch": 10.267295597484276,
      "grad_norm": 4.1971588134765625,
      "learning_rate": 2.8190587565720842e-05,
      "loss": 0.1012,
      "num_input_tokens_seen": 12813328,
      "step": 19590
    },
    {
      "epoch": 10.269916142557651,
      "grad_norm": 3.5856003761291504,
      "learning_rate": 2.8179246449198986e-05,
      "loss": 0.0934,
      "num_input_tokens_seen": 12816336,
      "step": 19595
    },
    {
      "epoch": 10.272536687631026,
      "grad_norm": 5.391551494598389,
      "learning_rate": 2.8167904667615048e-05,
      "loss": 0.1229,
      "num_input_tokens_seen": 12820624,
      "step": 19600
    },
    {
      "epoch": 10.275157232704402,
      "grad_norm": 5.8097734451293945,
      "learning_rate": 2.8156562223341592e-05,
      "loss": 0.0892,
      "num_input_tokens_seen": 12824176,
      "step": 19605
    },
    {
      "epoch": 10.277777777777779,
      "grad_norm": 1.9698123931884766,
      "learning_rate": 2.814521911875133e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 12827376,
      "step": 19610
    },
    {
      "epoch": 10.280398322851154,
      "grad_norm": 1.1608713865280151,
      "learning_rate": 2.8133875356217114e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 12830704,
      "step": 19615
    },
    {
      "epoch": 10.283018867924529,
      "grad_norm": 1.9259916543960571,
      "learning_rate": 2.8122530938111923e-05,
      "loss": 0.1176,
      "num_input_tokens_seen": 12833616,
      "step": 19620
    },
    {
      "epoch": 10.285639412997904,
      "grad_norm": 1.5730959177017212,
      "learning_rate": 2.8111185866808888e-05,
      "loss": 0.1007,
      "num_input_tokens_seen": 12836688,
      "step": 19625
    },
    {
      "epoch": 10.28825995807128,
      "grad_norm": 1.45634925365448,
      "learning_rate": 2.809984014468126e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 12840656,
      "step": 19630
    },
    {
      "epoch": 10.290880503144654,
      "grad_norm": 5.373927593231201,
      "learning_rate": 2.8088493774102443e-05,
      "loss": 0.0956,
      "num_input_tokens_seen": 12843408,
      "step": 19635
    },
    {
      "epoch": 10.29350104821803,
      "grad_norm": 1.7779680490493774,
      "learning_rate": 2.807714675744596e-05,
      "loss": 0.1239,
      "num_input_tokens_seen": 12846224,
      "step": 19640
    },
    {
      "epoch": 10.296121593291405,
      "grad_norm": 0.7458820939064026,
      "learning_rate": 2.8065799097085475e-05,
      "loss": 0.109,
      "num_input_tokens_seen": 12850288,
      "step": 19645
    },
    {
      "epoch": 10.29874213836478,
      "grad_norm": 2.6315033435821533,
      "learning_rate": 2.8054450795394804e-05,
      "loss": 0.1149,
      "num_input_tokens_seen": 12853424,
      "step": 19650
    },
    {
      "epoch": 10.301362683438155,
      "grad_norm": 3.3320353031158447,
      "learning_rate": 2.8043101854747878e-05,
      "loss": 0.0935,
      "num_input_tokens_seen": 12856656,
      "step": 19655
    },
    {
      "epoch": 10.30398322851153,
      "grad_norm": 1.6368122100830078,
      "learning_rate": 2.8031752277518748e-05,
      "loss": 0.1111,
      "num_input_tokens_seen": 12859248,
      "step": 19660
    },
    {
      "epoch": 10.306603773584905,
      "grad_norm": 3.5695765018463135,
      "learning_rate": 2.802040206608164e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 12862160,
      "step": 19665
    },
    {
      "epoch": 10.30922431865828,
      "grad_norm": 4.233432292938232,
      "learning_rate": 2.8009051222810877e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 12864304,
      "step": 19670
    },
    {
      "epoch": 10.311844863731656,
      "grad_norm": 9.380059242248535,
      "learning_rate": 2.7997699750080918e-05,
      "loss": 0.114,
      "num_input_tokens_seen": 12868752,
      "step": 19675
    },
    {
      "epoch": 10.314465408805031,
      "grad_norm": 1.2962405681610107,
      "learning_rate": 2.7986347650266386e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 12872112,
      "step": 19680
    },
    {
      "epoch": 10.317085953878406,
      "grad_norm": 3.559828281402588,
      "learning_rate": 2.797499492574199e-05,
      "loss": 0.0848,
      "num_input_tokens_seen": 12874576,
      "step": 19685
    },
    {
      "epoch": 10.319706498951781,
      "grad_norm": 2.1246933937072754,
      "learning_rate": 2.7963641578882594e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 12877232,
      "step": 19690
    },
    {
      "epoch": 10.322327044025156,
      "grad_norm": 1.0043655633926392,
      "learning_rate": 2.7952287612063193e-05,
      "loss": 0.1049,
      "num_input_tokens_seen": 12880656,
      "step": 19695
    },
    {
      "epoch": 10.324947589098532,
      "grad_norm": 2.1767468452453613,
      "learning_rate": 2.79409330276589e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 12883472,
      "step": 19700
    },
    {
      "epoch": 10.327568134171909,
      "grad_norm": 10.942880630493164,
      "learning_rate": 2.7929577828044974e-05,
      "loss": 0.1178,
      "num_input_tokens_seen": 12885776,
      "step": 19705
    },
    {
      "epoch": 10.330188679245284,
      "grad_norm": 2.213909387588501,
      "learning_rate": 2.7918222015596784e-05,
      "loss": 0.0884,
      "num_input_tokens_seen": 12888688,
      "step": 19710
    },
    {
      "epoch": 10.332809224318659,
      "grad_norm": 2.1164424419403076,
      "learning_rate": 2.7906865592689847e-05,
      "loss": 0.0942,
      "num_input_tokens_seen": 12891664,
      "step": 19715
    },
    {
      "epoch": 10.335429769392034,
      "grad_norm": 6.262276649475098,
      "learning_rate": 2.789550856169979e-05,
      "loss": 0.1011,
      "num_input_tokens_seen": 12894832,
      "step": 19720
    },
    {
      "epoch": 10.33805031446541,
      "grad_norm": 2.750121831893921,
      "learning_rate": 2.7884150925002374e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 12897744,
      "step": 19725
    },
    {
      "epoch": 10.340670859538784,
      "grad_norm": 1.61972975730896,
      "learning_rate": 2.787279268497349e-05,
      "loss": 0.1562,
      "num_input_tokens_seen": 12901008,
      "step": 19730
    },
    {
      "epoch": 10.34329140461216,
      "grad_norm": 1.682004451751709,
      "learning_rate": 2.786143384398915e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 12903792,
      "step": 19735
    },
    {
      "epoch": 10.345911949685535,
      "grad_norm": 3.99544358253479,
      "learning_rate": 2.785007440442549e-05,
      "loss": 0.0932,
      "num_input_tokens_seen": 12907120,
      "step": 19740
    },
    {
      "epoch": 10.34853249475891,
      "grad_norm": 5.090416431427002,
      "learning_rate": 2.7838714368658785e-05,
      "loss": 0.1459,
      "num_input_tokens_seen": 12910320,
      "step": 19745
    },
    {
      "epoch": 10.351153039832285,
      "grad_norm": 2.891932725906372,
      "learning_rate": 2.7827353739065427e-05,
      "loss": 0.0668,
      "num_input_tokens_seen": 12913008,
      "step": 19750
    },
    {
      "epoch": 10.35377358490566,
      "grad_norm": 2.337157726287842,
      "learning_rate": 2.781599251802191e-05,
      "loss": 0.1214,
      "num_input_tokens_seen": 12916432,
      "step": 19755
    },
    {
      "epoch": 10.356394129979035,
      "grad_norm": 1.749462366104126,
      "learning_rate": 2.780463070790489e-05,
      "loss": 0.1104,
      "num_input_tokens_seen": 12919536,
      "step": 19760
    },
    {
      "epoch": 10.35901467505241,
      "grad_norm": 1.8424261808395386,
      "learning_rate": 2.7793268311091125e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 12922768,
      "step": 19765
    },
    {
      "epoch": 10.361635220125786,
      "grad_norm": 1.90337073802948,
      "learning_rate": 2.7781905329957496e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 12925840,
      "step": 19770
    },
    {
      "epoch": 10.364255765199161,
      "grad_norm": 7.361931324005127,
      "learning_rate": 2.7770541766881008e-05,
      "loss": 0.112,
      "num_input_tokens_seen": 12928656,
      "step": 19775
    },
    {
      "epoch": 10.366876310272536,
      "grad_norm": 2.1971828937530518,
      "learning_rate": 2.7759177624238786e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 12931504,
      "step": 19780
    },
    {
      "epoch": 10.369496855345911,
      "grad_norm": 2.4421582221984863,
      "learning_rate": 2.7747812904408094e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 12935056,
      "step": 19785
    },
    {
      "epoch": 10.372117400419286,
      "grad_norm": 4.228492736816406,
      "learning_rate": 2.7736447609766286e-05,
      "loss": 0.1059,
      "num_input_tokens_seen": 12938064,
      "step": 19790
    },
    {
      "epoch": 10.374737945492662,
      "grad_norm": 2.234114408493042,
      "learning_rate": 2.772508174269085e-05,
      "loss": 0.1172,
      "num_input_tokens_seen": 12941456,
      "step": 19795
    },
    {
      "epoch": 10.377358490566039,
      "grad_norm": 4.713961601257324,
      "learning_rate": 2.7713715305559414e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 12945232,
      "step": 19800
    },
    {
      "epoch": 10.379979035639414,
      "grad_norm": 1.4807506799697876,
      "learning_rate": 2.7702348300749692e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 12947760,
      "step": 19805
    },
    {
      "epoch": 10.382599580712789,
      "grad_norm": 2.522243022918701,
      "learning_rate": 2.7690980730639533e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 12951216,
      "step": 19810
    },
    {
      "epoch": 10.385220125786164,
      "grad_norm": 2.236520528793335,
      "learning_rate": 2.7679612597606902e-05,
      "loss": 0.0893,
      "num_input_tokens_seen": 12954544,
      "step": 19815
    },
    {
      "epoch": 10.38784067085954,
      "grad_norm": 3.9830355644226074,
      "learning_rate": 2.7668243904029894e-05,
      "loss": 0.1712,
      "num_input_tokens_seen": 12959568,
      "step": 19820
    },
    {
      "epoch": 10.390461215932914,
      "grad_norm": 1.9594290256500244,
      "learning_rate": 2.76568746522867e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 12962992,
      "step": 19825
    },
    {
      "epoch": 10.39308176100629,
      "grad_norm": 3.2228174209594727,
      "learning_rate": 2.7645504844755637e-05,
      "loss": 0.1136,
      "num_input_tokens_seen": 12966608,
      "step": 19830
    },
    {
      "epoch": 10.395702306079665,
      "grad_norm": 1.4969124794006348,
      "learning_rate": 2.7634134483815143e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 12969744,
      "step": 19835
    },
    {
      "epoch": 10.39832285115304,
      "grad_norm": 3.3281357288360596,
      "learning_rate": 2.762276357184377e-05,
      "loss": 0.1061,
      "num_input_tokens_seen": 12972784,
      "step": 19840
    },
    {
      "epoch": 10.400943396226415,
      "grad_norm": 2.070056676864624,
      "learning_rate": 2.7611392111220176e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 12976944,
      "step": 19845
    },
    {
      "epoch": 10.40356394129979,
      "grad_norm": 2.8572113513946533,
      "learning_rate": 2.7600020104323154e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 12979984,
      "step": 19850
    },
    {
      "epoch": 10.406184486373165,
      "grad_norm": 9.591239929199219,
      "learning_rate": 2.7588647553531576e-05,
      "loss": 0.098,
      "num_input_tokens_seen": 12982704,
      "step": 19855
    },
    {
      "epoch": 10.40880503144654,
      "grad_norm": 3.2528233528137207,
      "learning_rate": 2.7577274461224463e-05,
      "loss": 0.0843,
      "num_input_tokens_seen": 12985680,
      "step": 19860
    },
    {
      "epoch": 10.411425576519916,
      "grad_norm": 1.8171287775039673,
      "learning_rate": 2.756590082978095e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 12989200,
      "step": 19865
    },
    {
      "epoch": 10.414046121593291,
      "grad_norm": 3.5233514308929443,
      "learning_rate": 2.7554526661580243e-05,
      "loss": 0.1084,
      "num_input_tokens_seen": 12992560,
      "step": 19870
    },
    {
      "epoch": 10.416666666666666,
      "grad_norm": 2.1400420665740967,
      "learning_rate": 2.7543151959001708e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 12995472,
      "step": 19875
    },
    {
      "epoch": 10.419287211740041,
      "grad_norm": 2.705656051635742,
      "learning_rate": 2.7531776724424797e-05,
      "loss": 0.1571,
      "num_input_tokens_seen": 12999024,
      "step": 19880
    },
    {
      "epoch": 10.421907756813416,
      "grad_norm": 2.605175733566284,
      "learning_rate": 2.752040096022907e-05,
      "loss": 0.119,
      "num_input_tokens_seen": 13001744,
      "step": 19885
    },
    {
      "epoch": 10.424528301886792,
      "grad_norm": 2.4846351146698,
      "learning_rate": 2.750902466879423e-05,
      "loss": 0.0895,
      "num_input_tokens_seen": 13004272,
      "step": 19890
    },
    {
      "epoch": 10.427148846960169,
      "grad_norm": 2.1646668910980225,
      "learning_rate": 2.7497647852500048e-05,
      "loss": 0.094,
      "num_input_tokens_seen": 13007472,
      "step": 19895
    },
    {
      "epoch": 10.429769392033544,
      "grad_norm": 3.912348985671997,
      "learning_rate": 2.7486270513726424e-05,
      "loss": 0.1138,
      "num_input_tokens_seen": 13011120,
      "step": 19900
    },
    {
      "epoch": 10.432389937106919,
      "grad_norm": 2.9955155849456787,
      "learning_rate": 2.7474892654853384e-05,
      "loss": 0.0857,
      "num_input_tokens_seen": 13013872,
      "step": 19905
    },
    {
      "epoch": 10.435010482180294,
      "grad_norm": 1.7883018255233765,
      "learning_rate": 2.7463514278261034e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 13016592,
      "step": 19910
    },
    {
      "epoch": 10.43763102725367,
      "grad_norm": 1.73722505569458,
      "learning_rate": 2.7452135386329596e-05,
      "loss": 0.0867,
      "num_input_tokens_seen": 13019888,
      "step": 19915
    },
    {
      "epoch": 10.440251572327044,
      "grad_norm": 5.330514907836914,
      "learning_rate": 2.7440755981439414e-05,
      "loss": 0.1312,
      "num_input_tokens_seen": 13023664,
      "step": 19920
    },
    {
      "epoch": 10.44287211740042,
      "grad_norm": 6.981319427490234,
      "learning_rate": 2.7429376065970924e-05,
      "loss": 0.1114,
      "num_input_tokens_seen": 13027152,
      "step": 19925
    },
    {
      "epoch": 10.445492662473795,
      "grad_norm": 2.047924518585205,
      "learning_rate": 2.7417995642304678e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 13030160,
      "step": 19930
    },
    {
      "epoch": 10.44811320754717,
      "grad_norm": 4.7119669914245605,
      "learning_rate": 2.7406614712821333e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 13034000,
      "step": 19935
    },
    {
      "epoch": 10.450733752620545,
      "grad_norm": 1.9519071578979492,
      "learning_rate": 2.739523327990164e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 13037392,
      "step": 19940
    },
    {
      "epoch": 10.45335429769392,
      "grad_norm": 4.547567844390869,
      "learning_rate": 2.7383851345926477e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 13040496,
      "step": 19945
    },
    {
      "epoch": 10.455974842767295,
      "grad_norm": 5.682966709136963,
      "learning_rate": 2.7372468913276804e-05,
      "loss": 0.1435,
      "num_input_tokens_seen": 13043248,
      "step": 19950
    },
    {
      "epoch": 10.45859538784067,
      "grad_norm": 5.696193695068359,
      "learning_rate": 2.7361085984333705e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 13046512,
      "step": 19955
    },
    {
      "epoch": 10.461215932914046,
      "grad_norm": 2.496316432952881,
      "learning_rate": 2.7349702561478356e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 13049616,
      "step": 19960
    },
    {
      "epoch": 10.463836477987421,
      "grad_norm": 1.7921370267868042,
      "learning_rate": 2.733831864709203e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 13052048,
      "step": 19965
    },
    {
      "epoch": 10.466457023060796,
      "grad_norm": 2.5434389114379883,
      "learning_rate": 2.732693424355613e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 13054096,
      "step": 19970
    },
    {
      "epoch": 10.469077568134171,
      "grad_norm": 1.4763884544372559,
      "learning_rate": 2.731554935325213e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 13057488,
      "step": 19975
    },
    {
      "epoch": 10.471698113207546,
      "grad_norm": 1.4916741847991943,
      "learning_rate": 2.7304163978561613e-05,
      "loss": 0.0757,
      "num_input_tokens_seen": 13060336,
      "step": 19980
    },
    {
      "epoch": 10.474318658280922,
      "grad_norm": 1.6347203254699707,
      "learning_rate": 2.729277812186629e-05,
      "loss": 0.1144,
      "num_input_tokens_seen": 13063728,
      "step": 19985
    },
    {
      "epoch": 10.476939203354299,
      "grad_norm": 2.546656370162964,
      "learning_rate": 2.7281391785547945e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 13066256,
      "step": 19990
    },
    {
      "epoch": 10.479559748427674,
      "grad_norm": 3.717703342437744,
      "learning_rate": 2.7270004971988456e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 13069424,
      "step": 19995
    },
    {
      "epoch": 10.482180293501049,
      "grad_norm": 1.250247836112976,
      "learning_rate": 2.7258617683569825e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 13072560,
      "step": 20000
    },
    {
      "epoch": 10.484800838574424,
      "grad_norm": 1.6533101797103882,
      "learning_rate": 2.7247229922674143e-05,
      "loss": 0.1097,
      "num_input_tokens_seen": 13076176,
      "step": 20005
    },
    {
      "epoch": 10.4874213836478,
      "grad_norm": 2.746234655380249,
      "learning_rate": 2.7235841691683605e-05,
      "loss": 0.0988,
      "num_input_tokens_seen": 13079984,
      "step": 20010
    },
    {
      "epoch": 10.490041928721174,
      "grad_norm": 2.429727554321289,
      "learning_rate": 2.7224452992980487e-05,
      "loss": 0.0677,
      "num_input_tokens_seen": 13082928,
      "step": 20015
    },
    {
      "epoch": 10.49266247379455,
      "grad_norm": 2.2475733757019043,
      "learning_rate": 2.7213063828947182e-05,
      "loss": 0.1116,
      "num_input_tokens_seen": 13085776,
      "step": 20020
    },
    {
      "epoch": 10.495283018867925,
      "grad_norm": 4.344204902648926,
      "learning_rate": 2.7201674201966176e-05,
      "loss": 0.1028,
      "num_input_tokens_seen": 13088656,
      "step": 20025
    },
    {
      "epoch": 10.4979035639413,
      "grad_norm": 2.760237455368042,
      "learning_rate": 2.7190284114420035e-05,
      "loss": 0.119,
      "num_input_tokens_seen": 13091728,
      "step": 20030
    },
    {
      "epoch": 10.500524109014675,
      "grad_norm": 2.081613063812256,
      "learning_rate": 2.717889356869146e-05,
      "loss": 0.1402,
      "num_input_tokens_seen": 13096048,
      "step": 20035
    },
    {
      "epoch": 10.50314465408805,
      "grad_norm": 2.897740602493286,
      "learning_rate": 2.7167502567163194e-05,
      "loss": 0.1736,
      "num_input_tokens_seen": 13098608,
      "step": 20040
    },
    {
      "epoch": 10.505765199161425,
      "grad_norm": 4.47283935546875,
      "learning_rate": 2.715611111221813e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 13101872,
      "step": 20045
    },
    {
      "epoch": 10.5083857442348,
      "grad_norm": 1.39231276512146,
      "learning_rate": 2.7144719206239217e-05,
      "loss": 0.1253,
      "num_input_tokens_seen": 13109072,
      "step": 20050
    },
    {
      "epoch": 10.511006289308176,
      "grad_norm": 2.609469175338745,
      "learning_rate": 2.7133326851609508e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 13112080,
      "step": 20055
    },
    {
      "epoch": 10.51362683438155,
      "grad_norm": 3.5761306285858154,
      "learning_rate": 2.7121934050712166e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 13114704,
      "step": 20060
    },
    {
      "epoch": 10.516247379454926,
      "grad_norm": 1.319385290145874,
      "learning_rate": 2.711054080593043e-05,
      "loss": 0.0894,
      "num_input_tokens_seen": 13117360,
      "step": 20065
    },
    {
      "epoch": 10.518867924528301,
      "grad_norm": 1.091490387916565,
      "learning_rate": 2.7099147119647627e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 13121104,
      "step": 20070
    },
    {
      "epoch": 10.521488469601676,
      "grad_norm": 1.4701822996139526,
      "learning_rate": 2.7087752994247206e-05,
      "loss": 0.1016,
      "num_input_tokens_seen": 13124464,
      "step": 20075
    },
    {
      "epoch": 10.524109014675052,
      "grad_norm": 1.8448139429092407,
      "learning_rate": 2.707635843211267e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 13127024,
      "step": 20080
    },
    {
      "epoch": 10.526729559748428,
      "grad_norm": 3.533104419708252,
      "learning_rate": 2.7064963435627637e-05,
      "loss": 0.0973,
      "num_input_tokens_seen": 13129936,
      "step": 20085
    },
    {
      "epoch": 10.529350104821804,
      "grad_norm": 2.217118978500366,
      "learning_rate": 2.7053568007175823e-05,
      "loss": 0.1331,
      "num_input_tokens_seen": 13132656,
      "step": 20090
    },
    {
      "epoch": 10.531970649895179,
      "grad_norm": 2.048823356628418,
      "learning_rate": 2.7042172149140994e-05,
      "loss": 0.1417,
      "num_input_tokens_seen": 13136528,
      "step": 20095
    },
    {
      "epoch": 10.534591194968554,
      "grad_norm": 1.7019836902618408,
      "learning_rate": 2.703077586390706e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 13139568,
      "step": 20100
    },
    {
      "epoch": 10.53721174004193,
      "grad_norm": 1.6155120134353638,
      "learning_rate": 2.7019379153857982e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 13142704,
      "step": 20105
    },
    {
      "epoch": 10.539832285115304,
      "grad_norm": 2.0247535705566406,
      "learning_rate": 2.700798202137782e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 13145936,
      "step": 20110
    },
    {
      "epoch": 10.54245283018868,
      "grad_norm": 4.235253810882568,
      "learning_rate": 2.6996584468850732e-05,
      "loss": 0.1388,
      "num_input_tokens_seen": 13149360,
      "step": 20115
    },
    {
      "epoch": 10.545073375262055,
      "grad_norm": 1.87954843044281,
      "learning_rate": 2.6985186498660946e-05,
      "loss": 0.1041,
      "num_input_tokens_seen": 13151920,
      "step": 20120
    },
    {
      "epoch": 10.54769392033543,
      "grad_norm": 2.0506324768066406,
      "learning_rate": 2.69737881131928e-05,
      "loss": 0.1314,
      "num_input_tokens_seen": 13155024,
      "step": 20125
    },
    {
      "epoch": 10.550314465408805,
      "grad_norm": 1.5846354961395264,
      "learning_rate": 2.6962389314830692e-05,
      "loss": 0.1333,
      "num_input_tokens_seen": 13158320,
      "step": 20130
    },
    {
      "epoch": 10.55293501048218,
      "grad_norm": 2.2579030990600586,
      "learning_rate": 2.6950990105959135e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 13161040,
      "step": 20135
    },
    {
      "epoch": 10.555555555555555,
      "grad_norm": 3.607776641845703,
      "learning_rate": 2.6939590488962703e-05,
      "loss": 0.0904,
      "num_input_tokens_seen": 13163984,
      "step": 20140
    },
    {
      "epoch": 10.55817610062893,
      "grad_norm": 5.613548278808594,
      "learning_rate": 2.6928190466226067e-05,
      "loss": 0.138,
      "num_input_tokens_seen": 13167888,
      "step": 20145
    },
    {
      "epoch": 10.560796645702306,
      "grad_norm": 10.675152778625488,
      "learning_rate": 2.691679004013398e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 13171312,
      "step": 20150
    },
    {
      "epoch": 10.56341719077568,
      "grad_norm": 2.341482162475586,
      "learning_rate": 2.6905389213071285e-05,
      "loss": 0.0897,
      "num_input_tokens_seen": 13174192,
      "step": 20155
    },
    {
      "epoch": 10.566037735849056,
      "grad_norm": 1.7046278715133667,
      "learning_rate": 2.6893987987422907e-05,
      "loss": 0.1445,
      "num_input_tokens_seen": 13178736,
      "step": 20160
    },
    {
      "epoch": 10.568658280922431,
      "grad_norm": 2.4851207733154297,
      "learning_rate": 2.6882586365573854e-05,
      "loss": 0.0879,
      "num_input_tokens_seen": 13181392,
      "step": 20165
    },
    {
      "epoch": 10.571278825995806,
      "grad_norm": 2.5494866371154785,
      "learning_rate": 2.6871184349909206e-05,
      "loss": 0.1193,
      "num_input_tokens_seen": 13184496,
      "step": 20170
    },
    {
      "epoch": 10.573899371069182,
      "grad_norm": 1.8178236484527588,
      "learning_rate": 2.685978194281414e-05,
      "loss": 0.1665,
      "num_input_tokens_seen": 13187760,
      "step": 20175
    },
    {
      "epoch": 10.576519916142558,
      "grad_norm": 5.472928524017334,
      "learning_rate": 2.6848379146673914e-05,
      "loss": 0.0987,
      "num_input_tokens_seen": 13190640,
      "step": 20180
    },
    {
      "epoch": 10.579140461215934,
      "grad_norm": 1.6823463439941406,
      "learning_rate": 2.6836975963873843e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 13193840,
      "step": 20185
    },
    {
      "epoch": 10.581761006289309,
      "grad_norm": 3.5296270847320557,
      "learning_rate": 2.6825572396799377e-05,
      "loss": 0.1067,
      "num_input_tokens_seen": 13197648,
      "step": 20190
    },
    {
      "epoch": 10.584381551362684,
      "grad_norm": 1.6042431592941284,
      "learning_rate": 2.6814168447835986e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 13201808,
      "step": 20195
    },
    {
      "epoch": 10.58700209643606,
      "grad_norm": 2.8797447681427,
      "learning_rate": 2.6802764119369245e-05,
      "loss": 0.1845,
      "num_input_tokens_seen": 13204976,
      "step": 20200
    },
    {
      "epoch": 10.589622641509434,
      "grad_norm": 4.083111763000488,
      "learning_rate": 2.6791359413784828e-05,
      "loss": 0.1274,
      "num_input_tokens_seen": 13207856,
      "step": 20205
    },
    {
      "epoch": 10.59224318658281,
      "grad_norm": 4.778023719787598,
      "learning_rate": 2.677995433346845e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 13210768,
      "step": 20210
    },
    {
      "epoch": 10.594863731656185,
      "grad_norm": 4.188339710235596,
      "learning_rate": 2.6768548880805934e-05,
      "loss": 0.1334,
      "num_input_tokens_seen": 13213776,
      "step": 20215
    },
    {
      "epoch": 10.59748427672956,
      "grad_norm": 1.4742542505264282,
      "learning_rate": 2.6757143058183177e-05,
      "loss": 0.2043,
      "num_input_tokens_seen": 13217104,
      "step": 20220
    },
    {
      "epoch": 10.600104821802935,
      "grad_norm": 2.397953748703003,
      "learning_rate": 2.6745736867986132e-05,
      "loss": 0.0985,
      "num_input_tokens_seen": 13220048,
      "step": 20225
    },
    {
      "epoch": 10.60272536687631,
      "grad_norm": 3.9614298343658447,
      "learning_rate": 2.673433031260085e-05,
      "loss": 0.1218,
      "num_input_tokens_seen": 13223376,
      "step": 20230
    },
    {
      "epoch": 10.605345911949685,
      "grad_norm": 3.360318183898926,
      "learning_rate": 2.6722923394413456e-05,
      "loss": 0.157,
      "num_input_tokens_seen": 13226512,
      "step": 20235
    },
    {
      "epoch": 10.60796645702306,
      "grad_norm": 1.4806978702545166,
      "learning_rate": 2.6711516115810136e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 13229744,
      "step": 20240
    },
    {
      "epoch": 10.610587002096436,
      "grad_norm": 2.673654794692993,
      "learning_rate": 2.6700108479177188e-05,
      "loss": 0.1286,
      "num_input_tokens_seen": 13233552,
      "step": 20245
    },
    {
      "epoch": 10.61320754716981,
      "grad_norm": 3.2405552864074707,
      "learning_rate": 2.6688700486900926e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 13237008,
      "step": 20250
    },
    {
      "epoch": 10.615828092243186,
      "grad_norm": 5.132383346557617,
      "learning_rate": 2.6677292141367787e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 13240080,
      "step": 20255
    },
    {
      "epoch": 10.618448637316561,
      "grad_norm": 5.080284118652344,
      "learning_rate": 2.6665883444964278e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 13243408,
      "step": 20260
    },
    {
      "epoch": 10.621069182389936,
      "grad_norm": 2.421762466430664,
      "learning_rate": 2.665447440007695e-05,
      "loss": 0.1128,
      "num_input_tokens_seen": 13247152,
      "step": 20265
    },
    {
      "epoch": 10.623689727463312,
      "grad_norm": 1.6186466217041016,
      "learning_rate": 2.664306500909246e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 13251376,
      "step": 20270
    },
    {
      "epoch": 10.626310272536688,
      "grad_norm": 1.9012529850006104,
      "learning_rate": 2.663165527439751e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 13254992,
      "step": 20275
    },
    {
      "epoch": 10.628930817610064,
      "grad_norm": 3.945425510406494,
      "learning_rate": 2.662024519837889e-05,
      "loss": 0.0837,
      "num_input_tokens_seen": 13257712,
      "step": 20280
    },
    {
      "epoch": 10.631551362683439,
      "grad_norm": 1.8704878091812134,
      "learning_rate": 2.6608834783423465e-05,
      "loss": 0.0761,
      "num_input_tokens_seen": 13262192,
      "step": 20285
    },
    {
      "epoch": 10.634171907756814,
      "grad_norm": 2.118913173675537,
      "learning_rate": 2.6597424031918156e-05,
      "loss": 0.1174,
      "num_input_tokens_seen": 13266832,
      "step": 20290
    },
    {
      "epoch": 10.63679245283019,
      "grad_norm": 2.277348279953003,
      "learning_rate": 2.6586012946249954e-05,
      "loss": 0.0786,
      "num_input_tokens_seen": 13269424,
      "step": 20295
    },
    {
      "epoch": 10.639412997903564,
      "grad_norm": 2.3617193698883057,
      "learning_rate": 2.6574601528805953e-05,
      "loss": 0.1151,
      "num_input_tokens_seen": 13272144,
      "step": 20300
    },
    {
      "epoch": 10.64203354297694,
      "grad_norm": 1.956504464149475,
      "learning_rate": 2.6563189781973268e-05,
      "loss": 0.071,
      "num_input_tokens_seen": 13275376,
      "step": 20305
    },
    {
      "epoch": 10.644654088050315,
      "grad_norm": 2.593306064605713,
      "learning_rate": 2.655177770813912e-05,
      "loss": 0.1092,
      "num_input_tokens_seen": 13278320,
      "step": 20310
    },
    {
      "epoch": 10.64727463312369,
      "grad_norm": 3.385564088821411,
      "learning_rate": 2.654036530969077e-05,
      "loss": 0.1191,
      "num_input_tokens_seen": 13280816,
      "step": 20315
    },
    {
      "epoch": 10.649895178197065,
      "grad_norm": 7.940119743347168,
      "learning_rate": 2.6528952589015576e-05,
      "loss": 0.2262,
      "num_input_tokens_seen": 13284784,
      "step": 20320
    },
    {
      "epoch": 10.65251572327044,
      "grad_norm": 2.079033613204956,
      "learning_rate": 2.6517539548500947e-05,
      "loss": 0.1403,
      "num_input_tokens_seen": 13287376,
      "step": 20325
    },
    {
      "epoch": 10.655136268343815,
      "grad_norm": 7.805866718292236,
      "learning_rate": 2.6506126190534354e-05,
      "loss": 0.0817,
      "num_input_tokens_seen": 13290928,
      "step": 20330
    },
    {
      "epoch": 10.65775681341719,
      "grad_norm": 1.7644799947738647,
      "learning_rate": 2.649471251750334e-05,
      "loss": 0.1086,
      "num_input_tokens_seen": 13294192,
      "step": 20335
    },
    {
      "epoch": 10.660377358490566,
      "grad_norm": 2.6793806552886963,
      "learning_rate": 2.6483298531795527e-05,
      "loss": 0.1344,
      "num_input_tokens_seen": 13297456,
      "step": 20340
    },
    {
      "epoch": 10.66299790356394,
      "grad_norm": 1.467166543006897,
      "learning_rate": 2.6471884235798583e-05,
      "loss": 0.1439,
      "num_input_tokens_seen": 13301008,
      "step": 20345
    },
    {
      "epoch": 10.665618448637316,
      "grad_norm": 1.8921204805374146,
      "learning_rate": 2.6460469631900247e-05,
      "loss": 0.1099,
      "num_input_tokens_seen": 13304368,
      "step": 20350
    },
    {
      "epoch": 10.668238993710691,
      "grad_norm": 2.782087802886963,
      "learning_rate": 2.644905472248832e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 13306864,
      "step": 20355
    },
    {
      "epoch": 10.670859538784066,
      "grad_norm": 3.30806827545166,
      "learning_rate": 2.6437639509950686e-05,
      "loss": 0.0811,
      "num_input_tokens_seen": 13310352,
      "step": 20360
    },
    {
      "epoch": 10.673480083857442,
      "grad_norm": 2.8493196964263916,
      "learning_rate": 2.6426223996675264e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 13313488,
      "step": 20365
    },
    {
      "epoch": 10.676100628930818,
      "grad_norm": 3.8480029106140137,
      "learning_rate": 2.641480818505005e-05,
      "loss": 0.1108,
      "num_input_tokens_seen": 13316112,
      "step": 20370
    },
    {
      "epoch": 10.678721174004194,
      "grad_norm": 1.183100700378418,
      "learning_rate": 2.6403392077463108e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 13320528,
      "step": 20375
    },
    {
      "epoch": 10.681341719077569,
      "grad_norm": 4.821445941925049,
      "learning_rate": 2.639197567630256e-05,
      "loss": 0.0716,
      "num_input_tokens_seen": 13323888,
      "step": 20380
    },
    {
      "epoch": 10.683962264150944,
      "grad_norm": 3.2458012104034424,
      "learning_rate": 2.638055898395657e-05,
      "loss": 0.1524,
      "num_input_tokens_seen": 13327568,
      "step": 20385
    },
    {
      "epoch": 10.68658280922432,
      "grad_norm": 3.8029606342315674,
      "learning_rate": 2.6369142002813403e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 13330096,
      "step": 20390
    },
    {
      "epoch": 10.689203354297694,
      "grad_norm": 1.6747891902923584,
      "learning_rate": 2.6357724735261335e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 13332560,
      "step": 20395
    },
    {
      "epoch": 10.69182389937107,
      "grad_norm": 1.8404179811477661,
      "learning_rate": 2.6346307183688752e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 13336080,
      "step": 20400
    },
    {
      "epoch": 10.694444444444445,
      "grad_norm": 2.556652784347534,
      "learning_rate": 2.6334889350484064e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 13339696,
      "step": 20405
    },
    {
      "epoch": 10.69706498951782,
      "grad_norm": 1.5040465593338013,
      "learning_rate": 2.632347123803575e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 13343760,
      "step": 20410
    },
    {
      "epoch": 10.699685534591195,
      "grad_norm": 2.4972035884857178,
      "learning_rate": 2.6312052848732353e-05,
      "loss": 0.1146,
      "num_input_tokens_seen": 13346928,
      "step": 20415
    },
    {
      "epoch": 10.70230607966457,
      "grad_norm": 2.662614583969116,
      "learning_rate": 2.6300634184962475e-05,
      "loss": 0.1161,
      "num_input_tokens_seen": 13349968,
      "step": 20420
    },
    {
      "epoch": 10.704926624737945,
      "grad_norm": 1.8899883031845093,
      "learning_rate": 2.628921524911475e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 13353232,
      "step": 20425
    },
    {
      "epoch": 10.70754716981132,
      "grad_norm": 5.497635364532471,
      "learning_rate": 2.6277796043577924e-05,
      "loss": 0.1353,
      "num_input_tokens_seen": 13356496,
      "step": 20430
    },
    {
      "epoch": 10.710167714884696,
      "grad_norm": 1.2202731370925903,
      "learning_rate": 2.6266376570740725e-05,
      "loss": 0.1073,
      "num_input_tokens_seen": 13360656,
      "step": 20435
    },
    {
      "epoch": 10.71278825995807,
      "grad_norm": 2.558917284011841,
      "learning_rate": 2.6254956832992005e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 13363984,
      "step": 20440
    },
    {
      "epoch": 10.715408805031446,
      "grad_norm": 4.61099910736084,
      "learning_rate": 2.6243536832720627e-05,
      "loss": 0.1227,
      "num_input_tokens_seen": 13367664,
      "step": 20445
    },
    {
      "epoch": 10.718029350104821,
      "grad_norm": 2.7794015407562256,
      "learning_rate": 2.623211657231553e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 13369744,
      "step": 20450
    },
    {
      "epoch": 10.720649895178196,
      "grad_norm": 2.957653045654297,
      "learning_rate": 2.6220696054165704e-05,
      "loss": 0.1002,
      "num_input_tokens_seen": 13373584,
      "step": 20455
    },
    {
      "epoch": 10.723270440251572,
      "grad_norm": 1.9138013124465942,
      "learning_rate": 2.6209275280660194e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 13377232,
      "step": 20460
    },
    {
      "epoch": 10.725890985324948,
      "grad_norm": 3.5633010864257812,
      "learning_rate": 2.6197854254188085e-05,
      "loss": 0.1879,
      "num_input_tokens_seen": 13380944,
      "step": 20465
    },
    {
      "epoch": 10.728511530398324,
      "grad_norm": 2.5036020278930664,
      "learning_rate": 2.618643297713854e-05,
      "loss": 0.1021,
      "num_input_tokens_seen": 13385072,
      "step": 20470
    },
    {
      "epoch": 10.731132075471699,
      "grad_norm": 2.6710615158081055,
      "learning_rate": 2.617501145190075e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 13388016,
      "step": 20475
    },
    {
      "epoch": 10.733752620545074,
      "grad_norm": 2.0380520820617676,
      "learning_rate": 2.616358968086397e-05,
      "loss": 0.128,
      "num_input_tokens_seen": 13391760,
      "step": 20480
    },
    {
      "epoch": 10.73637316561845,
      "grad_norm": 2.4853310585021973,
      "learning_rate": 2.6152167666417493e-05,
      "loss": 0.1627,
      "num_input_tokens_seen": 13396144,
      "step": 20485
    },
    {
      "epoch": 10.738993710691824,
      "grad_norm": 3.8640804290771484,
      "learning_rate": 2.6140745410950694e-05,
      "loss": 0.0621,
      "num_input_tokens_seen": 13399824,
      "step": 20490
    },
    {
      "epoch": 10.7416142557652,
      "grad_norm": 0.9000827670097351,
      "learning_rate": 2.6129322916852973e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 13403248,
      "step": 20495
    },
    {
      "epoch": 10.744234800838575,
      "grad_norm": 2.609245538711548,
      "learning_rate": 2.611790018651377e-05,
      "loss": 0.0601,
      "num_input_tokens_seen": 13405904,
      "step": 20500
    },
    {
      "epoch": 10.74685534591195,
      "grad_norm": 3.210604190826416,
      "learning_rate": 2.6106477222322613e-05,
      "loss": 0.0743,
      "num_input_tokens_seen": 13409360,
      "step": 20505
    },
    {
      "epoch": 10.749475890985325,
      "grad_norm": 2.5390241146087646,
      "learning_rate": 2.6095054026669037e-05,
      "loss": 0.0936,
      "num_input_tokens_seen": 13412304,
      "step": 20510
    },
    {
      "epoch": 10.7520964360587,
      "grad_norm": 2.0091092586517334,
      "learning_rate": 2.608363060194265e-05,
      "loss": 0.1181,
      "num_input_tokens_seen": 13415568,
      "step": 20515
    },
    {
      "epoch": 10.754716981132075,
      "grad_norm": 3.0413429737091064,
      "learning_rate": 2.6072206950533117e-05,
      "loss": 0.1109,
      "num_input_tokens_seen": 13418160,
      "step": 20520
    },
    {
      "epoch": 10.75733752620545,
      "grad_norm": 2.2940752506256104,
      "learning_rate": 2.606078307483011e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 13421168,
      "step": 20525
    },
    {
      "epoch": 10.759958071278826,
      "grad_norm": 3.2382357120513916,
      "learning_rate": 2.6049358977223392e-05,
      "loss": 0.0989,
      "num_input_tokens_seen": 13423824,
      "step": 20530
    },
    {
      "epoch": 10.7625786163522,
      "grad_norm": 3.5332157611846924,
      "learning_rate": 2.6037934660102747e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 13426928,
      "step": 20535
    },
    {
      "epoch": 10.765199161425576,
      "grad_norm": 1.7884238958358765,
      "learning_rate": 2.6026510125858006e-05,
      "loss": 0.0858,
      "num_input_tokens_seen": 13429712,
      "step": 20540
    },
    {
      "epoch": 10.767819706498951,
      "grad_norm": 1.669400691986084,
      "learning_rate": 2.6015085376879063e-05,
      "loss": 0.0766,
      "num_input_tokens_seen": 13433648,
      "step": 20545
    },
    {
      "epoch": 10.770440251572326,
      "grad_norm": 1.99573814868927,
      "learning_rate": 2.600366041555584e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 13436048,
      "step": 20550
    },
    {
      "epoch": 10.773060796645701,
      "grad_norm": 2.466878652572632,
      "learning_rate": 2.5992235244278302e-05,
      "loss": 0.1378,
      "num_input_tokens_seen": 13438672,
      "step": 20555
    },
    {
      "epoch": 10.775681341719078,
      "grad_norm": 2.141594886779785,
      "learning_rate": 2.5980809865436485e-05,
      "loss": 0.0946,
      "num_input_tokens_seen": 13440976,
      "step": 20560
    },
    {
      "epoch": 10.778301886792454,
      "grad_norm": 1.1368002891540527,
      "learning_rate": 2.5969384281420424e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 13444528,
      "step": 20565
    },
    {
      "epoch": 10.780922431865829,
      "grad_norm": 6.170438289642334,
      "learning_rate": 2.5957958494620227e-05,
      "loss": 0.1098,
      "num_input_tokens_seen": 13447664,
      "step": 20570
    },
    {
      "epoch": 10.783542976939204,
      "grad_norm": 2.454846143722534,
      "learning_rate": 2.594653250742605e-05,
      "loss": 0.1064,
      "num_input_tokens_seen": 13451312,
      "step": 20575
    },
    {
      "epoch": 10.786163522012579,
      "grad_norm": 2.369144916534424,
      "learning_rate": 2.5935106322228065e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 13453776,
      "step": 20580
    },
    {
      "epoch": 10.788784067085954,
      "grad_norm": 3.1425182819366455,
      "learning_rate": 2.5923679941416514e-05,
      "loss": 0.0896,
      "num_input_tokens_seen": 13456656,
      "step": 20585
    },
    {
      "epoch": 10.79140461215933,
      "grad_norm": 1.9730761051177979,
      "learning_rate": 2.591225336738165e-05,
      "loss": 0.0671,
      "num_input_tokens_seen": 13459440,
      "step": 20590
    },
    {
      "epoch": 10.794025157232705,
      "grad_norm": 5.243406295776367,
      "learning_rate": 2.5900826602513782e-05,
      "loss": 0.1513,
      "num_input_tokens_seen": 13462608,
      "step": 20595
    },
    {
      "epoch": 10.79664570230608,
      "grad_norm": 1.5392050743103027,
      "learning_rate": 2.5889399649203277e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 13465424,
      "step": 20600
    },
    {
      "epoch": 10.799266247379455,
      "grad_norm": 3.1422574520111084,
      "learning_rate": 2.5877972509840514e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 13468688,
      "step": 20605
    },
    {
      "epoch": 10.80188679245283,
      "grad_norm": 1.812524437904358,
      "learning_rate": 2.5866545186815912e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 13471312,
      "step": 20610
    },
    {
      "epoch": 10.804507337526205,
      "grad_norm": 4.55320930480957,
      "learning_rate": 2.5855117682519954e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 13474352,
      "step": 20615
    },
    {
      "epoch": 10.80712788259958,
      "grad_norm": 7.5690083503723145,
      "learning_rate": 2.584368999934312e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 13478704,
      "step": 20620
    },
    {
      "epoch": 10.809748427672956,
      "grad_norm": 2.692603349685669,
      "learning_rate": 2.5832262139675966e-05,
      "loss": 0.139,
      "num_input_tokens_seen": 13481904,
      "step": 20625
    },
    {
      "epoch": 10.81236897274633,
      "grad_norm": 1.634232759475708,
      "learning_rate": 2.582083410590907e-05,
      "loss": 0.2028,
      "num_input_tokens_seen": 13484624,
      "step": 20630
    },
    {
      "epoch": 10.814989517819706,
      "grad_norm": 2.351433277130127,
      "learning_rate": 2.5809405900433036e-05,
      "loss": 0.074,
      "num_input_tokens_seen": 13488368,
      "step": 20635
    },
    {
      "epoch": 10.817610062893081,
      "grad_norm": 2.584157943725586,
      "learning_rate": 2.579797752563853e-05,
      "loss": 0.0907,
      "num_input_tokens_seen": 13490672,
      "step": 20640
    },
    {
      "epoch": 10.820230607966456,
      "grad_norm": 0.958092212677002,
      "learning_rate": 2.5786548983916232e-05,
      "loss": 0.0955,
      "num_input_tokens_seen": 13495984,
      "step": 20645
    },
    {
      "epoch": 10.822851153039831,
      "grad_norm": 9.191693305969238,
      "learning_rate": 2.577512027765685e-05,
      "loss": 0.0691,
      "num_input_tokens_seen": 13498416,
      "step": 20650
    },
    {
      "epoch": 10.825471698113208,
      "grad_norm": 1.9647798538208008,
      "learning_rate": 2.5763691409251155e-05,
      "loss": 0.1302,
      "num_input_tokens_seen": 13502512,
      "step": 20655
    },
    {
      "epoch": 10.828092243186584,
      "grad_norm": 1.8987363576889038,
      "learning_rate": 2.5752262381089932e-05,
      "loss": 0.1062,
      "num_input_tokens_seen": 13506416,
      "step": 20660
    },
    {
      "epoch": 10.830712788259959,
      "grad_norm": 2.390625476837158,
      "learning_rate": 2.5740833195563996e-05,
      "loss": 0.1791,
      "num_input_tokens_seen": 13509392,
      "step": 20665
    },
    {
      "epoch": 10.833333333333334,
      "grad_norm": 3.73569655418396,
      "learning_rate": 2.572940385506421e-05,
      "loss": 0.122,
      "num_input_tokens_seen": 13514032,
      "step": 20670
    },
    {
      "epoch": 10.835953878406709,
      "grad_norm": 3.0924341678619385,
      "learning_rate": 2.5717974361981457e-05,
      "loss": 0.1135,
      "num_input_tokens_seen": 13517968,
      "step": 20675
    },
    {
      "epoch": 10.838574423480084,
      "grad_norm": 1.416072130203247,
      "learning_rate": 2.5706544718706665e-05,
      "loss": 0.1129,
      "num_input_tokens_seen": 13521616,
      "step": 20680
    },
    {
      "epoch": 10.84119496855346,
      "grad_norm": 3.1157801151275635,
      "learning_rate": 2.569511492763077e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 13524592,
      "step": 20685
    },
    {
      "epoch": 10.843815513626835,
      "grad_norm": 7.909873008728027,
      "learning_rate": 2.5683684991144764e-05,
      "loss": 0.1215,
      "num_input_tokens_seen": 13528048,
      "step": 20690
    },
    {
      "epoch": 10.84643605870021,
      "grad_norm": 2.3604190349578857,
      "learning_rate": 2.5672254911639666e-05,
      "loss": 0.1032,
      "num_input_tokens_seen": 13531120,
      "step": 20695
    },
    {
      "epoch": 10.849056603773585,
      "grad_norm": 1.9138259887695312,
      "learning_rate": 2.5660824691506496e-05,
      "loss": 0.1202,
      "num_input_tokens_seen": 13534288,
      "step": 20700
    },
    {
      "epoch": 10.85167714884696,
      "grad_norm": 4.207533836364746,
      "learning_rate": 2.5649394333136356e-05,
      "loss": 0.2127,
      "num_input_tokens_seen": 13538032,
      "step": 20705
    },
    {
      "epoch": 10.854297693920335,
      "grad_norm": 8.556184768676758,
      "learning_rate": 2.5637963838920316e-05,
      "loss": 0.1226,
      "num_input_tokens_seen": 13541008,
      "step": 20710
    },
    {
      "epoch": 10.85691823899371,
      "grad_norm": 3.348630428314209,
      "learning_rate": 2.5626533211249525e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 13544304,
      "step": 20715
    },
    {
      "epoch": 10.859538784067086,
      "grad_norm": 0.9014471173286438,
      "learning_rate": 2.5615102452515138e-05,
      "loss": 0.1685,
      "num_input_tokens_seen": 13548048,
      "step": 20720
    },
    {
      "epoch": 10.86215932914046,
      "grad_norm": 1.4726145267486572,
      "learning_rate": 2.5603671565108328e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 13554288,
      "step": 20725
    },
    {
      "epoch": 10.864779874213836,
      "grad_norm": 2.5097479820251465,
      "learning_rate": 2.5592240551420317e-05,
      "loss": 0.1198,
      "num_input_tokens_seen": 13558320,
      "step": 20730
    },
    {
      "epoch": 10.867400419287211,
      "grad_norm": 1.8041367530822754,
      "learning_rate": 2.5580809413842343e-05,
      "loss": 0.1227,
      "num_input_tokens_seen": 13561328,
      "step": 20735
    },
    {
      "epoch": 10.870020964360586,
      "grad_norm": 2.3423900604248047,
      "learning_rate": 2.5569378154765655e-05,
      "loss": 0.1538,
      "num_input_tokens_seen": 13566224,
      "step": 20740
    },
    {
      "epoch": 10.872641509433961,
      "grad_norm": 1.9686973094940186,
      "learning_rate": 2.5557946776581572e-05,
      "loss": 0.2048,
      "num_input_tokens_seen": 13568752,
      "step": 20745
    },
    {
      "epoch": 10.875262054507338,
      "grad_norm": 2.8608715534210205,
      "learning_rate": 2.554651528168137e-05,
      "loss": 0.1031,
      "num_input_tokens_seen": 13571664,
      "step": 20750
    },
    {
      "epoch": 10.877882599580714,
      "grad_norm": 1.7187527418136597,
      "learning_rate": 2.5535083672456407e-05,
      "loss": 0.1306,
      "num_input_tokens_seen": 13574704,
      "step": 20755
    },
    {
      "epoch": 10.880503144654089,
      "grad_norm": 5.330021381378174,
      "learning_rate": 2.5523651951298055e-05,
      "loss": 0.1364,
      "num_input_tokens_seen": 13577264,
      "step": 20760
    },
    {
      "epoch": 10.883123689727464,
      "grad_norm": 3.928757905960083,
      "learning_rate": 2.5512220120597686e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 13580240,
      "step": 20765
    },
    {
      "epoch": 10.885744234800839,
      "grad_norm": 3.294156789779663,
      "learning_rate": 2.550078818274671e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 13583472,
      "step": 20770
    },
    {
      "epoch": 10.888364779874214,
      "grad_norm": 3.4119668006896973,
      "learning_rate": 2.548935614013656e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 13586640,
      "step": 20775
    },
    {
      "epoch": 10.89098532494759,
      "grad_norm": 1.9026596546173096,
      "learning_rate": 2.5477923995158683e-05,
      "loss": 0.07,
      "num_input_tokens_seen": 13589680,
      "step": 20780
    },
    {
      "epoch": 10.893605870020965,
      "grad_norm": 1.9570353031158447,
      "learning_rate": 2.5466491750204562e-05,
      "loss": 0.1453,
      "num_input_tokens_seen": 13593072,
      "step": 20785
    },
    {
      "epoch": 10.89622641509434,
      "grad_norm": 2.0782418251037598,
      "learning_rate": 2.5455059407665693e-05,
      "loss": 0.1138,
      "num_input_tokens_seen": 13596624,
      "step": 20790
    },
    {
      "epoch": 10.898846960167715,
      "grad_norm": 1.915497899055481,
      "learning_rate": 2.544362696993358e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 13599088,
      "step": 20795
    },
    {
      "epoch": 10.90146750524109,
      "grad_norm": 5.337425708770752,
      "learning_rate": 2.5432194439399775e-05,
      "loss": 0.0751,
      "num_input_tokens_seen": 13601552,
      "step": 20800
    },
    {
      "epoch": 10.904088050314465,
      "grad_norm": 2.4741973876953125,
      "learning_rate": 2.542076181845582e-05,
      "loss": 0.0776,
      "num_input_tokens_seen": 13604144,
      "step": 20805
    },
    {
      "epoch": 10.90670859538784,
      "grad_norm": 1.225643277168274,
      "learning_rate": 2.540932910949329e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 13606928,
      "step": 20810
    },
    {
      "epoch": 10.909329140461216,
      "grad_norm": 2.957714796066284,
      "learning_rate": 2.5397896314903785e-05,
      "loss": 0.0728,
      "num_input_tokens_seen": 13610320,
      "step": 20815
    },
    {
      "epoch": 10.91194968553459,
      "grad_norm": 1.4653693437576294,
      "learning_rate": 2.5386463437078905e-05,
      "loss": 0.2891,
      "num_input_tokens_seen": 13613680,
      "step": 20820
    },
    {
      "epoch": 10.914570230607966,
      "grad_norm": 6.949742794036865,
      "learning_rate": 2.5375030478410282e-05,
      "loss": 0.0971,
      "num_input_tokens_seen": 13617264,
      "step": 20825
    },
    {
      "epoch": 10.917190775681341,
      "grad_norm": 2.701188564300537,
      "learning_rate": 2.536359744128957e-05,
      "loss": 0.1192,
      "num_input_tokens_seen": 13620208,
      "step": 20830
    },
    {
      "epoch": 10.919811320754716,
      "grad_norm": 2.106327772140503,
      "learning_rate": 2.5352164328108413e-05,
      "loss": 0.1268,
      "num_input_tokens_seen": 13623248,
      "step": 20835
    },
    {
      "epoch": 10.922431865828091,
      "grad_norm": 1.7912248373031616,
      "learning_rate": 2.53407311412585e-05,
      "loss": 0.0769,
      "num_input_tokens_seen": 13626672,
      "step": 20840
    },
    {
      "epoch": 10.925052410901468,
      "grad_norm": 2.0949509143829346,
      "learning_rate": 2.5329297883131526e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 13629776,
      "step": 20845
    },
    {
      "epoch": 10.927672955974844,
      "grad_norm": 3.2676568031311035,
      "learning_rate": 2.5317864556119182e-05,
      "loss": 0.154,
      "num_input_tokens_seen": 13632336,
      "step": 20850
    },
    {
      "epoch": 10.930293501048219,
      "grad_norm": 1.9031919240951538,
      "learning_rate": 2.53064311626132e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 13635600,
      "step": 20855
    },
    {
      "epoch": 10.932914046121594,
      "grad_norm": 2.5848467350006104,
      "learning_rate": 2.5294997705005318e-05,
      "loss": 0.1066,
      "num_input_tokens_seen": 13639248,
      "step": 20860
    },
    {
      "epoch": 10.935534591194969,
      "grad_norm": 4.516345500946045,
      "learning_rate": 2.5283564185687282e-05,
      "loss": 0.0962,
      "num_input_tokens_seen": 13641872,
      "step": 20865
    },
    {
      "epoch": 10.938155136268344,
      "grad_norm": 2.080580711364746,
      "learning_rate": 2.5272130607050848e-05,
      "loss": 0.1056,
      "num_input_tokens_seen": 13645424,
      "step": 20870
    },
    {
      "epoch": 10.94077568134172,
      "grad_norm": 2.346353769302368,
      "learning_rate": 2.5260696971487802e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 13648368,
      "step": 20875
    },
    {
      "epoch": 10.943396226415095,
      "grad_norm": 3.2231838703155518,
      "learning_rate": 2.5249263281389923e-05,
      "loss": 0.0972,
      "num_input_tokens_seen": 13650704,
      "step": 20880
    },
    {
      "epoch": 10.94601677148847,
      "grad_norm": 1.5664448738098145,
      "learning_rate": 2.5237829539149004e-05,
      "loss": 0.1106,
      "num_input_tokens_seen": 13653552,
      "step": 20885
    },
    {
      "epoch": 10.948637316561845,
      "grad_norm": 1.4076197147369385,
      "learning_rate": 2.5226395747156866e-05,
      "loss": 0.1464,
      "num_input_tokens_seen": 13657488,
      "step": 20890
    },
    {
      "epoch": 10.95125786163522,
      "grad_norm": 3.253906726837158,
      "learning_rate": 2.5214961907805317e-05,
      "loss": 0.1177,
      "num_input_tokens_seen": 13660976,
      "step": 20895
    },
    {
      "epoch": 10.953878406708595,
      "grad_norm": 1.9467628002166748,
      "learning_rate": 2.520352802348619e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 13664080,
      "step": 20900
    },
    {
      "epoch": 10.95649895178197,
      "grad_norm": 2.4770736694335938,
      "learning_rate": 2.5192094096591323e-05,
      "loss": 0.1971,
      "num_input_tokens_seen": 13666960,
      "step": 20905
    },
    {
      "epoch": 10.959119496855346,
      "grad_norm": 1.823742151260376,
      "learning_rate": 2.5180660129512556e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 13670832,
      "step": 20910
    },
    {
      "epoch": 10.96174004192872,
      "grad_norm": 3.3499410152435303,
      "learning_rate": 2.516922612464176e-05,
      "loss": 0.1299,
      "num_input_tokens_seen": 13673456,
      "step": 20915
    },
    {
      "epoch": 10.964360587002096,
      "grad_norm": 3.9785990715026855,
      "learning_rate": 2.5157792084370783e-05,
      "loss": 0.0688,
      "num_input_tokens_seen": 13676208,
      "step": 20920
    },
    {
      "epoch": 10.966981132075471,
      "grad_norm": 2.2902157306671143,
      "learning_rate": 2.5146358011091502e-05,
      "loss": 0.1562,
      "num_input_tokens_seen": 13679184,
      "step": 20925
    },
    {
      "epoch": 10.969601677148846,
      "grad_norm": 3.174952507019043,
      "learning_rate": 2.5134923907195805e-05,
      "loss": 0.1126,
      "num_input_tokens_seen": 13682896,
      "step": 20930
    },
    {
      "epoch": 10.972222222222221,
      "grad_norm": 8.853450775146484,
      "learning_rate": 2.5123489775075558e-05,
      "loss": 0.1427,
      "num_input_tokens_seen": 13686064,
      "step": 20935
    },
    {
      "epoch": 10.974842767295598,
      "grad_norm": 2.372352123260498,
      "learning_rate": 2.511205561712266e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 13689424,
      "step": 20940
    },
    {
      "epoch": 10.977463312368974,
      "grad_norm": 1.5466147661209106,
      "learning_rate": 2.5100621435729004e-05,
      "loss": 0.1597,
      "num_input_tokens_seen": 13692496,
      "step": 20945
    },
    {
      "epoch": 10.980083857442349,
      "grad_norm": 4.335494041442871,
      "learning_rate": 2.5089187233286492e-05,
      "loss": 0.1121,
      "num_input_tokens_seen": 13695728,
      "step": 20950
    },
    {
      "epoch": 10.982704402515724,
      "grad_norm": 1.9266589879989624,
      "learning_rate": 2.5077753012187026e-05,
      "loss": 0.1125,
      "num_input_tokens_seen": 13698640,
      "step": 20955
    },
    {
      "epoch": 10.985324947589099,
      "grad_norm": 4.444062232971191,
      "learning_rate": 2.506631877482252e-05,
      "loss": 0.09,
      "num_input_tokens_seen": 13702608,
      "step": 20960
    },
    {
      "epoch": 10.987945492662474,
      "grad_norm": 7.6909284591674805,
      "learning_rate": 2.505488452358488e-05,
      "loss": 0.1617,
      "num_input_tokens_seen": 13706352,
      "step": 20965
    },
    {
      "epoch": 10.99056603773585,
      "grad_norm": 2.140793561935425,
      "learning_rate": 2.504345026086603e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 13711056,
      "step": 20970
    },
    {
      "epoch": 10.993186582809225,
      "grad_norm": 1.2002378702163696,
      "learning_rate": 2.5032015989057877e-05,
      "loss": 0.0739,
      "num_input_tokens_seen": 13714480,
      "step": 20975
    },
    {
      "epoch": 10.9958071278826,
      "grad_norm": 1.4389245510101318,
      "learning_rate": 2.5020581710552348e-05,
      "loss": 0.0808,
      "num_input_tokens_seen": 13717136,
      "step": 20980
    },
    {
      "epoch": 10.998427672955975,
      "grad_norm": 1.1364061832427979,
      "learning_rate": 2.5009147427741358e-05,
      "loss": 0.1702,
      "num_input_tokens_seen": 13720464,
      "step": 20985
    },
    {
      "epoch": 11.0,
      "eval_loss": 0.891692042350769,
      "eval_runtime": 11.6359,
      "eval_samples_per_second": 72.878,
      "eval_steps_per_second": 18.219,
      "num_input_tokens_seen": 13721824,
      "step": 20988
    },
    {
      "epoch": 11.00104821802935,
      "grad_norm": 7.559231281280518,
      "learning_rate": 2.4997713143016822e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 13723072,
      "step": 20990
    },
    {
      "epoch": 11.003668763102725,
      "grad_norm": 1.652830958366394,
      "learning_rate": 2.498627885877068e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 13726400,
      "step": 20995
    },
    {
      "epoch": 11.0062893081761,
      "grad_norm": 0.748428463935852,
      "learning_rate": 2.4974844577394846e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 13729280,
      "step": 21000
    },
    {
      "epoch": 11.008909853249476,
      "grad_norm": 2.620236873626709,
      "learning_rate": 2.496341030128123e-05,
      "loss": 0.0653,
      "num_input_tokens_seen": 13732416,
      "step": 21005
    },
    {
      "epoch": 11.01153039832285,
      "grad_norm": 1.7061060667037964,
      "learning_rate": 2.495197603282177e-05,
      "loss": 0.1422,
      "num_input_tokens_seen": 13735232,
      "step": 21010
    },
    {
      "epoch": 11.014150943396226,
      "grad_norm": 2.074873447418213,
      "learning_rate": 2.4940541774408368e-05,
      "loss": 0.0938,
      "num_input_tokens_seen": 13738656,
      "step": 21015
    },
    {
      "epoch": 11.016771488469601,
      "grad_norm": 2.206027030944824,
      "learning_rate": 2.4929107528432953e-05,
      "loss": 0.0828,
      "num_input_tokens_seen": 13741568,
      "step": 21020
    },
    {
      "epoch": 11.019392033542976,
      "grad_norm": 0.9797276854515076,
      "learning_rate": 2.4917673297287446e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 13744192,
      "step": 21025
    },
    {
      "epoch": 11.022012578616351,
      "grad_norm": 16.309555053710938,
      "learning_rate": 2.4906239083363728e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 13746656,
      "step": 21030
    },
    {
      "epoch": 11.024633123689727,
      "grad_norm": 2.1485087871551514,
      "learning_rate": 2.4894804889053732e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 13749920,
      "step": 21035
    },
    {
      "epoch": 11.027253668763104,
      "grad_norm": 1.693690538406372,
      "learning_rate": 2.4883370716749347e-05,
      "loss": 0.0649,
      "num_input_tokens_seen": 13752864,
      "step": 21040
    },
    {
      "epoch": 11.029874213836479,
      "grad_norm": 1.3735302686691284,
      "learning_rate": 2.4871936568842486e-05,
      "loss": 0.0555,
      "num_input_tokens_seen": 13756192,
      "step": 21045
    },
    {
      "epoch": 11.032494758909854,
      "grad_norm": 2.844832181930542,
      "learning_rate": 2.4860502447725033e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 13759968,
      "step": 21050
    },
    {
      "epoch": 11.035115303983229,
      "grad_norm": 2.0780255794525146,
      "learning_rate": 2.4849068355788873e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 13763072,
      "step": 21055
    },
    {
      "epoch": 11.037735849056604,
      "grad_norm": 13.577875137329102,
      "learning_rate": 2.4837634295425892e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 13765728,
      "step": 21060
    },
    {
      "epoch": 11.04035639412998,
      "grad_norm": 1.3527148962020874,
      "learning_rate": 2.4826200269027957e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 13769504,
      "step": 21065
    },
    {
      "epoch": 11.042976939203355,
      "grad_norm": 1.2371832132339478,
      "learning_rate": 2.481476627898695e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 13773760,
      "step": 21070
    },
    {
      "epoch": 11.04559748427673,
      "grad_norm": 1.8280794620513916,
      "learning_rate": 2.4803332327694724e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 13777088,
      "step": 21075
    },
    {
      "epoch": 11.048218029350105,
      "grad_norm": 1.0692769289016724,
      "learning_rate": 2.4791898417543136e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 13781120,
      "step": 21080
    },
    {
      "epoch": 11.05083857442348,
      "grad_norm": 0.4469650983810425,
      "learning_rate": 2.4780464550924014e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 13784064,
      "step": 21085
    },
    {
      "epoch": 11.053459119496855,
      "grad_norm": 2.543787956237793,
      "learning_rate": 2.4769030730229213e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 13787104,
      "step": 21090
    },
    {
      "epoch": 11.05607966457023,
      "grad_norm": 2.35208797454834,
      "learning_rate": 2.475759695785054e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 13789824,
      "step": 21095
    },
    {
      "epoch": 11.058700209643606,
      "grad_norm": 1.9910616874694824,
      "learning_rate": 2.4746163236179822e-05,
      "loss": 0.1042,
      "num_input_tokens_seen": 13793152,
      "step": 21100
    },
    {
      "epoch": 11.06132075471698,
      "grad_norm": 2.832045316696167,
      "learning_rate": 2.4734729567608858e-05,
      "loss": 0.1227,
      "num_input_tokens_seen": 13796288,
      "step": 21105
    },
    {
      "epoch": 11.063941299790356,
      "grad_norm": 1.4413230419158936,
      "learning_rate": 2.472329595452945e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 13799072,
      "step": 21110
    },
    {
      "epoch": 11.066561844863731,
      "grad_norm": 2.3507399559020996,
      "learning_rate": 2.4711862399333367e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 13802560,
      "step": 21115
    },
    {
      "epoch": 11.069182389937106,
      "grad_norm": 2.4457693099975586,
      "learning_rate": 2.4700428904412377e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 13806592,
      "step": 21120
    },
    {
      "epoch": 11.071802935010481,
      "grad_norm": 3.2865335941314697,
      "learning_rate": 2.4688995472158245e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 13809568,
      "step": 21125
    },
    {
      "epoch": 11.074423480083857,
      "grad_norm": 3.891639471054077,
      "learning_rate": 2.4677562104962712e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 13812960,
      "step": 21130
    },
    {
      "epoch": 11.077044025157234,
      "grad_norm": 5.72788667678833,
      "learning_rate": 2.4666128805217513e-05,
      "loss": 0.1754,
      "num_input_tokens_seen": 13815712,
      "step": 21135
    },
    {
      "epoch": 11.079664570230609,
      "grad_norm": 1.2585480213165283,
      "learning_rate": 2.465469557531436e-05,
      "loss": 0.0711,
      "num_input_tokens_seen": 13818560,
      "step": 21140
    },
    {
      "epoch": 11.082285115303984,
      "grad_norm": 1.7379786968231201,
      "learning_rate": 2.4643262417644944e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 13823264,
      "step": 21145
    },
    {
      "epoch": 11.084905660377359,
      "grad_norm": 5.72144889831543,
      "learning_rate": 2.463182933460096e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 13826464,
      "step": 21150
    },
    {
      "epoch": 11.087526205450734,
      "grad_norm": 2.2555325031280518,
      "learning_rate": 2.4620396328574087e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 13829280,
      "step": 21155
    },
    {
      "epoch": 11.09014675052411,
      "grad_norm": 1.789955496788025,
      "learning_rate": 2.4608963401955965e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 13833344,
      "step": 21160
    },
    {
      "epoch": 11.092767295597485,
      "grad_norm": 1.5224488973617554,
      "learning_rate": 2.4597530557138247e-05,
      "loss": 0.1086,
      "num_input_tokens_seen": 13837088,
      "step": 21165
    },
    {
      "epoch": 11.09538784067086,
      "grad_norm": 3.4172167778015137,
      "learning_rate": 2.4586097796512538e-05,
      "loss": 0.0767,
      "num_input_tokens_seen": 13839488,
      "step": 21170
    },
    {
      "epoch": 11.098008385744235,
      "grad_norm": 3.252310037612915,
      "learning_rate": 2.457466512247045e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 13842528,
      "step": 21175
    },
    {
      "epoch": 11.10062893081761,
      "grad_norm": 2.694448232650757,
      "learning_rate": 2.4563232537403563e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 13845312,
      "step": 21180
    },
    {
      "epoch": 11.103249475890985,
      "grad_norm": 7.011220455169678,
      "learning_rate": 2.4551800043703442e-05,
      "loss": 0.0566,
      "num_input_tokens_seen": 13848256,
      "step": 21185
    },
    {
      "epoch": 11.10587002096436,
      "grad_norm": 0.8109320998191833,
      "learning_rate": 2.454036764376165e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 13852064,
      "step": 21190
    },
    {
      "epoch": 11.108490566037736,
      "grad_norm": 2.008065938949585,
      "learning_rate": 2.4528935339969705e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 13855040,
      "step": 21195
    },
    {
      "epoch": 11.11111111111111,
      "grad_norm": 1.3947327136993408,
      "learning_rate": 2.45175031347191e-05,
      "loss": 0.0658,
      "num_input_tokens_seen": 13858528,
      "step": 21200
    },
    {
      "epoch": 11.113731656184486,
      "grad_norm": 2.303281545639038,
      "learning_rate": 2.4506071030401344e-05,
      "loss": 0.1183,
      "num_input_tokens_seen": 13862048,
      "step": 21205
    },
    {
      "epoch": 11.116352201257861,
      "grad_norm": 3.1292054653167725,
      "learning_rate": 2.4494639029407884e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 13864864,
      "step": 21210
    },
    {
      "epoch": 11.118972746331236,
      "grad_norm": 2.115318536758423,
      "learning_rate": 2.448320713413018e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 13868256,
      "step": 21215
    },
    {
      "epoch": 11.121593291404611,
      "grad_norm": 5.178234100341797,
      "learning_rate": 2.4471775346959654e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 13870944,
      "step": 21220
    },
    {
      "epoch": 11.124213836477987,
      "grad_norm": 3.868936061859131,
      "learning_rate": 2.446034367028769e-05,
      "loss": 0.0968,
      "num_input_tokens_seen": 13874016,
      "step": 21225
    },
    {
      "epoch": 11.126834381551364,
      "grad_norm": 2.776374578475952,
      "learning_rate": 2.444891210650567e-05,
      "loss": 0.1201,
      "num_input_tokens_seen": 13881184,
      "step": 21230
    },
    {
      "epoch": 11.129454926624739,
      "grad_norm": 2.0449678897857666,
      "learning_rate": 2.4437480658004958e-05,
      "loss": 0.0726,
      "num_input_tokens_seen": 13884224,
      "step": 21235
    },
    {
      "epoch": 11.132075471698114,
      "grad_norm": 8.263922691345215,
      "learning_rate": 2.4426049327176868e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 13888064,
      "step": 21240
    },
    {
      "epoch": 11.134696016771489,
      "grad_norm": 2.5752594470977783,
      "learning_rate": 2.4414618116412716e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 13890912,
      "step": 21245
    },
    {
      "epoch": 11.137316561844864,
      "grad_norm": 2.278376340866089,
      "learning_rate": 2.440318702810378e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 13894432,
      "step": 21250
    },
    {
      "epoch": 11.13993710691824,
      "grad_norm": 6.832921981811523,
      "learning_rate": 2.43917560646413e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 13898048,
      "step": 21255
    },
    {
      "epoch": 11.142557651991615,
      "grad_norm": 2.9356942176818848,
      "learning_rate": 2.4380325228416518e-05,
      "loss": 0.1799,
      "num_input_tokens_seen": 13901184,
      "step": 21260
    },
    {
      "epoch": 11.14517819706499,
      "grad_norm": 1.8663933277130127,
      "learning_rate": 2.4368894521820624e-05,
      "loss": 0.0488,
      "num_input_tokens_seen": 13904864,
      "step": 21265
    },
    {
      "epoch": 11.147798742138365,
      "grad_norm": 1.9546617269515991,
      "learning_rate": 2.43574639472448e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 13908096,
      "step": 21270
    },
    {
      "epoch": 11.15041928721174,
      "grad_norm": 2.3040568828582764,
      "learning_rate": 2.4346033507080178e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 13911392,
      "step": 21275
    },
    {
      "epoch": 11.153039832285115,
      "grad_norm": 1.8206466436386108,
      "learning_rate": 2.4334603203717903e-05,
      "loss": 0.1079,
      "num_input_tokens_seen": 13915104,
      "step": 21280
    },
    {
      "epoch": 11.15566037735849,
      "grad_norm": 10.496461868286133,
      "learning_rate": 2.432317303954904e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 13918176,
      "step": 21285
    },
    {
      "epoch": 11.158280922431866,
      "grad_norm": 1.8433029651641846,
      "learning_rate": 2.4311743016964648e-05,
      "loss": 0.0482,
      "num_input_tokens_seen": 13921120,
      "step": 21290
    },
    {
      "epoch": 11.16090146750524,
      "grad_norm": 4.543839454650879,
      "learning_rate": 2.4300313138355772e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 13924224,
      "step": 21295
    },
    {
      "epoch": 11.163522012578616,
      "grad_norm": 2.6639552116394043,
      "learning_rate": 2.4288883406113396e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 13927488,
      "step": 21300
    },
    {
      "epoch": 11.166142557651991,
      "grad_norm": 2.159092903137207,
      "learning_rate": 2.4277453822628512e-05,
      "loss": 0.0517,
      "num_input_tokens_seen": 13930880,
      "step": 21305
    },
    {
      "epoch": 11.168763102725366,
      "grad_norm": 1.697906255722046,
      "learning_rate": 2.4266024390292037e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 13933728,
      "step": 21310
    },
    {
      "epoch": 11.171383647798741,
      "grad_norm": 12.914244651794434,
      "learning_rate": 2.425459511149489e-05,
      "loss": 0.0807,
      "num_input_tokens_seen": 13937664,
      "step": 21315
    },
    {
      "epoch": 11.174004192872117,
      "grad_norm": 1.0495290756225586,
      "learning_rate": 2.424316598862793e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 13940544,
      "step": 21320
    },
    {
      "epoch": 11.176624737945493,
      "grad_norm": 1.3700542449951172,
      "learning_rate": 2.4231737024082023e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 13943712,
      "step": 21325
    },
    {
      "epoch": 11.179245283018869,
      "grad_norm": 1.791951298713684,
      "learning_rate": 2.4220308220247954e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 13947200,
      "step": 21330
    },
    {
      "epoch": 11.181865828092244,
      "grad_norm": 3.2823543548583984,
      "learning_rate": 2.4208879579516528e-05,
      "loss": 0.339,
      "num_input_tokens_seen": 13950240,
      "step": 21335
    },
    {
      "epoch": 11.184486373165619,
      "grad_norm": 1.6253395080566406,
      "learning_rate": 2.419745110427846e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 13953440,
      "step": 21340
    },
    {
      "epoch": 11.187106918238994,
      "grad_norm": 3.4615979194641113,
      "learning_rate": 2.418602279692446e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 13956416,
      "step": 21345
    },
    {
      "epoch": 11.18972746331237,
      "grad_norm": 2.0697412490844727,
      "learning_rate": 2.417459465984521e-05,
      "loss": 0.0781,
      "num_input_tokens_seen": 13959840,
      "step": 21350
    },
    {
      "epoch": 11.192348008385745,
      "grad_norm": 5.209142684936523,
      "learning_rate": 2.4163166695431336e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 13963168,
      "step": 21355
    },
    {
      "epoch": 11.19496855345912,
      "grad_norm": 1.921728491783142,
      "learning_rate": 2.4151738906073446e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 13965632,
      "step": 21360
    },
    {
      "epoch": 11.197589098532495,
      "grad_norm": 2.4537768363952637,
      "learning_rate": 2.4140311294162105e-05,
      "loss": 0.063,
      "num_input_tokens_seen": 13969184,
      "step": 21365
    },
    {
      "epoch": 11.20020964360587,
      "grad_norm": 2.3293209075927734,
      "learning_rate": 2.412888386208782e-05,
      "loss": 0.0627,
      "num_input_tokens_seen": 13972800,
      "step": 21370
    },
    {
      "epoch": 11.202830188679245,
      "grad_norm": 3.4114246368408203,
      "learning_rate": 2.4117456612241103e-05,
      "loss": 0.1053,
      "num_input_tokens_seen": 13976128,
      "step": 21375
    },
    {
      "epoch": 11.20545073375262,
      "grad_norm": 2.788464307785034,
      "learning_rate": 2.410602954701239e-05,
      "loss": 0.0746,
      "num_input_tokens_seen": 13979424,
      "step": 21380
    },
    {
      "epoch": 11.208071278825996,
      "grad_norm": 0.6070854663848877,
      "learning_rate": 2.409460266879209e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 13981856,
      "step": 21385
    },
    {
      "epoch": 11.21069182389937,
      "grad_norm": 1.4313030242919922,
      "learning_rate": 2.4083175979970592e-05,
      "loss": 0.0402,
      "num_input_tokens_seen": 13985248,
      "step": 21390
    },
    {
      "epoch": 11.213312368972746,
      "grad_norm": 2.4010136127471924,
      "learning_rate": 2.4071749482938226e-05,
      "loss": 0.1368,
      "num_input_tokens_seen": 13988576,
      "step": 21395
    },
    {
      "epoch": 11.215932914046121,
      "grad_norm": 6.246856689453125,
      "learning_rate": 2.4060323180085266e-05,
      "loss": 0.0676,
      "num_input_tokens_seen": 13991232,
      "step": 21400
    },
    {
      "epoch": 11.218553459119496,
      "grad_norm": 1.9758285284042358,
      "learning_rate": 2.404889707380198e-05,
      "loss": 0.0546,
      "num_input_tokens_seen": 13994592,
      "step": 21405
    },
    {
      "epoch": 11.221174004192871,
      "grad_norm": 6.169962406158447,
      "learning_rate": 2.403747116647857e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 13997280,
      "step": 21410
    },
    {
      "epoch": 11.223794549266247,
      "grad_norm": 3.899163246154785,
      "learning_rate": 2.4026045460505216e-05,
      "loss": 0.0717,
      "num_input_tokens_seen": 14000384,
      "step": 21415
    },
    {
      "epoch": 11.226415094339623,
      "grad_norm": 3.1802420616149902,
      "learning_rate": 2.4014619958272047e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 14003872,
      "step": 21420
    },
    {
      "epoch": 11.229035639412999,
      "grad_norm": 2.1172070503234863,
      "learning_rate": 2.4003194662169132e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 14006976,
      "step": 21425
    },
    {
      "epoch": 11.231656184486374,
      "grad_norm": 1.8737415075302124,
      "learning_rate": 2.3991769574586522e-05,
      "loss": 0.1709,
      "num_input_tokens_seen": 14009472,
      "step": 21430
    },
    {
      "epoch": 11.234276729559749,
      "grad_norm": 1.6729120016098022,
      "learning_rate": 2.3980344697914215e-05,
      "loss": 0.0816,
      "num_input_tokens_seen": 14012480,
      "step": 21435
    },
    {
      "epoch": 11.236897274633124,
      "grad_norm": 1.6710816621780396,
      "learning_rate": 2.3968920034542168e-05,
      "loss": 0.0594,
      "num_input_tokens_seen": 14015072,
      "step": 21440
    },
    {
      "epoch": 11.2395178197065,
      "grad_norm": 2.555330514907837,
      "learning_rate": 2.3957495586860278e-05,
      "loss": 0.0551,
      "num_input_tokens_seen": 14018496,
      "step": 21445
    },
    {
      "epoch": 11.242138364779874,
      "grad_norm": 0.7818471789360046,
      "learning_rate": 2.394607135725843e-05,
      "loss": 0.1548,
      "num_input_tokens_seen": 14021600,
      "step": 21450
    },
    {
      "epoch": 11.24475890985325,
      "grad_norm": 3.5078375339508057,
      "learning_rate": 2.3934647348126426e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 14024544,
      "step": 21455
    },
    {
      "epoch": 11.247379454926625,
      "grad_norm": 1.9894264936447144,
      "learning_rate": 2.3923223561854035e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 14027968,
      "step": 21460
    },
    {
      "epoch": 11.25,
      "grad_norm": 2.5484836101531982,
      "learning_rate": 2.391180000083099e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 14030688,
      "step": 21465
    },
    {
      "epoch": 11.252620545073375,
      "grad_norm": 2.2355072498321533,
      "learning_rate": 2.390037666744698e-05,
      "loss": 0.0581,
      "num_input_tokens_seen": 14033408,
      "step": 21470
    },
    {
      "epoch": 11.25524109014675,
      "grad_norm": 1.610683560371399,
      "learning_rate": 2.388895356409162e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 14036416,
      "step": 21475
    },
    {
      "epoch": 11.257861635220126,
      "grad_norm": 2.85935115814209,
      "learning_rate": 2.3877530693154506e-05,
      "loss": 0.0833,
      "num_input_tokens_seen": 14039872,
      "step": 21480
    },
    {
      "epoch": 11.2604821802935,
      "grad_norm": 2.1968555450439453,
      "learning_rate": 2.3866108057025162e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 14042624,
      "step": 21485
    },
    {
      "epoch": 11.263102725366876,
      "grad_norm": 2.166698932647705,
      "learning_rate": 2.385468565809307e-05,
      "loss": 0.1123,
      "num_input_tokens_seen": 14045984,
      "step": 21490
    },
    {
      "epoch": 11.265723270440251,
      "grad_norm": 1.9044413566589355,
      "learning_rate": 2.3843263498747675e-05,
      "loss": 0.1004,
      "num_input_tokens_seen": 14049152,
      "step": 21495
    },
    {
      "epoch": 11.268343815513626,
      "grad_norm": 3.313552141189575,
      "learning_rate": 2.383184158137836e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 14052320,
      "step": 21500
    },
    {
      "epoch": 11.270964360587001,
      "grad_norm": 2.0853731632232666,
      "learning_rate": 2.3820419908374465e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 14055584,
      "step": 21505
    },
    {
      "epoch": 11.273584905660377,
      "grad_norm": 1.5803658962249756,
      "learning_rate": 2.3808998482125265e-05,
      "loss": 0.066,
      "num_input_tokens_seen": 14058816,
      "step": 21510
    },
    {
      "epoch": 11.276205450733753,
      "grad_norm": 2.095848798751831,
      "learning_rate": 2.3797577305019987e-05,
      "loss": 0.1572,
      "num_input_tokens_seen": 14062528,
      "step": 21515
    },
    {
      "epoch": 11.278825995807129,
      "grad_norm": 3.931993246078491,
      "learning_rate": 2.378615637944783e-05,
      "loss": 0.092,
      "num_input_tokens_seen": 14065280,
      "step": 21520
    },
    {
      "epoch": 11.281446540880504,
      "grad_norm": 6.452297210693359,
      "learning_rate": 2.3774735707797895e-05,
      "loss": 0.0705,
      "num_input_tokens_seen": 14068160,
      "step": 21525
    },
    {
      "epoch": 11.284067085953879,
      "grad_norm": 1.1311516761779785,
      "learning_rate": 2.376331529245928e-05,
      "loss": 0.0585,
      "num_input_tokens_seen": 14071712,
      "step": 21530
    },
    {
      "epoch": 11.286687631027254,
      "grad_norm": 2.1236612796783447,
      "learning_rate": 2.3751895135821006e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 14074912,
      "step": 21535
    },
    {
      "epoch": 11.28930817610063,
      "grad_norm": 2.6200902462005615,
      "learning_rate": 2.374047524027202e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 14078912,
      "step": 21540
    },
    {
      "epoch": 11.291928721174004,
      "grad_norm": 0.6540394425392151,
      "learning_rate": 2.3729055608201243e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 14082304,
      "step": 21545
    },
    {
      "epoch": 11.29454926624738,
      "grad_norm": 1.8642687797546387,
      "learning_rate": 2.3717636241997537e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 14085056,
      "step": 21550
    },
    {
      "epoch": 11.297169811320755,
      "grad_norm": 2.285910129547119,
      "learning_rate": 2.3706217144049696e-05,
      "loss": 0.0715,
      "num_input_tokens_seen": 14088128,
      "step": 21555
    },
    {
      "epoch": 11.29979035639413,
      "grad_norm": 1.990512728691101,
      "learning_rate": 2.369479831674647e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 14090912,
      "step": 21560
    },
    {
      "epoch": 11.302410901467505,
      "grad_norm": 1.6248890161514282,
      "learning_rate": 2.3683379762476555e-05,
      "loss": 0.1167,
      "num_input_tokens_seen": 14094464,
      "step": 21565
    },
    {
      "epoch": 11.30503144654088,
      "grad_norm": 2.8328912258148193,
      "learning_rate": 2.3671961483628562e-05,
      "loss": 0.093,
      "num_input_tokens_seen": 14097280,
      "step": 21570
    },
    {
      "epoch": 11.307651991614255,
      "grad_norm": 1.0200973749160767,
      "learning_rate": 2.3660543482591078e-05,
      "loss": 0.0773,
      "num_input_tokens_seen": 14100224,
      "step": 21575
    },
    {
      "epoch": 11.31027253668763,
      "grad_norm": 0.8349900841712952,
      "learning_rate": 2.364912576175261e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 14103744,
      "step": 21580
    },
    {
      "epoch": 11.312893081761006,
      "grad_norm": 4.272836208343506,
      "learning_rate": 2.363770832350163e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 14107008,
      "step": 21585
    },
    {
      "epoch": 11.315513626834381,
      "grad_norm": 1.1005576848983765,
      "learning_rate": 2.362629117022653e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 14110176,
      "step": 21590
    },
    {
      "epoch": 11.318134171907756,
      "grad_norm": 4.490825176239014,
      "learning_rate": 2.3614874304315633e-05,
      "loss": 0.1663,
      "num_input_tokens_seen": 14115968,
      "step": 21595
    },
    {
      "epoch": 11.320754716981131,
      "grad_norm": 1.2792484760284424,
      "learning_rate": 2.3603457728157237e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 14120864,
      "step": 21600
    },
    {
      "epoch": 11.323375262054507,
      "grad_norm": 1.7103885412216187,
      "learning_rate": 2.3592041444139547e-05,
      "loss": 0.0943,
      "num_input_tokens_seen": 14124576,
      "step": 21605
    },
    {
      "epoch": 11.325995807127883,
      "grad_norm": 1.2637996673583984,
      "learning_rate": 2.358062545465073e-05,
      "loss": 0.1147,
      "num_input_tokens_seen": 14127648,
      "step": 21610
    },
    {
      "epoch": 11.328616352201259,
      "grad_norm": 3.098773717880249,
      "learning_rate": 2.3569209762078866e-05,
      "loss": 0.069,
      "num_input_tokens_seen": 14130400,
      "step": 21615
    },
    {
      "epoch": 11.331236897274634,
      "grad_norm": 4.379977703094482,
      "learning_rate": 2.355779436881201e-05,
      "loss": 0.0791,
      "num_input_tokens_seen": 14134080,
      "step": 21620
    },
    {
      "epoch": 11.333857442348009,
      "grad_norm": 2.615100622177124,
      "learning_rate": 2.3546379277238107e-05,
      "loss": 0.0835,
      "num_input_tokens_seen": 14138080,
      "step": 21625
    },
    {
      "epoch": 11.336477987421384,
      "grad_norm": 2.0074992179870605,
      "learning_rate": 2.353496448974508e-05,
      "loss": 0.0681,
      "num_input_tokens_seen": 14142432,
      "step": 21630
    },
    {
      "epoch": 11.33909853249476,
      "grad_norm": 2.727760076522827,
      "learning_rate": 2.3523550008720763e-05,
      "loss": 0.089,
      "num_input_tokens_seen": 14145440,
      "step": 21635
    },
    {
      "epoch": 11.341719077568134,
      "grad_norm": 2.7183544635772705,
      "learning_rate": 2.3512135836552945e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 14148672,
      "step": 21640
    },
    {
      "epoch": 11.34433962264151,
      "grad_norm": 1.7370837926864624,
      "learning_rate": 2.350072197562933e-05,
      "loss": 0.078,
      "num_input_tokens_seen": 14152000,
      "step": 21645
    },
    {
      "epoch": 11.346960167714885,
      "grad_norm": 2.464024066925049,
      "learning_rate": 2.3489308428337586e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 14154912,
      "step": 21650
    },
    {
      "epoch": 11.34958071278826,
      "grad_norm": 1.3234881162643433,
      "learning_rate": 2.3477895197065275e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 14158208,
      "step": 21655
    },
    {
      "epoch": 11.352201257861635,
      "grad_norm": 1.0336544513702393,
      "learning_rate": 2.346648228419992e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 14161920,
      "step": 21660
    },
    {
      "epoch": 11.35482180293501,
      "grad_norm": 3.1413378715515137,
      "learning_rate": 2.3455069692128982e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 14165120,
      "step": 21665
    },
    {
      "epoch": 11.357442348008385,
      "grad_norm": 2.530691385269165,
      "learning_rate": 2.3443657423239832e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 14168000,
      "step": 21670
    },
    {
      "epoch": 11.36006289308176,
      "grad_norm": 2.6887247562408447,
      "learning_rate": 2.34322454799198e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 14171584,
      "step": 21675
    },
    {
      "epoch": 11.362683438155136,
      "grad_norm": 2.713172435760498,
      "learning_rate": 2.3420833864556133e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 14174336,
      "step": 21680
    },
    {
      "epoch": 11.365303983228511,
      "grad_norm": 5.991670608520508,
      "learning_rate": 2.3409422579535995e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 14177344,
      "step": 21685
    },
    {
      "epoch": 11.367924528301886,
      "grad_norm": 1.9211145639419556,
      "learning_rate": 2.339801162724651e-05,
      "loss": 0.0672,
      "num_input_tokens_seen": 14180608,
      "step": 21690
    },
    {
      "epoch": 11.370545073375261,
      "grad_norm": 2.556917905807495,
      "learning_rate": 2.338660101007471e-05,
      "loss": 0.1152,
      "num_input_tokens_seen": 14183872,
      "step": 21695
    },
    {
      "epoch": 11.373165618448636,
      "grad_norm": 3.612764358520508,
      "learning_rate": 2.3375190730407578e-05,
      "loss": 0.1258,
      "num_input_tokens_seen": 14186720,
      "step": 21700
    },
    {
      "epoch": 11.375786163522013,
      "grad_norm": 1.5085676908493042,
      "learning_rate": 2.336378079063202e-05,
      "loss": 0.0979,
      "num_input_tokens_seen": 14190336,
      "step": 21705
    },
    {
      "epoch": 11.378406708595389,
      "grad_norm": 2.945164442062378,
      "learning_rate": 2.335237119313484e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 14193760,
      "step": 21710
    },
    {
      "epoch": 11.381027253668764,
      "grad_norm": 2.098662853240967,
      "learning_rate": 2.3340961940302807e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 14196096,
      "step": 21715
    },
    {
      "epoch": 11.383647798742139,
      "grad_norm": 2.187368392944336,
      "learning_rate": 2.3329553034522616e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 14199296,
      "step": 21720
    },
    {
      "epoch": 11.386268343815514,
      "grad_norm": 2.41707444190979,
      "learning_rate": 2.3318144478180866e-05,
      "loss": 0.0747,
      "num_input_tokens_seen": 14202528,
      "step": 21725
    },
    {
      "epoch": 11.38888888888889,
      "grad_norm": 0.9051204323768616,
      "learning_rate": 2.330673627366411e-05,
      "loss": 0.075,
      "num_input_tokens_seen": 14205568,
      "step": 21730
    },
    {
      "epoch": 11.391509433962264,
      "grad_norm": 1.7523789405822754,
      "learning_rate": 2.329532842335881e-05,
      "loss": 0.0684,
      "num_input_tokens_seen": 14209056,
      "step": 21735
    },
    {
      "epoch": 11.39412997903564,
      "grad_norm": 4.371617317199707,
      "learning_rate": 2.328392092965135e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 14214720,
      "step": 21740
    },
    {
      "epoch": 11.396750524109015,
      "grad_norm": 1.750943660736084,
      "learning_rate": 2.3272513794928054e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 14217664,
      "step": 21745
    },
    {
      "epoch": 11.39937106918239,
      "grad_norm": 1.5667273998260498,
      "learning_rate": 2.326110702157516e-05,
      "loss": 0.0678,
      "num_input_tokens_seen": 14221216,
      "step": 21750
    },
    {
      "epoch": 11.401991614255765,
      "grad_norm": 2.713300943374634,
      "learning_rate": 2.3249700611978847e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 14224000,
      "step": 21755
    },
    {
      "epoch": 11.40461215932914,
      "grad_norm": 2.2261300086975098,
      "learning_rate": 2.323829456852519e-05,
      "loss": 0.0789,
      "num_input_tokens_seen": 14227872,
      "step": 21760
    },
    {
      "epoch": 11.407232704402515,
      "grad_norm": 3.6400301456451416,
      "learning_rate": 2.3226888893600223e-05,
      "loss": 0.0826,
      "num_input_tokens_seen": 14231936,
      "step": 21765
    },
    {
      "epoch": 11.40985324947589,
      "grad_norm": 2.5976338386535645,
      "learning_rate": 2.3215483589589865e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 14235328,
      "step": 21770
    },
    {
      "epoch": 11.412473794549266,
      "grad_norm": 2.857581853866577,
      "learning_rate": 2.3204078658879974e-05,
      "loss": 0.0847,
      "num_input_tokens_seen": 14238848,
      "step": 21775
    },
    {
      "epoch": 11.415094339622641,
      "grad_norm": 3.0743651390075684,
      "learning_rate": 2.3192674103856337e-05,
      "loss": 0.0938,
      "num_input_tokens_seen": 14242336,
      "step": 21780
    },
    {
      "epoch": 11.417714884696016,
      "grad_norm": 1.8160182237625122,
      "learning_rate": 2.318126992690467e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 14245728,
      "step": 21785
    },
    {
      "epoch": 11.420335429769391,
      "grad_norm": 2.1764729022979736,
      "learning_rate": 2.3169866130410585e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 14248448,
      "step": 21790
    },
    {
      "epoch": 11.422955974842766,
      "grad_norm": 2.5180444717407227,
      "learning_rate": 2.315846271675962e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 14251648,
      "step": 21795
    },
    {
      "epoch": 11.425576519916142,
      "grad_norm": 2.018460512161255,
      "learning_rate": 2.3147059688337244e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 14254624,
      "step": 21800
    },
    {
      "epoch": 11.428197064989519,
      "grad_norm": 2.576293706893921,
      "learning_rate": 2.313565704752884e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 14257728,
      "step": 21805
    },
    {
      "epoch": 11.430817610062894,
      "grad_norm": 3.325991630554199,
      "learning_rate": 2.3124254796719718e-05,
      "loss": 0.1395,
      "num_input_tokens_seen": 14262176,
      "step": 21810
    },
    {
      "epoch": 11.433438155136269,
      "grad_norm": 3.875612735748291,
      "learning_rate": 2.3112852938295088e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 14265760,
      "step": 21815
    },
    {
      "epoch": 11.436058700209644,
      "grad_norm": 8.159932136535645,
      "learning_rate": 2.3101451474640104e-05,
      "loss": 0.0772,
      "num_input_tokens_seen": 14268768,
      "step": 21820
    },
    {
      "epoch": 11.43867924528302,
      "grad_norm": 2.3571536540985107,
      "learning_rate": 2.3090050408139808e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 14272064,
      "step": 21825
    },
    {
      "epoch": 11.441299790356394,
      "grad_norm": 2.5649919509887695,
      "learning_rate": 2.307864974117917e-05,
      "loss": 0.1113,
      "num_input_tokens_seen": 14275456,
      "step": 21830
    },
    {
      "epoch": 11.44392033542977,
      "grad_norm": 2.033226251602173,
      "learning_rate": 2.3067249476143087e-05,
      "loss": 0.0822,
      "num_input_tokens_seen": 14279712,
      "step": 21835
    },
    {
      "epoch": 11.446540880503145,
      "grad_norm": 1.5491219758987427,
      "learning_rate": 2.3055849615416363e-05,
      "loss": 0.0638,
      "num_input_tokens_seen": 14283616,
      "step": 21840
    },
    {
      "epoch": 11.44916142557652,
      "grad_norm": 1.920383334159851,
      "learning_rate": 2.304445016138372e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 14286720,
      "step": 21845
    },
    {
      "epoch": 11.451781970649895,
      "grad_norm": 3.6370749473571777,
      "learning_rate": 2.303305111642981e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 14291232,
      "step": 21850
    },
    {
      "epoch": 11.45440251572327,
      "grad_norm": 2.675304412841797,
      "learning_rate": 2.3021652482939144e-05,
      "loss": 0.0666,
      "num_input_tokens_seen": 14294496,
      "step": 21855
    },
    {
      "epoch": 11.457023060796645,
      "grad_norm": 1.5487200021743774,
      "learning_rate": 2.3010254263296214e-05,
      "loss": 0.0395,
      "num_input_tokens_seen": 14297088,
      "step": 21860
    },
    {
      "epoch": 11.45964360587002,
      "grad_norm": 3.6875159740448,
      "learning_rate": 2.2998856459885396e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 14301056,
      "step": 21865
    },
    {
      "epoch": 11.462264150943396,
      "grad_norm": 1.9406445026397705,
      "learning_rate": 2.298745907509097e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 14303968,
      "step": 21870
    },
    {
      "epoch": 11.464884696016771,
      "grad_norm": 0.9751476049423218,
      "learning_rate": 2.2976062111297156e-05,
      "loss": 0.0885,
      "num_input_tokens_seen": 14308256,
      "step": 21875
    },
    {
      "epoch": 11.467505241090146,
      "grad_norm": 2.363027811050415,
      "learning_rate": 2.296466557088805e-05,
      "loss": 0.1131,
      "num_input_tokens_seen": 14311424,
      "step": 21880
    },
    {
      "epoch": 11.470125786163521,
      "grad_norm": 3.860950469970703,
      "learning_rate": 2.2953269456247688e-05,
      "loss": 0.092,
      "num_input_tokens_seen": 14314752,
      "step": 21885
    },
    {
      "epoch": 11.472746331236896,
      "grad_norm": 3.8900351524353027,
      "learning_rate": 2.2941873769760006e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 14317760,
      "step": 21890
    },
    {
      "epoch": 11.475366876310272,
      "grad_norm": 1.8066694736480713,
      "learning_rate": 2.2930478513808844e-05,
      "loss": 0.099,
      "num_input_tokens_seen": 14321440,
      "step": 21895
    },
    {
      "epoch": 11.477987421383649,
      "grad_norm": 0.42927566170692444,
      "learning_rate": 2.2919083690777976e-05,
      "loss": 0.0748,
      "num_input_tokens_seen": 14324512,
      "step": 21900
    },
    {
      "epoch": 11.480607966457024,
      "grad_norm": 2.690580129623413,
      "learning_rate": 2.2907689303051062e-05,
      "loss": 0.1023,
      "num_input_tokens_seen": 14327648,
      "step": 21905
    },
    {
      "epoch": 11.483228511530399,
      "grad_norm": 4.639955997467041,
      "learning_rate": 2.289629535301167e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 14331040,
      "step": 21910
    },
    {
      "epoch": 11.485849056603774,
      "grad_norm": 2.829500436782837,
      "learning_rate": 2.288490184304329e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 14333920,
      "step": 21915
    },
    {
      "epoch": 11.48846960167715,
      "grad_norm": 1.349851369857788,
      "learning_rate": 2.2873508775529308e-05,
      "loss": 0.0611,
      "num_input_tokens_seen": 14337056,
      "step": 21920
    },
    {
      "epoch": 11.491090146750524,
      "grad_norm": 1.2310917377471924,
      "learning_rate": 2.2862116152853034e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 14339584,
      "step": 21925
    },
    {
      "epoch": 11.4937106918239,
      "grad_norm": 0.5623429417610168,
      "learning_rate": 2.285072397739767e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 14345024,
      "step": 21930
    },
    {
      "epoch": 11.496331236897275,
      "grad_norm": 2.9025115966796875,
      "learning_rate": 2.283933225154634e-05,
      "loss": 0.0473,
      "num_input_tokens_seen": 14348288,
      "step": 21935
    },
    {
      "epoch": 11.49895178197065,
      "grad_norm": 5.820590496063232,
      "learning_rate": 2.2827940977682045e-05,
      "loss": 0.1087,
      "num_input_tokens_seen": 14352000,
      "step": 21940
    },
    {
      "epoch": 11.501572327044025,
      "grad_norm": 2.104787826538086,
      "learning_rate": 2.2816550158187717e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 14354336,
      "step": 21945
    },
    {
      "epoch": 11.5041928721174,
      "grad_norm": 4.255884170532227,
      "learning_rate": 2.2805159795446182e-05,
      "loss": 0.1108,
      "num_input_tokens_seen": 14357728,
      "step": 21950
    },
    {
      "epoch": 11.506813417190775,
      "grad_norm": 2.2386155128479004,
      "learning_rate": 2.2793769891840184e-05,
      "loss": 0.097,
      "num_input_tokens_seen": 14360384,
      "step": 21955
    },
    {
      "epoch": 11.50943396226415,
      "grad_norm": 1.8779245615005493,
      "learning_rate": 2.278238044975236e-05,
      "loss": 0.0951,
      "num_input_tokens_seen": 14364800,
      "step": 21960
    },
    {
      "epoch": 11.512054507337526,
      "grad_norm": 2.3516860008239746,
      "learning_rate": 2.2770991471565236e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 14367872,
      "step": 21965
    },
    {
      "epoch": 11.514675052410901,
      "grad_norm": 0.8574848771095276,
      "learning_rate": 2.2759602959661265e-05,
      "loss": 0.0925,
      "num_input_tokens_seen": 14371872,
      "step": 21970
    },
    {
      "epoch": 11.517295597484276,
      "grad_norm": 3.8316051959991455,
      "learning_rate": 2.2748214916422796e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 14374752,
      "step": 21975
    },
    {
      "epoch": 11.519916142557651,
      "grad_norm": 5.269882678985596,
      "learning_rate": 2.2736827344232082e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 14377792,
      "step": 21980
    },
    {
      "epoch": 11.522536687631026,
      "grad_norm": 1.3503124713897705,
      "learning_rate": 2.272544024547126e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 14382496,
      "step": 21985
    },
    {
      "epoch": 11.525157232704402,
      "grad_norm": 2.086456537246704,
      "learning_rate": 2.27140536225224e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 14389184,
      "step": 21990
    },
    {
      "epoch": 11.527777777777779,
      "grad_norm": 3.1271417140960693,
      "learning_rate": 2.270266747776744e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 14391904,
      "step": 21995
    },
    {
      "epoch": 11.530398322851154,
      "grad_norm": 4.029504776000977,
      "learning_rate": 2.2691281813588226e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 14394880,
      "step": 22000
    },
    {
      "epoch": 11.533018867924529,
      "grad_norm": 2.482280969619751,
      "learning_rate": 2.267989663236652e-05,
      "loss": 0.0875,
      "num_input_tokens_seen": 14399168,
      "step": 22005
    },
    {
      "epoch": 11.535639412997904,
      "grad_norm": 3.3827106952667236,
      "learning_rate": 2.2668511936483975e-05,
      "loss": 0.0707,
      "num_input_tokens_seen": 14402848,
      "step": 22010
    },
    {
      "epoch": 11.53825995807128,
      "grad_norm": 7.324382781982422,
      "learning_rate": 2.2657127728322124e-05,
      "loss": 0.0846,
      "num_input_tokens_seen": 14406272,
      "step": 22015
    },
    {
      "epoch": 11.540880503144654,
      "grad_norm": 4.467774391174316,
      "learning_rate": 2.264574401026244e-05,
      "loss": 0.0582,
      "num_input_tokens_seen": 14409120,
      "step": 22020
    },
    {
      "epoch": 11.54350104821803,
      "grad_norm": 1.9461424350738525,
      "learning_rate": 2.2634360784686242e-05,
      "loss": 0.0813,
      "num_input_tokens_seen": 14412224,
      "step": 22025
    },
    {
      "epoch": 11.546121593291405,
      "grad_norm": 2.3976738452911377,
      "learning_rate": 2.2622978053974776e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 14415616,
      "step": 22030
    },
    {
      "epoch": 11.54874213836478,
      "grad_norm": 4.219113826751709,
      "learning_rate": 2.2611595820509186e-05,
      "loss": 0.1536,
      "num_input_tokens_seen": 14419392,
      "step": 22035
    },
    {
      "epoch": 11.551362683438155,
      "grad_norm": 2.934234857559204,
      "learning_rate": 2.2600214086670494e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 14423040,
      "step": 22040
    },
    {
      "epoch": 11.55398322851153,
      "grad_norm": 4.441740989685059,
      "learning_rate": 2.2588832854839644e-05,
      "loss": 0.1369,
      "num_input_tokens_seen": 14426656,
      "step": 22045
    },
    {
      "epoch": 11.556603773584905,
      "grad_norm": 1.936124324798584,
      "learning_rate": 2.2577452127397463e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 14429792,
      "step": 22050
    },
    {
      "epoch": 11.55922431865828,
      "grad_norm": 1.499586820602417,
      "learning_rate": 2.256607190672464e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 14432992,
      "step": 22055
    },
    {
      "epoch": 11.561844863731656,
      "grad_norm": 0.36637407541275024,
      "learning_rate": 2.2554692195201817e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 14436448,
      "step": 22060
    },
    {
      "epoch": 11.564465408805031,
      "grad_norm": 3.8187854290008545,
      "learning_rate": 2.254331299520948e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 14438880,
      "step": 22065
    },
    {
      "epoch": 11.567085953878406,
      "grad_norm": 6.0192084312438965,
      "learning_rate": 2.253193430912805e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 14441248,
      "step": 22070
    },
    {
      "epoch": 11.569706498951781,
      "grad_norm": 2.9849612712860107,
      "learning_rate": 2.2520556139337803e-05,
      "loss": 0.0656,
      "num_input_tokens_seen": 14444576,
      "step": 22075
    },
    {
      "epoch": 11.572327044025156,
      "grad_norm": 2.3249759674072266,
      "learning_rate": 2.250917848821892e-05,
      "loss": 0.0763,
      "num_input_tokens_seen": 14447968,
      "step": 22080
    },
    {
      "epoch": 11.574947589098532,
      "grad_norm": 1.931740641593933,
      "learning_rate": 2.2497801358151472e-05,
      "loss": 0.1052,
      "num_input_tokens_seen": 14450816,
      "step": 22085
    },
    {
      "epoch": 11.577568134171909,
      "grad_norm": 1.1835132837295532,
      "learning_rate": 2.248642475151545e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 14453728,
      "step": 22090
    },
    {
      "epoch": 11.580188679245284,
      "grad_norm": 1.9501041173934937,
      "learning_rate": 2.2475048670690678e-05,
      "loss": 0.1936,
      "num_input_tokens_seen": 14456512,
      "step": 22095
    },
    {
      "epoch": 11.582809224318659,
      "grad_norm": 2.166278600692749,
      "learning_rate": 2.2463673118056927e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 14460160,
      "step": 22100
    },
    {
      "epoch": 11.585429769392034,
      "grad_norm": 2.2390658855438232,
      "learning_rate": 2.245229809599383e-05,
      "loss": 0.0741,
      "num_input_tokens_seen": 14462848,
      "step": 22105
    },
    {
      "epoch": 11.58805031446541,
      "grad_norm": 1.0661617517471313,
      "learning_rate": 2.2440923606880896e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 14466592,
      "step": 22110
    },
    {
      "epoch": 11.590670859538784,
      "grad_norm": 0.4629058837890625,
      "learning_rate": 2.242954965309755e-05,
      "loss": 0.065,
      "num_input_tokens_seen": 14468960,
      "step": 22115
    },
    {
      "epoch": 11.59329140461216,
      "grad_norm": 1.5452672243118286,
      "learning_rate": 2.2418176237023088e-05,
      "loss": 0.0754,
      "num_input_tokens_seen": 14472864,
      "step": 22120
    },
    {
      "epoch": 11.595911949685535,
      "grad_norm": 1.8896870613098145,
      "learning_rate": 2.2406803361036706e-05,
      "loss": 0.0793,
      "num_input_tokens_seen": 14476768,
      "step": 22125
    },
    {
      "epoch": 11.59853249475891,
      "grad_norm": 3.440314769744873,
      "learning_rate": 2.2395431027517475e-05,
      "loss": 0.0735,
      "num_input_tokens_seen": 14479872,
      "step": 22130
    },
    {
      "epoch": 11.601153039832285,
      "grad_norm": 1.8205572366714478,
      "learning_rate": 2.2384059238844364e-05,
      "loss": 0.0872,
      "num_input_tokens_seen": 14482400,
      "step": 22135
    },
    {
      "epoch": 11.60377358490566,
      "grad_norm": 2.8455650806427,
      "learning_rate": 2.2372687997396215e-05,
      "loss": 0.0758,
      "num_input_tokens_seen": 14485696,
      "step": 22140
    },
    {
      "epoch": 11.606394129979035,
      "grad_norm": 3.279855728149414,
      "learning_rate": 2.2361317305551756e-05,
      "loss": 0.1186,
      "num_input_tokens_seen": 14488800,
      "step": 22145
    },
    {
      "epoch": 11.60901467505241,
      "grad_norm": 3.938203811645508,
      "learning_rate": 2.2349947165689615e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 14491520,
      "step": 22150
    },
    {
      "epoch": 11.611635220125786,
      "grad_norm": 1.0709524154663086,
      "learning_rate": 2.2338577580188296e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 14494176,
      "step": 22155
    },
    {
      "epoch": 11.614255765199161,
      "grad_norm": 2.384812831878662,
      "learning_rate": 2.2327208551426188e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 14496640,
      "step": 22160
    },
    {
      "epoch": 11.616876310272536,
      "grad_norm": 4.821502685546875,
      "learning_rate": 2.2315840081781546e-05,
      "loss": 0.0883,
      "num_input_tokens_seen": 14499232,
      "step": 22165
    },
    {
      "epoch": 11.619496855345911,
      "grad_norm": 3.656668186187744,
      "learning_rate": 2.230447217363255e-05,
      "loss": 0.0637,
      "num_input_tokens_seen": 14501600,
      "step": 22170
    },
    {
      "epoch": 11.622117400419286,
      "grad_norm": 2.5646121501922607,
      "learning_rate": 2.2293104829357212e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 14504544,
      "step": 22175
    },
    {
      "epoch": 11.624737945492662,
      "grad_norm": 4.919783592224121,
      "learning_rate": 2.2281738051333467e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 14507904,
      "step": 22180
    },
    {
      "epoch": 11.627358490566039,
      "grad_norm": 1.8910489082336426,
      "learning_rate": 2.2270371841939105e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 14511264,
      "step": 22185
    },
    {
      "epoch": 11.629979035639414,
      "grad_norm": 2.3326711654663086,
      "learning_rate": 2.2259006203551824e-05,
      "loss": 0.0665,
      "num_input_tokens_seen": 14514752,
      "step": 22190
    },
    {
      "epoch": 11.632599580712789,
      "grad_norm": 1.7866597175598145,
      "learning_rate": 2.2247641138549168e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 14517312,
      "step": 22195
    },
    {
      "epoch": 11.635220125786164,
      "grad_norm": 8.086155891418457,
      "learning_rate": 2.2236276649308583e-05,
      "loss": 0.0804,
      "num_input_tokens_seen": 14520800,
      "step": 22200
    },
    {
      "epoch": 11.63784067085954,
      "grad_norm": 7.407040119171143,
      "learning_rate": 2.2224912738207395e-05,
      "loss": 0.0515,
      "num_input_tokens_seen": 14524000,
      "step": 22205
    },
    {
      "epoch": 11.640461215932914,
      "grad_norm": 4.755507469177246,
      "learning_rate": 2.2213549407622796e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 14526944,
      "step": 22210
    },
    {
      "epoch": 11.64308176100629,
      "grad_norm": 1.5119118690490723,
      "learning_rate": 2.220218665993188e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 14530368,
      "step": 22215
    },
    {
      "epoch": 11.645702306079665,
      "grad_norm": 3.328733205795288,
      "learning_rate": 2.21908244975116e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 14533824,
      "step": 22220
    },
    {
      "epoch": 11.64832285115304,
      "grad_norm": 2.9856526851654053,
      "learning_rate": 2.2179462922738776e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 14539232,
      "step": 22225
    },
    {
      "epoch": 11.650943396226415,
      "grad_norm": 1.337234616279602,
      "learning_rate": 2.2168101937990135e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 14542784,
      "step": 22230
    },
    {
      "epoch": 11.65356394129979,
      "grad_norm": 7.370135307312012,
      "learning_rate": 2.2156741545642253e-05,
      "loss": 0.0877,
      "num_input_tokens_seen": 14545696,
      "step": 22235
    },
    {
      "epoch": 11.656184486373165,
      "grad_norm": 3.136749505996704,
      "learning_rate": 2.2145381748071606e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 14549344,
      "step": 22240
    },
    {
      "epoch": 11.65880503144654,
      "grad_norm": 2.754157781600952,
      "learning_rate": 2.2134022547654538e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 14552096,
      "step": 22245
    },
    {
      "epoch": 11.661425576519916,
      "grad_norm": 2.1412720680236816,
      "learning_rate": 2.212266394676725e-05,
      "loss": 0.064,
      "num_input_tokens_seen": 14556384,
      "step": 22250
    },
    {
      "epoch": 11.664046121593291,
      "grad_norm": 1.887702465057373,
      "learning_rate": 2.2111305947785836e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 14559680,
      "step": 22255
    },
    {
      "epoch": 11.666666666666666,
      "grad_norm": 2.577284336090088,
      "learning_rate": 2.2099948553086263e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 14562592,
      "step": 22260
    },
    {
      "epoch": 11.669287211740041,
      "grad_norm": 1.6951123476028442,
      "learning_rate": 2.2088591765044365e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 14564960,
      "step": 22265
    },
    {
      "epoch": 11.671907756813416,
      "grad_norm": 3.1775383949279785,
      "learning_rate": 2.207723558603586e-05,
      "loss": 0.1105,
      "num_input_tokens_seen": 14568384,
      "step": 22270
    },
    {
      "epoch": 11.674528301886792,
      "grad_norm": 1.695327639579773,
      "learning_rate": 2.206588001843633e-05,
      "loss": 0.0753,
      "num_input_tokens_seen": 14571680,
      "step": 22275
    },
    {
      "epoch": 11.677148846960169,
      "grad_norm": 2.3922836780548096,
      "learning_rate": 2.2054525064621216e-05,
      "loss": 0.1019,
      "num_input_tokens_seen": 14575136,
      "step": 22280
    },
    {
      "epoch": 11.679769392033544,
      "grad_norm": 1.9690519571304321,
      "learning_rate": 2.2043170726965858e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 14577888,
      "step": 22285
    },
    {
      "epoch": 11.682389937106919,
      "grad_norm": 1.9688336849212646,
      "learning_rate": 2.2031817007845447e-05,
      "loss": 0.0651,
      "num_input_tokens_seen": 14581248,
      "step": 22290
    },
    {
      "epoch": 11.685010482180294,
      "grad_norm": 3.4091062545776367,
      "learning_rate": 2.202046390963506e-05,
      "loss": 0.0924,
      "num_input_tokens_seen": 14584928,
      "step": 22295
    },
    {
      "epoch": 11.68763102725367,
      "grad_norm": 0.9061406850814819,
      "learning_rate": 2.2009111434709627e-05,
      "loss": 0.0928,
      "num_input_tokens_seen": 14587552,
      "step": 22300
    },
    {
      "epoch": 11.690251572327044,
      "grad_norm": 5.520125389099121,
      "learning_rate": 2.1997759585443968e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 14589760,
      "step": 22305
    },
    {
      "epoch": 11.69287211740042,
      "grad_norm": 1.6357083320617676,
      "learning_rate": 2.198640836421275e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 14592448,
      "step": 22310
    },
    {
      "epoch": 11.695492662473795,
      "grad_norm": 2.939089298248291,
      "learning_rate": 2.1975057773390515e-05,
      "loss": 0.0785,
      "num_input_tokens_seen": 14595808,
      "step": 22315
    },
    {
      "epoch": 11.69811320754717,
      "grad_norm": 2.0255792140960693,
      "learning_rate": 2.1963707815351683e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 14598624,
      "step": 22320
    },
    {
      "epoch": 11.700733752620545,
      "grad_norm": 3.6645848751068115,
      "learning_rate": 2.1952358492470544e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 14601536,
      "step": 22325
    },
    {
      "epoch": 11.70335429769392,
      "grad_norm": 7.690096855163574,
      "learning_rate": 2.194100980712124e-05,
      "loss": 0.0524,
      "num_input_tokens_seen": 14604736,
      "step": 22330
    },
    {
      "epoch": 11.705974842767295,
      "grad_norm": 1.4811666011810303,
      "learning_rate": 2.1929661761677774e-05,
      "loss": 0.072,
      "num_input_tokens_seen": 14607840,
      "step": 22335
    },
    {
      "epoch": 11.70859538784067,
      "grad_norm": 2.853278636932373,
      "learning_rate": 2.191831435851405e-05,
      "loss": 0.0947,
      "num_input_tokens_seen": 14610816,
      "step": 22340
    },
    {
      "epoch": 11.711215932914046,
      "grad_norm": 5.719001293182373,
      "learning_rate": 2.1906967600003793e-05,
      "loss": 0.103,
      "num_input_tokens_seen": 14614464,
      "step": 22345
    },
    {
      "epoch": 11.713836477987421,
      "grad_norm": 2.438734531402588,
      "learning_rate": 2.1895621488520636e-05,
      "loss": 0.0532,
      "num_input_tokens_seen": 14616704,
      "step": 22350
    },
    {
      "epoch": 11.716457023060796,
      "grad_norm": 1.330259919166565,
      "learning_rate": 2.1884276026438044e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 14620832,
      "step": 22355
    },
    {
      "epoch": 11.719077568134171,
      "grad_norm": 2.710670232772827,
      "learning_rate": 2.187293121612937e-05,
      "loss": 0.0687,
      "num_input_tokens_seen": 14623328,
      "step": 22360
    },
    {
      "epoch": 11.721698113207546,
      "grad_norm": 3.3244314193725586,
      "learning_rate": 2.1861587059967804e-05,
      "loss": 0.0986,
      "num_input_tokens_seen": 14627520,
      "step": 22365
    },
    {
      "epoch": 11.724318658280922,
      "grad_norm": 2.1293022632598877,
      "learning_rate": 2.1850243560326414e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 14631712,
      "step": 22370
    },
    {
      "epoch": 11.726939203354299,
      "grad_norm": 3.500028371810913,
      "learning_rate": 2.1838900719578143e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 14634912,
      "step": 22375
    },
    {
      "epoch": 11.729559748427674,
      "grad_norm": 1.2442375421524048,
      "learning_rate": 2.1827558540095772e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 14637440,
      "step": 22380
    },
    {
      "epoch": 11.732180293501049,
      "grad_norm": 1.1733587980270386,
      "learning_rate": 2.1816217024251966e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 14642336,
      "step": 22385
    },
    {
      "epoch": 11.734800838574424,
      "grad_norm": 1.4675368070602417,
      "learning_rate": 2.1804876174419242e-05,
      "loss": 0.1068,
      "num_input_tokens_seen": 14646016,
      "step": 22390
    },
    {
      "epoch": 11.7374213836478,
      "grad_norm": 3.2395808696746826,
      "learning_rate": 2.179353599296996e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 14648384,
      "step": 22395
    },
    {
      "epoch": 11.740041928721174,
      "grad_norm": 2.532921552658081,
      "learning_rate": 2.1782196482276368e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 14651552,
      "step": 22400
    },
    {
      "epoch": 11.74266247379455,
      "grad_norm": 2.990128517150879,
      "learning_rate": 2.1770857644710567e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 14654304,
      "step": 22405
    },
    {
      "epoch": 11.745283018867925,
      "grad_norm": 6.492458820343018,
      "learning_rate": 2.1759519482644495e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 14657472,
      "step": 22410
    },
    {
      "epoch": 11.7479035639413,
      "grad_norm": 2.950727939605713,
      "learning_rate": 2.174818199844999e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 14660992,
      "step": 22415
    },
    {
      "epoch": 11.750524109014675,
      "grad_norm": 3.0773496627807617,
      "learning_rate": 2.173684519449872e-05,
      "loss": 0.0981,
      "num_input_tokens_seen": 14665344,
      "step": 22420
    },
    {
      "epoch": 11.75314465408805,
      "grad_norm": 2.8646395206451416,
      "learning_rate": 2.1725509073162188e-05,
      "loss": 0.1037,
      "num_input_tokens_seen": 14668384,
      "step": 22425
    },
    {
      "epoch": 11.755765199161425,
      "grad_norm": 5.513684272766113,
      "learning_rate": 2.171417363681181e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 14671488,
      "step": 22430
    },
    {
      "epoch": 11.7583857442348,
      "grad_norm": 2.6103713512420654,
      "learning_rate": 2.1702838887818816e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 14674592,
      "step": 22435
    },
    {
      "epoch": 11.761006289308176,
      "grad_norm": 1.324309229850769,
      "learning_rate": 2.1691504828554315e-05,
      "loss": 0.1072,
      "num_input_tokens_seen": 14678176,
      "step": 22440
    },
    {
      "epoch": 11.76362683438155,
      "grad_norm": 4.457544326782227,
      "learning_rate": 2.1680171461389265e-05,
      "loss": 0.0787,
      "num_input_tokens_seen": 14681184,
      "step": 22445
    },
    {
      "epoch": 11.766247379454926,
      "grad_norm": 1.122814655303955,
      "learning_rate": 2.166883878869446e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 14684192,
      "step": 22450
    },
    {
      "epoch": 11.768867924528301,
      "grad_norm": 2.5071280002593994,
      "learning_rate": 2.1657506812840582e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 14687040,
      "step": 22455
    },
    {
      "epoch": 11.771488469601676,
      "grad_norm": 1.8449827432632446,
      "learning_rate": 2.164617553619814e-05,
      "loss": 0.0634,
      "num_input_tokens_seen": 14690592,
      "step": 22460
    },
    {
      "epoch": 11.774109014675052,
      "grad_norm": 1.415817141532898,
      "learning_rate": 2.1634844961137517e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 14693792,
      "step": 22465
    },
    {
      "epoch": 11.776729559748428,
      "grad_norm": 7.3889594078063965,
      "learning_rate": 2.162351509002893e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 14696544,
      "step": 22470
    },
    {
      "epoch": 11.779350104821804,
      "grad_norm": 2.4966602325439453,
      "learning_rate": 2.161218592524248e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 14700416,
      "step": 22475
    },
    {
      "epoch": 11.781970649895179,
      "grad_norm": 2.542381525039673,
      "learning_rate": 2.160085746914807e-05,
      "loss": 0.129,
      "num_input_tokens_seen": 14703648,
      "step": 22480
    },
    {
      "epoch": 11.784591194968554,
      "grad_norm": 1.831958293914795,
      "learning_rate": 2.1589529724115503e-05,
      "loss": 0.0921,
      "num_input_tokens_seen": 14706304,
      "step": 22485
    },
    {
      "epoch": 11.78721174004193,
      "grad_norm": 3.2471914291381836,
      "learning_rate": 2.1578202692514407e-05,
      "loss": 0.0964,
      "num_input_tokens_seen": 14708672,
      "step": 22490
    },
    {
      "epoch": 11.789832285115304,
      "grad_norm": 1.8373726606369019,
      "learning_rate": 2.1566876376714274e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 14711904,
      "step": 22495
    },
    {
      "epoch": 11.79245283018868,
      "grad_norm": 2.6874895095825195,
      "learning_rate": 2.155555077908443e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 14714080,
      "step": 22500
    },
    {
      "epoch": 11.795073375262055,
      "grad_norm": 1.5496834516525269,
      "learning_rate": 2.1544225901994077e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 14717152,
      "step": 22505
    },
    {
      "epoch": 11.79769392033543,
      "grad_norm": 1.4332855939865112,
      "learning_rate": 2.1532901747812234e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 14720064,
      "step": 22510
    },
    {
      "epoch": 11.800314465408805,
      "grad_norm": 3.177274465560913,
      "learning_rate": 2.1521578318907787e-05,
      "loss": 0.0806,
      "num_input_tokens_seen": 14722848,
      "step": 22515
    },
    {
      "epoch": 11.80293501048218,
      "grad_norm": 1.3621999025344849,
      "learning_rate": 2.1510255617649476e-05,
      "loss": 0.0759,
      "num_input_tokens_seen": 14725696,
      "step": 22520
    },
    {
      "epoch": 11.805555555555555,
      "grad_norm": 5.752293109893799,
      "learning_rate": 2.149893364640587e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 14728800,
      "step": 22525
    },
    {
      "epoch": 11.80817610062893,
      "grad_norm": 1.4751592874526978,
      "learning_rate": 2.1487612407545417e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 14732608,
      "step": 22530
    },
    {
      "epoch": 11.810796645702306,
      "grad_norm": 1.9588772058486938,
      "learning_rate": 2.147629190343637e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 14736640,
      "step": 22535
    },
    {
      "epoch": 11.81341719077568,
      "grad_norm": 3.0276196002960205,
      "learning_rate": 2.1464972136446852e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 14739296,
      "step": 22540
    },
    {
      "epoch": 11.816037735849056,
      "grad_norm": 0.9736467599868774,
      "learning_rate": 2.1453653108944838e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 14742048,
      "step": 22545
    },
    {
      "epoch": 11.818658280922431,
      "grad_norm": 0.6784443259239197,
      "learning_rate": 2.1442334823298128e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 14745056,
      "step": 22550
    },
    {
      "epoch": 11.821278825995806,
      "grad_norm": 2.4236276149749756,
      "learning_rate": 2.1431017281874386e-05,
      "loss": 0.1927,
      "num_input_tokens_seen": 14748064,
      "step": 22555
    },
    {
      "epoch": 11.823899371069182,
      "grad_norm": 2.153506278991699,
      "learning_rate": 2.1419700487041123e-05,
      "loss": 0.0778,
      "num_input_tokens_seen": 14751264,
      "step": 22560
    },
    {
      "epoch": 11.826519916142558,
      "grad_norm": 3.3025503158569336,
      "learning_rate": 2.140838444116566e-05,
      "loss": 0.0941,
      "num_input_tokens_seen": 14755168,
      "step": 22565
    },
    {
      "epoch": 11.829140461215934,
      "grad_norm": 2.228379011154175,
      "learning_rate": 2.13970691466152e-05,
      "loss": 0.0614,
      "num_input_tokens_seen": 14757408,
      "step": 22570
    },
    {
      "epoch": 11.831761006289309,
      "grad_norm": 2.0434489250183105,
      "learning_rate": 2.1385754605756773e-05,
      "loss": 0.0736,
      "num_input_tokens_seen": 14760288,
      "step": 22575
    },
    {
      "epoch": 11.834381551362684,
      "grad_norm": 1.9739246368408203,
      "learning_rate": 2.137444082095724e-05,
      "loss": 0.0874,
      "num_input_tokens_seen": 14763616,
      "step": 22580
    },
    {
      "epoch": 11.83700209643606,
      "grad_norm": 0.7198569178581238,
      "learning_rate": 2.1363127794583335e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 14768544,
      "step": 22585
    },
    {
      "epoch": 11.839622641509434,
      "grad_norm": 3.7281477451324463,
      "learning_rate": 2.1351815529001605e-05,
      "loss": 0.0905,
      "num_input_tokens_seen": 14771456,
      "step": 22590
    },
    {
      "epoch": 11.84224318658281,
      "grad_norm": 2.013056755065918,
      "learning_rate": 2.1340504026578438e-05,
      "loss": 0.153,
      "num_input_tokens_seen": 14774880,
      "step": 22595
    },
    {
      "epoch": 11.844863731656185,
      "grad_norm": 2.4162001609802246,
      "learning_rate": 2.1329193289680084e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 14778432,
      "step": 22600
    },
    {
      "epoch": 11.84748427672956,
      "grad_norm": 3.3848180770874023,
      "learning_rate": 2.1317883320672608e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 14781984,
      "step": 22605
    },
    {
      "epoch": 11.850104821802935,
      "grad_norm": 4.662182807922363,
      "learning_rate": 2.1306574121921942e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 14785280,
      "step": 22610
    },
    {
      "epoch": 11.85272536687631,
      "grad_norm": 18.829204559326172,
      "learning_rate": 2.1295265695793836e-05,
      "loss": 0.1686,
      "num_input_tokens_seen": 14789568,
      "step": 22615
    },
    {
      "epoch": 11.855345911949685,
      "grad_norm": 6.5331597328186035,
      "learning_rate": 2.128395804465387e-05,
      "loss": 0.0533,
      "num_input_tokens_seen": 14792608,
      "step": 22620
    },
    {
      "epoch": 11.85796645702306,
      "grad_norm": 0.7717723250389099,
      "learning_rate": 2.1272651170867487e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 14797760,
      "step": 22625
    },
    {
      "epoch": 11.860587002096436,
      "grad_norm": 2.1552987098693848,
      "learning_rate": 2.1261345076799955e-05,
      "loss": 0.0525,
      "num_input_tokens_seen": 14800768,
      "step": 22630
    },
    {
      "epoch": 11.86320754716981,
      "grad_norm": 1.278619647026062,
      "learning_rate": 2.125003976481638e-05,
      "loss": 0.0898,
      "num_input_tokens_seen": 14804000,
      "step": 22635
    },
    {
      "epoch": 11.865828092243186,
      "grad_norm": 3.3543715476989746,
      "learning_rate": 2.123873523728171e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 14807392,
      "step": 22640
    },
    {
      "epoch": 11.868448637316561,
      "grad_norm": 12.247469902038574,
      "learning_rate": 2.122743149656072e-05,
      "loss": 0.0603,
      "num_input_tokens_seen": 14810816,
      "step": 22645
    },
    {
      "epoch": 11.871069182389936,
      "grad_norm": 1.121427059173584,
      "learning_rate": 2.1216128545018012e-05,
      "loss": 0.0403,
      "num_input_tokens_seen": 14813056,
      "step": 22650
    },
    {
      "epoch": 11.873689727463312,
      "grad_norm": 2.947744846343994,
      "learning_rate": 2.1204826385018052e-05,
      "loss": 0.0903,
      "num_input_tokens_seen": 14816736,
      "step": 22655
    },
    {
      "epoch": 11.876310272536688,
      "grad_norm": 2.0612759590148926,
      "learning_rate": 2.1193525018925106e-05,
      "loss": 0.1064,
      "num_input_tokens_seen": 14820160,
      "step": 22660
    },
    {
      "epoch": 11.878930817610064,
      "grad_norm": 3.5501811504364014,
      "learning_rate": 2.1182224449103304e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 14822944,
      "step": 22665
    },
    {
      "epoch": 11.881551362683439,
      "grad_norm": 2.6329092979431152,
      "learning_rate": 2.117092467791659e-05,
      "loss": 0.0606,
      "num_input_tokens_seen": 14826464,
      "step": 22670
    },
    {
      "epoch": 11.884171907756814,
      "grad_norm": 2.0911293029785156,
      "learning_rate": 2.115962570772876e-05,
      "loss": 0.0724,
      "num_input_tokens_seen": 14830080,
      "step": 22675
    },
    {
      "epoch": 11.88679245283019,
      "grad_norm": 1.704129934310913,
      "learning_rate": 2.1148327540903408e-05,
      "loss": 0.067,
      "num_input_tokens_seen": 14833280,
      "step": 22680
    },
    {
      "epoch": 11.889412997903564,
      "grad_norm": 3.066148519515991,
      "learning_rate": 2.113703017980399e-05,
      "loss": 0.1277,
      "num_input_tokens_seen": 14836768,
      "step": 22685
    },
    {
      "epoch": 11.89203354297694,
      "grad_norm": 1.9948211908340454,
      "learning_rate": 2.1125733626793793e-05,
      "loss": 0.0974,
      "num_input_tokens_seen": 14839456,
      "step": 22690
    },
    {
      "epoch": 11.894654088050315,
      "grad_norm": 1.4192264080047607,
      "learning_rate": 2.1114437884235915e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 14842272,
      "step": 22695
    },
    {
      "epoch": 11.89727463312369,
      "grad_norm": 1.9743928909301758,
      "learning_rate": 2.1103142954493313e-05,
      "loss": 0.1718,
      "num_input_tokens_seen": 14846304,
      "step": 22700
    },
    {
      "epoch": 11.899895178197065,
      "grad_norm": 2.252007484436035,
      "learning_rate": 2.1091848839928745e-05,
      "loss": 0.0577,
      "num_input_tokens_seen": 14849184,
      "step": 22705
    },
    {
      "epoch": 11.90251572327044,
      "grad_norm": 2.7251625061035156,
      "learning_rate": 2.1080555542904805e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 14852704,
      "step": 22710
    },
    {
      "epoch": 11.905136268343815,
      "grad_norm": 3.0975921154022217,
      "learning_rate": 2.106926306578393e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 14856288,
      "step": 22715
    },
    {
      "epoch": 11.90775681341719,
      "grad_norm": 2.922549247741699,
      "learning_rate": 2.1057971410928383e-05,
      "loss": 0.0779,
      "num_input_tokens_seen": 14859232,
      "step": 22720
    },
    {
      "epoch": 11.910377358490566,
      "grad_norm": 2.224548101425171,
      "learning_rate": 2.104668058070024e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 14862624,
      "step": 22725
    },
    {
      "epoch": 11.91299790356394,
      "grad_norm": 1.4298288822174072,
      "learning_rate": 2.1035390577461427e-05,
      "loss": 0.0827,
      "num_input_tokens_seen": 14866368,
      "step": 22730
    },
    {
      "epoch": 11.915618448637316,
      "grad_norm": 3.1869888305664062,
      "learning_rate": 2.102410140357367e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 14869024,
      "step": 22735
    },
    {
      "epoch": 11.918238993710691,
      "grad_norm": 17.240238189697266,
      "learning_rate": 2.1012813061398534e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 14871936,
      "step": 22740
    },
    {
      "epoch": 11.920859538784066,
      "grad_norm": 5.635097980499268,
      "learning_rate": 2.1001525553297422e-05,
      "loss": 0.1339,
      "num_input_tokens_seen": 14874400,
      "step": 22745
    },
    {
      "epoch": 11.923480083857442,
      "grad_norm": 17.00337791442871,
      "learning_rate": 2.099023888163154e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 14876864,
      "step": 22750
    },
    {
      "epoch": 11.926100628930818,
      "grad_norm": 2.323763132095337,
      "learning_rate": 2.0978953048761947e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 14879872,
      "step": 22755
    },
    {
      "epoch": 11.928721174004194,
      "grad_norm": 3.334259033203125,
      "learning_rate": 2.0967668057049507e-05,
      "loss": 0.0703,
      "num_input_tokens_seen": 14883424,
      "step": 22760
    },
    {
      "epoch": 11.931341719077569,
      "grad_norm": 2.933220148086548,
      "learning_rate": 2.0956383908854895e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 14887392,
      "step": 22765
    },
    {
      "epoch": 11.933962264150944,
      "grad_norm": 3.4957361221313477,
      "learning_rate": 2.0945100606538644e-05,
      "loss": 0.115,
      "num_input_tokens_seen": 14890688,
      "step": 22770
    },
    {
      "epoch": 11.93658280922432,
      "grad_norm": 3.4711248874664307,
      "learning_rate": 2.0933818152461077e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 14894464,
      "step": 22775
    },
    {
      "epoch": 11.939203354297694,
      "grad_norm": 1.132396936416626,
      "learning_rate": 2.092253654898237e-05,
      "loss": 0.0823,
      "num_input_tokens_seen": 14897024,
      "step": 22780
    },
    {
      "epoch": 11.94182389937107,
      "grad_norm": 2.4788150787353516,
      "learning_rate": 2.0911255798462494e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 14900480,
      "step": 22785
    },
    {
      "epoch": 11.944444444444445,
      "grad_norm": 4.207056999206543,
      "learning_rate": 2.0899975903261266e-05,
      "loss": 0.1178,
      "num_input_tokens_seen": 14903136,
      "step": 22790
    },
    {
      "epoch": 11.94706498951782,
      "grad_norm": 4.326562881469727,
      "learning_rate": 2.0888696865738295e-05,
      "loss": 0.0662,
      "num_input_tokens_seen": 14907040,
      "step": 22795
    },
    {
      "epoch": 11.949685534591195,
      "grad_norm": 5.544569492340088,
      "learning_rate": 2.087741868825304e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 14910304,
      "step": 22800
    },
    {
      "epoch": 11.95230607966457,
      "grad_norm": 1.2679191827774048,
      "learning_rate": 2.0866141373164758e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 14913312,
      "step": 22805
    },
    {
      "epoch": 11.954926624737945,
      "grad_norm": 2.5580270290374756,
      "learning_rate": 2.0854864922832544e-05,
      "loss": 0.0889,
      "num_input_tokens_seen": 14916512,
      "step": 22810
    },
    {
      "epoch": 11.95754716981132,
      "grad_norm": 1.6774295568466187,
      "learning_rate": 2.0843589339615304e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 14920416,
      "step": 22815
    },
    {
      "epoch": 11.960167714884696,
      "grad_norm": 2.2426023483276367,
      "learning_rate": 2.0832314625871742e-05,
      "loss": 0.0868,
      "num_input_tokens_seen": 14923840,
      "step": 22820
    },
    {
      "epoch": 11.96278825995807,
      "grad_norm": 2.450378894805908,
      "learning_rate": 2.0821040783960423e-05,
      "loss": 0.1033,
      "num_input_tokens_seen": 14927520,
      "step": 22825
    },
    {
      "epoch": 11.965408805031446,
      "grad_norm": 2.1135356426239014,
      "learning_rate": 2.0809767816239686e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 14930240,
      "step": 22830
    },
    {
      "epoch": 11.968029350104821,
      "grad_norm": 2.294041872024536,
      "learning_rate": 2.0798495725067725e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 14932928,
      "step": 22835
    },
    {
      "epoch": 11.970649895178196,
      "grad_norm": 4.074471473693848,
      "learning_rate": 2.078722451280252e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 14935616,
      "step": 22840
    },
    {
      "epoch": 11.973270440251572,
      "grad_norm": 8.274648666381836,
      "learning_rate": 2.0775954181801893e-05,
      "loss": 0.0876,
      "num_input_tokens_seen": 14938080,
      "step": 22845
    },
    {
      "epoch": 11.975890985324948,
      "grad_norm": 8.565385818481445,
      "learning_rate": 2.0764684734423455e-05,
      "loss": 0.1284,
      "num_input_tokens_seen": 14940320,
      "step": 22850
    },
    {
      "epoch": 11.978511530398324,
      "grad_norm": 2.5818514823913574,
      "learning_rate": 2.075341617302465e-05,
      "loss": 0.0812,
      "num_input_tokens_seen": 14944000,
      "step": 22855
    },
    {
      "epoch": 11.981132075471699,
      "grad_norm": 1.1761494874954224,
      "learning_rate": 2.074214849996274e-05,
      "loss": 0.0632,
      "num_input_tokens_seen": 14947168,
      "step": 22860
    },
    {
      "epoch": 11.983752620545074,
      "grad_norm": 2.8859379291534424,
      "learning_rate": 2.0730881717594776e-05,
      "loss": 0.0878,
      "num_input_tokens_seen": 14950592,
      "step": 22865
    },
    {
      "epoch": 11.98637316561845,
      "grad_norm": 2.3260860443115234,
      "learning_rate": 2.0719615828277657e-05,
      "loss": 0.0996,
      "num_input_tokens_seen": 14954176,
      "step": 22870
    },
    {
      "epoch": 11.988993710691824,
      "grad_norm": 3.7402758598327637,
      "learning_rate": 2.0708350834368083e-05,
      "loss": 0.0954,
      "num_input_tokens_seen": 14957248,
      "step": 22875
    },
    {
      "epoch": 11.9916142557652,
      "grad_norm": 3.106738328933716,
      "learning_rate": 2.069708673822255e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 14960640,
      "step": 22880
    },
    {
      "epoch": 11.994234800838575,
      "grad_norm": 1.0814305543899536,
      "learning_rate": 2.068582354219737e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 14963392,
      "step": 22885
    },
    {
      "epoch": 11.99685534591195,
      "grad_norm": 2.0518128871917725,
      "learning_rate": 2.0674561248648697e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 14966528,
      "step": 22890
    },
    {
      "epoch": 11.999475890985325,
      "grad_norm": 2.3900232315063477,
      "learning_rate": 2.0663299859932456e-05,
      "loss": 0.0829,
      "num_input_tokens_seen": 14970464,
      "step": 22895
    },
    {
      "epoch": 12.0,
      "eval_loss": 1.0327281951904297,
      "eval_runtime": 11.677,
      "eval_samples_per_second": 72.621,
      "eval_steps_per_second": 18.155,
      "num_input_tokens_seen": 14970528,
      "step": 22896
    },
    {
      "epoch": 12.0020964360587,
      "grad_norm": 0.427877277135849,
      "learning_rate": 2.065203937840442e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 14973600,
      "step": 22900
    },
    {
      "epoch": 12.004716981132075,
      "grad_norm": 0.9505634307861328,
      "learning_rate": 2.0640779806420136e-05,
      "loss": 0.0689,
      "num_input_tokens_seen": 14977536,
      "step": 22905
    },
    {
      "epoch": 12.00733752620545,
      "grad_norm": 1.3920342922210693,
      "learning_rate": 2.0629521146334978e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 14980640,
      "step": 22910
    },
    {
      "epoch": 12.009958071278826,
      "grad_norm": 1.4482178688049316,
      "learning_rate": 2.061826340050414e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 14983456,
      "step": 22915
    },
    {
      "epoch": 12.0125786163522,
      "grad_norm": 3.1881415843963623,
      "learning_rate": 2.0607006571282605e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 14987328,
      "step": 22920
    },
    {
      "epoch": 12.015199161425576,
      "grad_norm": 7.478388786315918,
      "learning_rate": 2.0595750661025183e-05,
      "loss": 0.05,
      "num_input_tokens_seen": 14990432,
      "step": 22925
    },
    {
      "epoch": 12.017819706498951,
      "grad_norm": 5.09908390045166,
      "learning_rate": 2.0584495672086483e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 14993824,
      "step": 22930
    },
    {
      "epoch": 12.020440251572326,
      "grad_norm": 0.9571104645729065,
      "learning_rate": 2.05732416068209e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 14996608,
      "step": 22935
    },
    {
      "epoch": 12.023060796645701,
      "grad_norm": 2.684094190597534,
      "learning_rate": 2.0561988467582676e-05,
      "loss": 0.0341,
      "num_input_tokens_seen": 15000128,
      "step": 22940
    },
    {
      "epoch": 12.025681341719078,
      "grad_norm": 2.230241537094116,
      "learning_rate": 2.0550736256725828e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 15003104,
      "step": 22945
    },
    {
      "epoch": 12.028301886792454,
      "grad_norm": 2.3928627967834473,
      "learning_rate": 2.0539484976604196e-05,
      "loss": 0.0228,
      "num_input_tokens_seen": 15005632,
      "step": 22950
    },
    {
      "epoch": 12.030922431865829,
      "grad_norm": 1.3312894105911255,
      "learning_rate": 2.052823462957142e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 15009024,
      "step": 22955
    },
    {
      "epoch": 12.033542976939204,
      "grad_norm": 2.1725869178771973,
      "learning_rate": 2.0516985217980953e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 15011680,
      "step": 22960
    },
    {
      "epoch": 12.036163522012579,
      "grad_norm": 3.119776487350464,
      "learning_rate": 2.050573674418602e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 15014656,
      "step": 22965
    },
    {
      "epoch": 12.038784067085954,
      "grad_norm": 1.094604730606079,
      "learning_rate": 2.0494489210539693e-05,
      "loss": 0.0145,
      "num_input_tokens_seen": 15017952,
      "step": 22970
    },
    {
      "epoch": 12.04140461215933,
      "grad_norm": 3.135528087615967,
      "learning_rate": 2.0483242619394817e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 15020736,
      "step": 22975
    },
    {
      "epoch": 12.044025157232705,
      "grad_norm": 4.670656204223633,
      "learning_rate": 2.0471996973104063e-05,
      "loss": 0.1593,
      "num_input_tokens_seen": 15024544,
      "step": 22980
    },
    {
      "epoch": 12.04664570230608,
      "grad_norm": 2.065124750137329,
      "learning_rate": 2.0460752274019886e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 15027424,
      "step": 22985
    },
    {
      "epoch": 12.049266247379455,
      "grad_norm": 2.0872960090637207,
      "learning_rate": 2.0449508524494537e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 15030976,
      "step": 22990
    },
    {
      "epoch": 12.05188679245283,
      "grad_norm": 1.3653440475463867,
      "learning_rate": 2.0438265726880092e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 15034464,
      "step": 22995
    },
    {
      "epoch": 12.054507337526205,
      "grad_norm": 1.6966280937194824,
      "learning_rate": 2.042702388352841e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 15039104,
      "step": 23000
    },
    {
      "epoch": 12.05712788259958,
      "grad_norm": 2.660842180252075,
      "learning_rate": 2.041578299679117e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 15041920,
      "step": 23005
    },
    {
      "epoch": 12.059748427672956,
      "grad_norm": 9.518631935119629,
      "learning_rate": 2.0404543069019818e-05,
      "loss": 0.08,
      "num_input_tokens_seen": 15045344,
      "step": 23010
    },
    {
      "epoch": 12.06236897274633,
      "grad_norm": 2.1005733013153076,
      "learning_rate": 2.0393304102565642e-05,
      "loss": 0.0631,
      "num_input_tokens_seen": 15048352,
      "step": 23015
    },
    {
      "epoch": 12.064989517819706,
      "grad_norm": 3.567523717880249,
      "learning_rate": 2.0382066099779685e-05,
      "loss": 0.0283,
      "num_input_tokens_seen": 15051680,
      "step": 23020
    },
    {
      "epoch": 12.067610062893081,
      "grad_norm": 1.895719289779663,
      "learning_rate": 2.0370829063012808e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 15054304,
      "step": 23025
    },
    {
      "epoch": 12.070230607966456,
      "grad_norm": 1.7399152517318726,
      "learning_rate": 2.035959299461568e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 15056640,
      "step": 23030
    },
    {
      "epoch": 12.072851153039831,
      "grad_norm": 1.6258578300476074,
      "learning_rate": 2.0348357896938762e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 15059456,
      "step": 23035
    },
    {
      "epoch": 12.075471698113208,
      "grad_norm": 3.3575730323791504,
      "learning_rate": 2.03371237723323e-05,
      "loss": 0.081,
      "num_input_tokens_seen": 15062720,
      "step": 23040
    },
    {
      "epoch": 12.078092243186584,
      "grad_norm": 1.7808343172073364,
      "learning_rate": 2.032589062314636e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 15065568,
      "step": 23045
    },
    {
      "epoch": 12.080712788259959,
      "grad_norm": 2.689497709274292,
      "learning_rate": 2.0314658451730767e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 15068224,
      "step": 23050
    },
    {
      "epoch": 12.083333333333334,
      "grad_norm": 4.191396236419678,
      "learning_rate": 2.030342726043517e-05,
      "loss": 0.0732,
      "num_input_tokens_seen": 15072032,
      "step": 23055
    },
    {
      "epoch": 12.085953878406709,
      "grad_norm": 1.459831714630127,
      "learning_rate": 2.0292197051609013e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 15075104,
      "step": 23060
    },
    {
      "epoch": 12.088574423480084,
      "grad_norm": 1.386896014213562,
      "learning_rate": 2.0280967827601518e-05,
      "loss": 0.0865,
      "num_input_tokens_seen": 15078336,
      "step": 23065
    },
    {
      "epoch": 12.09119496855346,
      "grad_norm": 0.6954485774040222,
      "learning_rate": 2.0269739590761723e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 15080736,
      "step": 23070
    },
    {
      "epoch": 12.093815513626835,
      "grad_norm": 9.04417896270752,
      "learning_rate": 2.0258512343438442e-05,
      "loss": 0.0771,
      "num_input_tokens_seen": 15083392,
      "step": 23075
    },
    {
      "epoch": 12.09643605870021,
      "grad_norm": 1.031791090965271,
      "learning_rate": 2.0247286087980276e-05,
      "loss": 0.0305,
      "num_input_tokens_seen": 15086784,
      "step": 23080
    },
    {
      "epoch": 12.099056603773585,
      "grad_norm": 1.345314621925354,
      "learning_rate": 2.0236060826735644e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 15089504,
      "step": 23085
    },
    {
      "epoch": 12.10167714884696,
      "grad_norm": 5.830625057220459,
      "learning_rate": 2.0224836562052732e-05,
      "loss": 0.061,
      "num_input_tokens_seen": 15092512,
      "step": 23090
    },
    {
      "epoch": 12.104297693920335,
      "grad_norm": 4.778675079345703,
      "learning_rate": 2.0213613296279533e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 15095808,
      "step": 23095
    },
    {
      "epoch": 12.10691823899371,
      "grad_norm": 1.3081282377243042,
      "learning_rate": 2.0202391031763844e-05,
      "loss": 0.0989,
      "num_input_tokens_seen": 15098720,
      "step": 23100
    },
    {
      "epoch": 12.109538784067086,
      "grad_norm": 2.64676833152771,
      "learning_rate": 2.0191169770853195e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 15102176,
      "step": 23105
    },
    {
      "epoch": 12.11215932914046,
      "grad_norm": 0.5134444832801819,
      "learning_rate": 2.0179949515894975e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 15104480,
      "step": 23110
    },
    {
      "epoch": 12.114779874213836,
      "grad_norm": 5.313261985778809,
      "learning_rate": 2.016873026923633e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 15107264,
      "step": 23115
    },
    {
      "epoch": 12.117400419287211,
      "grad_norm": 0.6992260813713074,
      "learning_rate": 2.0157512033224187e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 15111104,
      "step": 23120
    },
    {
      "epoch": 12.120020964360586,
      "grad_norm": 0.2140982300043106,
      "learning_rate": 2.0146294810205287e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 15114240,
      "step": 23125
    },
    {
      "epoch": 12.122641509433961,
      "grad_norm": 4.213108062744141,
      "learning_rate": 2.0135078602526146e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 15117600,
      "step": 23130
    },
    {
      "epoch": 12.125262054507338,
      "grad_norm": 1.4213775396347046,
      "learning_rate": 2.0123863412533046e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 15120416,
      "step": 23135
    },
    {
      "epoch": 12.127882599580714,
      "grad_norm": 2.115906000137329,
      "learning_rate": 2.0112649242572095e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 15124064,
      "step": 23140
    },
    {
      "epoch": 12.130503144654089,
      "grad_norm": 1.859954595565796,
      "learning_rate": 2.0101436094989164e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 15127168,
      "step": 23145
    },
    {
      "epoch": 12.133123689727464,
      "grad_norm": 2.4078426361083984,
      "learning_rate": 2.0090223972129918e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 15130208,
      "step": 23150
    },
    {
      "epoch": 12.135744234800839,
      "grad_norm": 5.228672027587891,
      "learning_rate": 2.0079012876339797e-05,
      "loss": 0.0992,
      "num_input_tokens_seen": 15133760,
      "step": 23155
    },
    {
      "epoch": 12.138364779874214,
      "grad_norm": 1.6527403593063354,
      "learning_rate": 2.0067802809964055e-05,
      "loss": 0.073,
      "num_input_tokens_seen": 15136992,
      "step": 23160
    },
    {
      "epoch": 12.14098532494759,
      "grad_norm": 3.040193796157837,
      "learning_rate": 2.0056593775347686e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 15139296,
      "step": 23165
    },
    {
      "epoch": 12.143605870020965,
      "grad_norm": 0.7758124470710754,
      "learning_rate": 2.00453857748355e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 15142656,
      "step": 23170
    },
    {
      "epoch": 12.14622641509434,
      "grad_norm": 1.9782687425613403,
      "learning_rate": 2.0034178810772082e-05,
      "loss": 0.0794,
      "num_input_tokens_seen": 15145824,
      "step": 23175
    },
    {
      "epoch": 12.148846960167715,
      "grad_norm": 3.0623514652252197,
      "learning_rate": 2.002297288550181e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 15148608,
      "step": 23180
    },
    {
      "epoch": 12.15146750524109,
      "grad_norm": 0.6486659049987793,
      "learning_rate": 2.0011768001368835e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 15151200,
      "step": 23185
    },
    {
      "epoch": 12.154088050314465,
      "grad_norm": 0.31464022397994995,
      "learning_rate": 2.0000564160717076e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 15154464,
      "step": 23190
    },
    {
      "epoch": 12.15670859538784,
      "grad_norm": 3.1950130462646484,
      "learning_rate": 1.9989361365890263e-05,
      "loss": 0.1288,
      "num_input_tokens_seen": 15158272,
      "step": 23195
    },
    {
      "epoch": 12.159329140461216,
      "grad_norm": 0.29163751006126404,
      "learning_rate": 1.997815961923188e-05,
      "loss": 0.0516,
      "num_input_tokens_seen": 15161152,
      "step": 23200
    },
    {
      "epoch": 12.16194968553459,
      "grad_norm": 1.728599190711975,
      "learning_rate": 1.996695892308522e-05,
      "loss": 0.0283,
      "num_input_tokens_seen": 15163616,
      "step": 23205
    },
    {
      "epoch": 12.164570230607966,
      "grad_norm": 2.866442918777466,
      "learning_rate": 1.9955759279793327e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 15167264,
      "step": 23210
    },
    {
      "epoch": 12.167190775681341,
      "grad_norm": 1.2423884868621826,
      "learning_rate": 1.9944560691699057e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 15170816,
      "step": 23215
    },
    {
      "epoch": 12.169811320754716,
      "grad_norm": 3.219182252883911,
      "learning_rate": 1.9933363161145006e-05,
      "loss": 0.0492,
      "num_input_tokens_seen": 15173280,
      "step": 23220
    },
    {
      "epoch": 12.172431865828091,
      "grad_norm": 2.695385217666626,
      "learning_rate": 1.9922166690473575e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 15176544,
      "step": 23225
    },
    {
      "epoch": 12.175052410901468,
      "grad_norm": 2.7481677532196045,
      "learning_rate": 1.991097128202695e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 15179488,
      "step": 23230
    },
    {
      "epoch": 12.177672955974844,
      "grad_norm": 2.2050139904022217,
      "learning_rate": 1.9899776938147065e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 15182016,
      "step": 23235
    },
    {
      "epoch": 12.180293501048219,
      "grad_norm": 2.256610155105591,
      "learning_rate": 1.9888583661175665e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 15185760,
      "step": 23240
    },
    {
      "epoch": 12.182914046121594,
      "grad_norm": 1.8704664707183838,
      "learning_rate": 1.9877391453454253e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 15188992,
      "step": 23245
    },
    {
      "epoch": 12.185534591194969,
      "grad_norm": 0.9754112958908081,
      "learning_rate": 1.98662003173241e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 15192768,
      "step": 23250
    },
    {
      "epoch": 12.188155136268344,
      "grad_norm": 3.2138607501983643,
      "learning_rate": 1.9855010255126266e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 15196256,
      "step": 23255
    },
    {
      "epoch": 12.19077568134172,
      "grad_norm": 0.34902098774909973,
      "learning_rate": 1.9843821269201598e-05,
      "loss": 0.0722,
      "num_input_tokens_seen": 15202848,
      "step": 23260
    },
    {
      "epoch": 12.193396226415095,
      "grad_norm": 6.25675106048584,
      "learning_rate": 1.983263336189069e-05,
      "loss": 0.0669,
      "num_input_tokens_seen": 15206240,
      "step": 23265
    },
    {
      "epoch": 12.19601677148847,
      "grad_norm": 1.6091947555541992,
      "learning_rate": 1.9821446535533937e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 15208864,
      "step": 23270
    },
    {
      "epoch": 12.198637316561845,
      "grad_norm": 2.442127227783203,
      "learning_rate": 1.9810260792471492e-05,
      "loss": 0.0487,
      "num_input_tokens_seen": 15211968,
      "step": 23275
    },
    {
      "epoch": 12.20125786163522,
      "grad_norm": 2.2786002159118652,
      "learning_rate": 1.979907613504327e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 15214656,
      "step": 23280
    },
    {
      "epoch": 12.203878406708595,
      "grad_norm": 1.913506031036377,
      "learning_rate": 1.978789256558899e-05,
      "loss": 0.0578,
      "num_input_tokens_seen": 15218048,
      "step": 23285
    },
    {
      "epoch": 12.20649895178197,
      "grad_norm": 1.801688551902771,
      "learning_rate": 1.977671008644812e-05,
      "loss": 0.0939,
      "num_input_tokens_seen": 15222112,
      "step": 23290
    },
    {
      "epoch": 12.209119496855346,
      "grad_norm": 1.752083420753479,
      "learning_rate": 1.9765528699959912e-05,
      "loss": 0.0554,
      "num_input_tokens_seen": 15225312,
      "step": 23295
    },
    {
      "epoch": 12.21174004192872,
      "grad_norm": 1.4351614713668823,
      "learning_rate": 1.975434840846339e-05,
      "loss": 0.1046,
      "num_input_tokens_seen": 15228288,
      "step": 23300
    },
    {
      "epoch": 12.214360587002096,
      "grad_norm": 1.408191204071045,
      "learning_rate": 1.9743169214297326e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 15232352,
      "step": 23305
    },
    {
      "epoch": 12.216981132075471,
      "grad_norm": 1.606174111366272,
      "learning_rate": 1.973199111980029e-05,
      "loss": 0.0593,
      "num_input_tokens_seen": 15237984,
      "step": 23310
    },
    {
      "epoch": 12.219601677148846,
      "grad_norm": 12.823089599609375,
      "learning_rate": 1.9720814127310608e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 15241184,
      "step": 23315
    },
    {
      "epoch": 12.222222222222221,
      "grad_norm": 2.1539430618286133,
      "learning_rate": 1.9709638239166386e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 15243744,
      "step": 23320
    },
    {
      "epoch": 12.224842767295598,
      "grad_norm": 1.4003762006759644,
      "learning_rate": 1.9698463457705485e-05,
      "loss": 0.1707,
      "num_input_tokens_seen": 15246432,
      "step": 23325
    },
    {
      "epoch": 12.227463312368974,
      "grad_norm": 2.9386634826660156,
      "learning_rate": 1.968728978526556e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 15249440,
      "step": 23330
    },
    {
      "epoch": 12.230083857442349,
      "grad_norm": 2.1660845279693604,
      "learning_rate": 1.9676117224183986e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 15253120,
      "step": 23335
    },
    {
      "epoch": 12.232704402515724,
      "grad_norm": 4.024454593658447,
      "learning_rate": 1.9664945776797956e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 15256224,
      "step": 23340
    },
    {
      "epoch": 12.235324947589099,
      "grad_norm": 1.5588881969451904,
      "learning_rate": 1.96537754454444e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 15258720,
      "step": 23345
    },
    {
      "epoch": 12.237945492662474,
      "grad_norm": 1.9781404733657837,
      "learning_rate": 1.964260623246003e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 15261760,
      "step": 23350
    },
    {
      "epoch": 12.24056603773585,
      "grad_norm": 1.3122398853302002,
      "learning_rate": 1.9631438140181324e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 15265472,
      "step": 23355
    },
    {
      "epoch": 12.243186582809225,
      "grad_norm": 1.377478003501892,
      "learning_rate": 1.9620271170944503e-05,
      "loss": 0.0347,
      "num_input_tokens_seen": 15268896,
      "step": 23360
    },
    {
      "epoch": 12.2458071278826,
      "grad_norm": 1.1498329639434814,
      "learning_rate": 1.960910532708558e-05,
      "loss": 0.0167,
      "num_input_tokens_seen": 15272320,
      "step": 23365
    },
    {
      "epoch": 12.248427672955975,
      "grad_norm": 3.837977409362793,
      "learning_rate": 1.9597940610940315e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 15275776,
      "step": 23370
    },
    {
      "epoch": 12.25104821802935,
      "grad_norm": 10.033125877380371,
      "learning_rate": 1.958677702484425e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 15278912,
      "step": 23375
    },
    {
      "epoch": 12.253668763102725,
      "grad_norm": 2.5182111263275146,
      "learning_rate": 1.957561457113267e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 15281536,
      "step": 23380
    },
    {
      "epoch": 12.2562893081761,
      "grad_norm": 4.008944988250732,
      "learning_rate": 1.9564453252140648e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 15284544,
      "step": 23385
    },
    {
      "epoch": 12.258909853249476,
      "grad_norm": 1.6526029109954834,
      "learning_rate": 1.9553293070202984e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 15287968,
      "step": 23390
    },
    {
      "epoch": 12.26153039832285,
      "grad_norm": 1.7246037721633911,
      "learning_rate": 1.9542134027654267e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 15291136,
      "step": 23395
    },
    {
      "epoch": 12.264150943396226,
      "grad_norm": 1.558302640914917,
      "learning_rate": 1.953097612682885e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 15293824,
      "step": 23400
    },
    {
      "epoch": 12.266771488469601,
      "grad_norm": 0.8943349719047546,
      "learning_rate": 1.9519819370060832e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 15296928,
      "step": 23405
    },
    {
      "epoch": 12.269392033542976,
      "grad_norm": 1.4507904052734375,
      "learning_rate": 1.9508663759684074e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 15299904,
      "step": 23410
    },
    {
      "epoch": 12.272012578616351,
      "grad_norm": 1.218790054321289,
      "learning_rate": 1.9497509298032224e-05,
      "loss": 0.0435,
      "num_input_tokens_seen": 15302624,
      "step": 23415
    },
    {
      "epoch": 12.274633123689728,
      "grad_norm": 1.801243782043457,
      "learning_rate": 1.9486355987438647e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 15305408,
      "step": 23420
    },
    {
      "epoch": 12.277253668763104,
      "grad_norm": 3.6758532524108887,
      "learning_rate": 1.947520383023649e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 15308288,
      "step": 23425
    },
    {
      "epoch": 12.279874213836479,
      "grad_norm": 0.9104042649269104,
      "learning_rate": 1.9464052828758666e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 15311200,
      "step": 23430
    },
    {
      "epoch": 12.282494758909854,
      "grad_norm": 3.8054442405700684,
      "learning_rate": 1.945290298533783e-05,
      "loss": 0.0694,
      "num_input_tokens_seen": 15314688,
      "step": 23435
    },
    {
      "epoch": 12.285115303983229,
      "grad_norm": 1.998371958732605,
      "learning_rate": 1.9441754302306406e-05,
      "loss": 0.0168,
      "num_input_tokens_seen": 15317632,
      "step": 23440
    },
    {
      "epoch": 12.287735849056604,
      "grad_norm": 1.2652747631072998,
      "learning_rate": 1.9430606781996578e-05,
      "loss": 0.0211,
      "num_input_tokens_seen": 15320288,
      "step": 23445
    },
    {
      "epoch": 12.29035639412998,
      "grad_norm": 1.3496267795562744,
      "learning_rate": 1.941946042674026e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 15323520,
      "step": 23450
    },
    {
      "epoch": 12.292976939203355,
      "grad_norm": 3.3820509910583496,
      "learning_rate": 1.940831523886916e-05,
      "loss": 0.0686,
      "num_input_tokens_seen": 15327008,
      "step": 23455
    },
    {
      "epoch": 12.29559748427673,
      "grad_norm": 1.8231463432312012,
      "learning_rate": 1.9397171220714714e-05,
      "loss": 0.0376,
      "num_input_tokens_seen": 15329952,
      "step": 23460
    },
    {
      "epoch": 12.298218029350105,
      "grad_norm": 2.7879397869110107,
      "learning_rate": 1.9386028374608134e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 15333056,
      "step": 23465
    },
    {
      "epoch": 12.30083857442348,
      "grad_norm": 2.3258354663848877,
      "learning_rate": 1.9374886702880372e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 15336736,
      "step": 23470
    },
    {
      "epoch": 12.303459119496855,
      "grad_norm": 2.003448724746704,
      "learning_rate": 1.936374620786213e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 15339712,
      "step": 23475
    },
    {
      "epoch": 12.30607966457023,
      "grad_norm": 2.5601601600646973,
      "learning_rate": 1.9352606891883885e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 15343232,
      "step": 23480
    },
    {
      "epoch": 12.308700209643606,
      "grad_norm": 1.7095154523849487,
      "learning_rate": 1.9341468757275844e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 15346816,
      "step": 23485
    },
    {
      "epoch": 12.31132075471698,
      "grad_norm": 2.3395538330078125,
      "learning_rate": 1.933033180636798e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 15349760,
      "step": 23490
    },
    {
      "epoch": 12.313941299790356,
      "grad_norm": 1.3373609781265259,
      "learning_rate": 1.9319196041490027e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 15353184,
      "step": 23495
    },
    {
      "epoch": 12.316561844863731,
      "grad_norm": 7.577740669250488,
      "learning_rate": 1.930806146497146e-05,
      "loss": 0.053,
      "num_input_tokens_seen": 15356512,
      "step": 23500
    },
    {
      "epoch": 12.319182389937106,
      "grad_norm": 0.8708258867263794,
      "learning_rate": 1.9296928079141485e-05,
      "loss": 0.0293,
      "num_input_tokens_seen": 15359104,
      "step": 23505
    },
    {
      "epoch": 12.321802935010481,
      "grad_norm": 4.108316421508789,
      "learning_rate": 1.9285795886329093e-05,
      "loss": 0.0675,
      "num_input_tokens_seen": 15362528,
      "step": 23510
    },
    {
      "epoch": 12.324423480083858,
      "grad_norm": 2.80364990234375,
      "learning_rate": 1.927466488886301e-05,
      "loss": 0.0695,
      "num_input_tokens_seen": 15366240,
      "step": 23515
    },
    {
      "epoch": 12.327044025157234,
      "grad_norm": 2.8364510536193848,
      "learning_rate": 1.926353508907172e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 15368704,
      "step": 23520
    },
    {
      "epoch": 12.329664570230609,
      "grad_norm": 1.9460965394973755,
      "learning_rate": 1.9252406489283437e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 15373312,
      "step": 23525
    },
    {
      "epoch": 12.332285115303984,
      "grad_norm": 1.124068260192871,
      "learning_rate": 1.9241279091826155e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 15377632,
      "step": 23530
    },
    {
      "epoch": 12.334905660377359,
      "grad_norm": 3.3258168697357178,
      "learning_rate": 1.923015289902758e-05,
      "loss": 0.0493,
      "num_input_tokens_seen": 15380864,
      "step": 23535
    },
    {
      "epoch": 12.337526205450734,
      "grad_norm": 1.8493711948394775,
      "learning_rate": 1.9219027913215193e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 15384480,
      "step": 23540
    },
    {
      "epoch": 12.34014675052411,
      "grad_norm": 2.219879388809204,
      "learning_rate": 1.9207904136716214e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 15387968,
      "step": 23545
    },
    {
      "epoch": 12.342767295597485,
      "grad_norm": 5.163074016571045,
      "learning_rate": 1.9196781571857608e-05,
      "loss": 0.0494,
      "num_input_tokens_seen": 15390528,
      "step": 23550
    },
    {
      "epoch": 12.34538784067086,
      "grad_norm": 3.39473819732666,
      "learning_rate": 1.9185660220966097e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 15394368,
      "step": 23555
    },
    {
      "epoch": 12.348008385744235,
      "grad_norm": 2.680886745452881,
      "learning_rate": 1.9174540086368135e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 15397600,
      "step": 23560
    },
    {
      "epoch": 12.35062893081761,
      "grad_norm": 1.4155724048614502,
      "learning_rate": 1.9163421170389918e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 15400288,
      "step": 23565
    },
    {
      "epoch": 12.353249475890985,
      "grad_norm": 5.050276279449463,
      "learning_rate": 1.9152303475357402e-05,
      "loss": 0.0749,
      "num_input_tokens_seen": 15404032,
      "step": 23570
    },
    {
      "epoch": 12.35587002096436,
      "grad_norm": 0.3928076922893524,
      "learning_rate": 1.9141187003596295e-05,
      "loss": 0.1301,
      "num_input_tokens_seen": 15407392,
      "step": 23575
    },
    {
      "epoch": 12.358490566037736,
      "grad_norm": 5.346042633056641,
      "learning_rate": 1.913007175743202e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 15410208,
      "step": 23580
    },
    {
      "epoch": 12.36111111111111,
      "grad_norm": 3.09086275100708,
      "learning_rate": 1.9118957739189773e-05,
      "loss": 0.0641,
      "num_input_tokens_seen": 15412992,
      "step": 23585
    },
    {
      "epoch": 12.363731656184486,
      "grad_norm": 2.070213556289673,
      "learning_rate": 1.910784495119447e-05,
      "loss": 0.0913,
      "num_input_tokens_seen": 15417056,
      "step": 23590
    },
    {
      "epoch": 12.366352201257861,
      "grad_norm": 1.4545689821243286,
      "learning_rate": 1.9096733395770778e-05,
      "loss": 0.0685,
      "num_input_tokens_seen": 15420032,
      "step": 23595
    },
    {
      "epoch": 12.368972746331236,
      "grad_norm": 1.1682157516479492,
      "learning_rate": 1.9085623075243114e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 15423776,
      "step": 23600
    },
    {
      "epoch": 12.371593291404611,
      "grad_norm": 2.4852654933929443,
      "learning_rate": 1.9074513991935623e-05,
      "loss": 0.0773,
      "num_input_tokens_seen": 15426880,
      "step": 23605
    },
    {
      "epoch": 12.374213836477987,
      "grad_norm": 6.179083347320557,
      "learning_rate": 1.906340614817221e-05,
      "loss": 0.0821,
      "num_input_tokens_seen": 15430688,
      "step": 23610
    },
    {
      "epoch": 12.376834381551364,
      "grad_norm": 4.175281047821045,
      "learning_rate": 1.9052299546276506e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 15433728,
      "step": 23615
    },
    {
      "epoch": 12.379454926624739,
      "grad_norm": 2.747422933578491,
      "learning_rate": 1.904119418857187e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 15437568,
      "step": 23620
    },
    {
      "epoch": 12.382075471698114,
      "grad_norm": 3.1167752742767334,
      "learning_rate": 1.903009007738144e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 15440384,
      "step": 23625
    },
    {
      "epoch": 12.384696016771489,
      "grad_norm": 1.6938122510910034,
      "learning_rate": 1.9018987215028045e-05,
      "loss": 0.0457,
      "num_input_tokens_seen": 15444032,
      "step": 23630
    },
    {
      "epoch": 12.387316561844864,
      "grad_norm": 2.1936252117156982,
      "learning_rate": 1.9007885603834298e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 15446400,
      "step": 23635
    },
    {
      "epoch": 12.38993710691824,
      "grad_norm": 2.185835123062134,
      "learning_rate": 1.8996785246122526e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 15450880,
      "step": 23640
    },
    {
      "epoch": 12.392557651991615,
      "grad_norm": 1.7064733505249023,
      "learning_rate": 1.898568614421478e-05,
      "loss": 0.0629,
      "num_input_tokens_seen": 15453856,
      "step": 23645
    },
    {
      "epoch": 12.39517819706499,
      "grad_norm": 0.969797670841217,
      "learning_rate": 1.897458830043288e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 15456320,
      "step": 23650
    },
    {
      "epoch": 12.397798742138365,
      "grad_norm": 2.450120210647583,
      "learning_rate": 1.8963491717098372e-05,
      "loss": 0.0745,
      "num_input_tokens_seen": 15459616,
      "step": 23655
    },
    {
      "epoch": 12.40041928721174,
      "grad_norm": 1.9591177701950073,
      "learning_rate": 1.8952396396532522e-05,
      "loss": 0.084,
      "num_input_tokens_seen": 15463520,
      "step": 23660
    },
    {
      "epoch": 12.403039832285115,
      "grad_norm": 1.8106091022491455,
      "learning_rate": 1.8941302341056356e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 15466976,
      "step": 23665
    },
    {
      "epoch": 12.40566037735849,
      "grad_norm": 3.855966091156006,
      "learning_rate": 1.8930209552990625e-05,
      "loss": 0.0683,
      "num_input_tokens_seen": 15470784,
      "step": 23670
    },
    {
      "epoch": 12.408280922431866,
      "grad_norm": 1.3894799947738647,
      "learning_rate": 1.8919118034655796e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 15474816,
      "step": 23675
    },
    {
      "epoch": 12.41090146750524,
      "grad_norm": 3.226384401321411,
      "learning_rate": 1.8908027788372107e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 15478016,
      "step": 23680
    },
    {
      "epoch": 12.413522012578616,
      "grad_norm": 4.548415660858154,
      "learning_rate": 1.88969388164595e-05,
      "loss": 0.0416,
      "num_input_tokens_seen": 15481536,
      "step": 23685
    },
    {
      "epoch": 12.416142557651991,
      "grad_norm": 1.6737029552459717,
      "learning_rate": 1.8885851121237666e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 15485504,
      "step": 23690
    },
    {
      "epoch": 12.418763102725366,
      "grad_norm": 2.583131790161133,
      "learning_rate": 1.8874764705026023e-05,
      "loss": 0.0976,
      "num_input_tokens_seen": 15489120,
      "step": 23695
    },
    {
      "epoch": 12.421383647798741,
      "grad_norm": 4.344069004058838,
      "learning_rate": 1.8863679570143728e-05,
      "loss": 0.0729,
      "num_input_tokens_seen": 15491840,
      "step": 23700
    },
    {
      "epoch": 12.424004192872117,
      "grad_norm": 14.870306015014648,
      "learning_rate": 1.885259571890966e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 15494880,
      "step": 23705
    },
    {
      "epoch": 12.426624737945493,
      "grad_norm": 0.9564418792724609,
      "learning_rate": 1.8841513153642427e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 15497312,
      "step": 23710
    },
    {
      "epoch": 12.429245283018869,
      "grad_norm": 1.2168911695480347,
      "learning_rate": 1.8830431876660388e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 15500416,
      "step": 23715
    },
    {
      "epoch": 12.431865828092244,
      "grad_norm": 15.431204795837402,
      "learning_rate": 1.8819351890281607e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 15503520,
      "step": 23720
    },
    {
      "epoch": 12.434486373165619,
      "grad_norm": 5.3895697593688965,
      "learning_rate": 1.880827319682391e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 15506656,
      "step": 23725
    },
    {
      "epoch": 12.437106918238994,
      "grad_norm": 3.1555216312408447,
      "learning_rate": 1.879719579860481e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 15509792,
      "step": 23730
    },
    {
      "epoch": 12.43972746331237,
      "grad_norm": 3.9432291984558105,
      "learning_rate": 1.878611969794159e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 15512768,
      "step": 23735
    },
    {
      "epoch": 12.442348008385745,
      "grad_norm": 1.8653337955474854,
      "learning_rate": 1.8775044897151232e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 15515712,
      "step": 23740
    },
    {
      "epoch": 12.44496855345912,
      "grad_norm": 1.496741771697998,
      "learning_rate": 1.876397139855047e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 15518528,
      "step": 23745
    },
    {
      "epoch": 12.447589098532495,
      "grad_norm": 1.5155608654022217,
      "learning_rate": 1.875289920445574e-05,
      "loss": 0.0526,
      "num_input_tokens_seen": 15521472,
      "step": 23750
    },
    {
      "epoch": 12.45020964360587,
      "grad_norm": 1.3552234172821045,
      "learning_rate": 1.8741828317183234e-05,
      "loss": 0.0166,
      "num_input_tokens_seen": 15524320,
      "step": 23755
    },
    {
      "epoch": 12.452830188679245,
      "grad_norm": 2.7194364070892334,
      "learning_rate": 1.873075873904885e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 15527168,
      "step": 23760
    },
    {
      "epoch": 12.45545073375262,
      "grad_norm": 1.9301860332489014,
      "learning_rate": 1.87196904723682e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 15532064,
      "step": 23765
    },
    {
      "epoch": 12.458071278825996,
      "grad_norm": 3.3861801624298096,
      "learning_rate": 1.8708623519456665e-05,
      "loss": 0.0251,
      "num_input_tokens_seen": 15534944,
      "step": 23770
    },
    {
      "epoch": 12.46069182389937,
      "grad_norm": 1.1083240509033203,
      "learning_rate": 1.8697557882629308e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 15538368,
      "step": 23775
    },
    {
      "epoch": 12.463312368972746,
      "grad_norm": 2.065549850463867,
      "learning_rate": 1.868649356420094e-05,
      "loss": 0.0803,
      "num_input_tokens_seen": 15541280,
      "step": 23780
    },
    {
      "epoch": 12.465932914046121,
      "grad_norm": 3.869366407394409,
      "learning_rate": 1.86754305664861e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 15544736,
      "step": 23785
    },
    {
      "epoch": 12.468553459119496,
      "grad_norm": 2.344041347503662,
      "learning_rate": 1.866436889179902e-05,
      "loss": 0.0657,
      "num_input_tokens_seen": 15549568,
      "step": 23790
    },
    {
      "epoch": 12.471174004192871,
      "grad_norm": 1.8903025388717651,
      "learning_rate": 1.865330854245369e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 15553184,
      "step": 23795
    },
    {
      "epoch": 12.473794549266247,
      "grad_norm": 1.4740569591522217,
      "learning_rate": 1.8642249520763804e-05,
      "loss": 0.0408,
      "num_input_tokens_seen": 15557120,
      "step": 23800
    },
    {
      "epoch": 12.476415094339623,
      "grad_norm": 2.6325056552886963,
      "learning_rate": 1.863119182904278e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 15561632,
      "step": 23805
    },
    {
      "epoch": 12.479035639412999,
      "grad_norm": 3.072584390640259,
      "learning_rate": 1.8620135469603777e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 15568736,
      "step": 23810
    },
    {
      "epoch": 12.481656184486374,
      "grad_norm": 1.6831729412078857,
      "learning_rate": 1.8609080444759653e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 15572352,
      "step": 23815
    },
    {
      "epoch": 12.484276729559749,
      "grad_norm": 2.340610980987549,
      "learning_rate": 1.8598026756822978e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 15574688,
      "step": 23820
    },
    {
      "epoch": 12.486897274633124,
      "grad_norm": 1.4035078287124634,
      "learning_rate": 1.8586974408106074e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 15577376,
      "step": 23825
    },
    {
      "epoch": 12.4895178197065,
      "grad_norm": 7.653781890869141,
      "learning_rate": 1.8575923400920954e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 15581280,
      "step": 23830
    },
    {
      "epoch": 12.492138364779874,
      "grad_norm": 2.0808773040771484,
      "learning_rate": 1.856487373757937e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 15584544,
      "step": 23835
    },
    {
      "epoch": 12.49475890985325,
      "grad_norm": 1.724210262298584,
      "learning_rate": 1.8553825420392792e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 15587520,
      "step": 23840
    },
    {
      "epoch": 12.497379454926625,
      "grad_norm": 3.5799734592437744,
      "learning_rate": 1.8542778451672385e-05,
      "loss": 0.0987,
      "num_input_tokens_seen": 15590624,
      "step": 23845
    },
    {
      "epoch": 12.5,
      "grad_norm": 2.1416220664978027,
      "learning_rate": 1.8531732833729063e-05,
      "loss": 0.0497,
      "num_input_tokens_seen": 15594560,
      "step": 23850
    },
    {
      "epoch": 12.502620545073375,
      "grad_norm": 4.257220268249512,
      "learning_rate": 1.852068856887343e-05,
      "loss": 0.095,
      "num_input_tokens_seen": 15598048,
      "step": 23855
    },
    {
      "epoch": 12.50524109014675,
      "grad_norm": 2.742546558380127,
      "learning_rate": 1.8509645659415837e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 15602048,
      "step": 23860
    },
    {
      "epoch": 12.507861635220126,
      "grad_norm": 2.09000825881958,
      "learning_rate": 1.849860410766632e-05,
      "loss": 0.0798,
      "num_input_tokens_seen": 15605152,
      "step": 23865
    },
    {
      "epoch": 12.5104821802935,
      "grad_norm": 2.0900135040283203,
      "learning_rate": 1.8487563915934664e-05,
      "loss": 0.0506,
      "num_input_tokens_seen": 15607808,
      "step": 23870
    },
    {
      "epoch": 12.513102725366876,
      "grad_norm": 0.31436705589294434,
      "learning_rate": 1.847652508653033e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 15613344,
      "step": 23875
    },
    {
      "epoch": 12.515723270440251,
      "grad_norm": 1.693334698677063,
      "learning_rate": 1.846548762176252e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 15616128,
      "step": 23880
    },
    {
      "epoch": 12.518343815513626,
      "grad_norm": 1.684117078781128,
      "learning_rate": 1.845445152394015e-05,
      "loss": 0.0961,
      "num_input_tokens_seen": 15619328,
      "step": 23885
    },
    {
      "epoch": 12.520964360587001,
      "grad_norm": 0.7763387560844421,
      "learning_rate": 1.8443416795371853e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 15622624,
      "step": 23890
    },
    {
      "epoch": 12.523584905660378,
      "grad_norm": 1.6381332874298096,
      "learning_rate": 1.8432383438365957e-05,
      "loss": 0.0365,
      "num_input_tokens_seen": 15626560,
      "step": 23895
    },
    {
      "epoch": 12.526205450733752,
      "grad_norm": 2.3678109645843506,
      "learning_rate": 1.8421351455230528e-05,
      "loss": 0.1021,
      "num_input_tokens_seen": 15629888,
      "step": 23900
    },
    {
      "epoch": 12.528825995807129,
      "grad_norm": 2.232717514038086,
      "learning_rate": 1.8410320848273315e-05,
      "loss": 0.0626,
      "num_input_tokens_seen": 15633312,
      "step": 23905
    },
    {
      "epoch": 12.531446540880504,
      "grad_norm": 3.43957257270813,
      "learning_rate": 1.8399291619801794e-05,
      "loss": 0.0663,
      "num_input_tokens_seen": 15636096,
      "step": 23910
    },
    {
      "epoch": 12.534067085953879,
      "grad_norm": 0.4192799925804138,
      "learning_rate": 1.838826377212317e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 15638496,
      "step": 23915
    },
    {
      "epoch": 12.536687631027254,
      "grad_norm": 0.8866831660270691,
      "learning_rate": 1.8377237307544326e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 15642848,
      "step": 23920
    },
    {
      "epoch": 12.53930817610063,
      "grad_norm": 1.686020016670227,
      "learning_rate": 1.8366212228371894e-05,
      "loss": 0.0869,
      "num_input_tokens_seen": 15646080,
      "step": 23925
    },
    {
      "epoch": 12.541928721174004,
      "grad_norm": 1.9442030191421509,
      "learning_rate": 1.8355188536912173e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 15648928,
      "step": 23930
    },
    {
      "epoch": 12.54454926624738,
      "grad_norm": 2.0261831283569336,
      "learning_rate": 1.83441662354712e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 15652608,
      "step": 23935
    },
    {
      "epoch": 12.547169811320755,
      "grad_norm": 4.100189208984375,
      "learning_rate": 1.8333145326354722e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 15656576,
      "step": 23940
    },
    {
      "epoch": 12.54979035639413,
      "grad_norm": 1.466820240020752,
      "learning_rate": 1.8322125811868175e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 15659264,
      "step": 23945
    },
    {
      "epoch": 12.552410901467505,
      "grad_norm": 1.8726445436477661,
      "learning_rate": 1.831110769431673e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 15662272,
      "step": 23950
    },
    {
      "epoch": 12.55503144654088,
      "grad_norm": 1.0575015544891357,
      "learning_rate": 1.8300090976005247e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 15665184,
      "step": 23955
    },
    {
      "epoch": 12.557651991614255,
      "grad_norm": 3.7841153144836426,
      "learning_rate": 1.8289075659238285e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 15668832,
      "step": 23960
    },
    {
      "epoch": 12.56027253668763,
      "grad_norm": 2.5052900314331055,
      "learning_rate": 1.8278061746320134e-05,
      "loss": 0.1336,
      "num_input_tokens_seen": 15672160,
      "step": 23965
    },
    {
      "epoch": 12.562893081761006,
      "grad_norm": 0.630037784576416,
      "learning_rate": 1.8267049239554786e-05,
      "loss": 0.0562,
      "num_input_tokens_seen": 15675296,
      "step": 23970
    },
    {
      "epoch": 12.565513626834381,
      "grad_norm": 3.4198155403137207,
      "learning_rate": 1.8256038141245912e-05,
      "loss": 0.0369,
      "num_input_tokens_seen": 15678784,
      "step": 23975
    },
    {
      "epoch": 12.568134171907756,
      "grad_norm": 2.2122690677642822,
      "learning_rate": 1.824502845369693e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 15680992,
      "step": 23980
    },
    {
      "epoch": 12.570754716981131,
      "grad_norm": 1.2480781078338623,
      "learning_rate": 1.8234020179210938e-05,
      "loss": 0.0349,
      "num_input_tokens_seen": 15683872,
      "step": 23985
    },
    {
      "epoch": 12.573375262054507,
      "grad_norm": 5.298614978790283,
      "learning_rate": 1.8223013320090715e-05,
      "loss": 0.0972,
      "num_input_tokens_seen": 15686432,
      "step": 23990
    },
    {
      "epoch": 12.575995807127882,
      "grad_norm": 1.030115008354187,
      "learning_rate": 1.82120078786388e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 15691296,
      "step": 23995
    },
    {
      "epoch": 12.578616352201259,
      "grad_norm": 2.49025297164917,
      "learning_rate": 1.820100385715739e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 15695712,
      "step": 24000
    },
    {
      "epoch": 12.581236897274634,
      "grad_norm": 1.3265578746795654,
      "learning_rate": 1.8190001257948414e-05,
      "loss": 0.0274,
      "num_input_tokens_seen": 15699264,
      "step": 24005
    },
    {
      "epoch": 12.583857442348009,
      "grad_norm": 4.568558692932129,
      "learning_rate": 1.8179000083313483e-05,
      "loss": 0.0652,
      "num_input_tokens_seen": 15701952,
      "step": 24010
    },
    {
      "epoch": 12.586477987421384,
      "grad_norm": 1.4795429706573486,
      "learning_rate": 1.8168000335553907e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 15704768,
      "step": 24015
    },
    {
      "epoch": 12.58909853249476,
      "grad_norm": 3.0806918144226074,
      "learning_rate": 1.8157002016970723e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 15708352,
      "step": 24020
    },
    {
      "epoch": 12.591719077568134,
      "grad_norm": 0.9881702661514282,
      "learning_rate": 1.8146005129864644e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 15711424,
      "step": 24025
    },
    {
      "epoch": 12.59433962264151,
      "grad_norm": 10.53112506866455,
      "learning_rate": 1.81350096765361e-05,
      "loss": 0.0557,
      "num_input_tokens_seen": 15714848,
      "step": 24030
    },
    {
      "epoch": 12.596960167714885,
      "grad_norm": 2.6270620822906494,
      "learning_rate": 1.812401565928521e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 15717664,
      "step": 24035
    },
    {
      "epoch": 12.59958071278826,
      "grad_norm": 5.073805809020996,
      "learning_rate": 1.8113023080411806e-05,
      "loss": 0.0545,
      "num_input_tokens_seen": 15720448,
      "step": 24040
    },
    {
      "epoch": 12.602201257861635,
      "grad_norm": 2.1584575176239014,
      "learning_rate": 1.810203194221539e-05,
      "loss": 0.0449,
      "num_input_tokens_seen": 15723552,
      "step": 24045
    },
    {
      "epoch": 12.60482180293501,
      "grad_norm": 5.961183071136475,
      "learning_rate": 1.8091042246995206e-05,
      "loss": 0.0182,
      "num_input_tokens_seen": 15727616,
      "step": 24050
    },
    {
      "epoch": 12.607442348008385,
      "grad_norm": 1.102170705795288,
      "learning_rate": 1.8080053997050154e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 15730240,
      "step": 24055
    },
    {
      "epoch": 12.61006289308176,
      "grad_norm": 2.178388833999634,
      "learning_rate": 1.8069067194678867e-05,
      "loss": 0.0588,
      "num_input_tokens_seen": 15734048,
      "step": 24060
    },
    {
      "epoch": 12.612683438155136,
      "grad_norm": 3.847747564315796,
      "learning_rate": 1.8058081842179642e-05,
      "loss": 0.0512,
      "num_input_tokens_seen": 15736896,
      "step": 24065
    },
    {
      "epoch": 12.615303983228511,
      "grad_norm": 6.428053379058838,
      "learning_rate": 1.804709794185051e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 15739872,
      "step": 24070
    },
    {
      "epoch": 12.617924528301886,
      "grad_norm": 0.8242686986923218,
      "learning_rate": 1.8036115495989164e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 15743200,
      "step": 24075
    },
    {
      "epoch": 12.620545073375261,
      "grad_norm": 2.5527334213256836,
      "learning_rate": 1.8025134506892997e-05,
      "loss": 0.0613,
      "num_input_tokens_seen": 15748224,
      "step": 24080
    },
    {
      "epoch": 12.623165618448636,
      "grad_norm": 3.597651243209839,
      "learning_rate": 1.8014154976859126e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 15751168,
      "step": 24085
    },
    {
      "epoch": 12.625786163522012,
      "grad_norm": 1.014121174812317,
      "learning_rate": 1.800317690818433e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 15754592,
      "step": 24090
    },
    {
      "epoch": 12.628406708595389,
      "grad_norm": 0.787413477897644,
      "learning_rate": 1.799220030316511e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 15758368,
      "step": 24095
    },
    {
      "epoch": 12.631027253668764,
      "grad_norm": 8.34557819366455,
      "learning_rate": 1.7981225164097633e-05,
      "loss": 0.0424,
      "num_input_tokens_seen": 15761504,
      "step": 24100
    },
    {
      "epoch": 12.633647798742139,
      "grad_norm": 2.4827070236206055,
      "learning_rate": 1.7970251493277774e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 15764224,
      "step": 24105
    },
    {
      "epoch": 12.636268343815514,
      "grad_norm": 3.4923720359802246,
      "learning_rate": 1.7959279293001102e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 15767072,
      "step": 24110
    },
    {
      "epoch": 12.63888888888889,
      "grad_norm": 6.064342975616455,
      "learning_rate": 1.7948308565562876e-05,
      "loss": 0.1237,
      "num_input_tokens_seen": 15769920,
      "step": 24115
    },
    {
      "epoch": 12.641509433962264,
      "grad_norm": 0.8949292302131653,
      "learning_rate": 1.793733931325805e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 15773472,
      "step": 24120
    },
    {
      "epoch": 12.64412997903564,
      "grad_norm": 3.299515724182129,
      "learning_rate": 1.792637153838127e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 15776064,
      "step": 24125
    },
    {
      "epoch": 12.646750524109015,
      "grad_norm": 0.9872038960456848,
      "learning_rate": 1.7915405243226864e-05,
      "loss": 0.0452,
      "num_input_tokens_seen": 15779552,
      "step": 24130
    },
    {
      "epoch": 12.64937106918239,
      "grad_norm": 0.5801597237586975,
      "learning_rate": 1.7904440430088847e-05,
      "loss": 0.0287,
      "num_input_tokens_seen": 15782272,
      "step": 24135
    },
    {
      "epoch": 12.651991614255765,
      "grad_norm": 3.20627498626709,
      "learning_rate": 1.7893477101260947e-05,
      "loss": 0.0733,
      "num_input_tokens_seen": 15785024,
      "step": 24140
    },
    {
      "epoch": 12.65461215932914,
      "grad_norm": 1.2983900308609009,
      "learning_rate": 1.7882515259036558e-05,
      "loss": 0.0773,
      "num_input_tokens_seen": 15788128,
      "step": 24145
    },
    {
      "epoch": 12.657232704402515,
      "grad_norm": 2.489621639251709,
      "learning_rate": 1.787155490570878e-05,
      "loss": 0.0489,
      "num_input_tokens_seen": 15790944,
      "step": 24150
    },
    {
      "epoch": 12.65985324947589,
      "grad_norm": 4.435710430145264,
      "learning_rate": 1.7860596043570398e-05,
      "loss": 0.0586,
      "num_input_tokens_seen": 15793920,
      "step": 24155
    },
    {
      "epoch": 12.662473794549266,
      "grad_norm": 1.4003323316574097,
      "learning_rate": 1.7849638674913863e-05,
      "loss": 0.0472,
      "num_input_tokens_seen": 15799552,
      "step": 24160
    },
    {
      "epoch": 12.665094339622641,
      "grad_norm": 1.1714273691177368,
      "learning_rate": 1.7838682802031344e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 15802656,
      "step": 24165
    },
    {
      "epoch": 12.667714884696016,
      "grad_norm": 1.2838811874389648,
      "learning_rate": 1.782772842721468e-05,
      "loss": 0.0501,
      "num_input_tokens_seen": 15806112,
      "step": 24170
    },
    {
      "epoch": 12.670335429769391,
      "grad_norm": 4.3638014793396,
      "learning_rate": 1.7816775552755406e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 15808928,
      "step": 24175
    },
    {
      "epoch": 12.672955974842766,
      "grad_norm": 1.6540234088897705,
      "learning_rate": 1.780582418094473e-05,
      "loss": 0.0752,
      "num_input_tokens_seen": 15812320,
      "step": 24180
    },
    {
      "epoch": 12.675576519916142,
      "grad_norm": 1.5942199230194092,
      "learning_rate": 1.7794874314073573e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 15816608,
      "step": 24185
    },
    {
      "epoch": 12.678197064989519,
      "grad_norm": 11.868943214416504,
      "learning_rate": 1.7783925954432503e-05,
      "loss": 0.0659,
      "num_input_tokens_seen": 15819680,
      "step": 24190
    },
    {
      "epoch": 12.680817610062894,
      "grad_norm": 3.713576078414917,
      "learning_rate": 1.777297910431179e-05,
      "loss": 0.0528,
      "num_input_tokens_seen": 15822848,
      "step": 24195
    },
    {
      "epoch": 12.683438155136269,
      "grad_norm": 1.798878788948059,
      "learning_rate": 1.7762033766001402e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 15825312,
      "step": 24200
    },
    {
      "epoch": 12.686058700209644,
      "grad_norm": 1.9073055982589722,
      "learning_rate": 1.7751089941790977e-05,
      "loss": 0.043,
      "num_input_tokens_seen": 15828672,
      "step": 24205
    },
    {
      "epoch": 12.68867924528302,
      "grad_norm": 3.8952701091766357,
      "learning_rate": 1.7740147633969844e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 15831648,
      "step": 24210
    },
    {
      "epoch": 12.691299790356394,
      "grad_norm": 7.113290309906006,
      "learning_rate": 1.772920684482699e-05,
      "loss": 0.0455,
      "num_input_tokens_seen": 15835456,
      "step": 24215
    },
    {
      "epoch": 12.69392033542977,
      "grad_norm": 0.8966953754425049,
      "learning_rate": 1.771826757665112e-05,
      "loss": 0.0765,
      "num_input_tokens_seen": 15839008,
      "step": 24220
    },
    {
      "epoch": 12.696540880503145,
      "grad_norm": 4.753725528717041,
      "learning_rate": 1.770732983173059e-05,
      "loss": 0.0585,
      "num_input_tokens_seen": 15842048,
      "step": 24225
    },
    {
      "epoch": 12.69916142557652,
      "grad_norm": 1.274763584136963,
      "learning_rate": 1.7696393612353467e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 15844992,
      "step": 24230
    },
    {
      "epoch": 12.701781970649895,
      "grad_norm": 1.6243547201156616,
      "learning_rate": 1.7685458920807474e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 15847008,
      "step": 24235
    },
    {
      "epoch": 12.70440251572327,
      "grad_norm": 3.9014720916748047,
      "learning_rate": 1.7674525759380033e-05,
      "loss": 0.1013,
      "num_input_tokens_seen": 15849760,
      "step": 24240
    },
    {
      "epoch": 12.707023060796645,
      "grad_norm": 1.6811267137527466,
      "learning_rate": 1.766359413035823e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 15852608,
      "step": 24245
    },
    {
      "epoch": 12.70964360587002,
      "grad_norm": 2.9184510707855225,
      "learning_rate": 1.7652664036028827e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 15855904,
      "step": 24250
    },
    {
      "epoch": 12.712264150943396,
      "grad_norm": 7.166663646697998,
      "learning_rate": 1.7641735478678288e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 15860224,
      "step": 24255
    },
    {
      "epoch": 12.714884696016771,
      "grad_norm": 7.392668724060059,
      "learning_rate": 1.763080846059274e-05,
      "loss": 0.0458,
      "num_input_tokens_seen": 15863136,
      "step": 24260
    },
    {
      "epoch": 12.717505241090146,
      "grad_norm": 12.447454452514648,
      "learning_rate": 1.7619882984057987e-05,
      "loss": 0.0595,
      "num_input_tokens_seen": 15866400,
      "step": 24265
    },
    {
      "epoch": 12.720125786163521,
      "grad_norm": 0.9719697833061218,
      "learning_rate": 1.760895905135953e-05,
      "loss": 0.0305,
      "num_input_tokens_seen": 15868896,
      "step": 24270
    },
    {
      "epoch": 12.722746331236896,
      "grad_norm": 1.7119488716125488,
      "learning_rate": 1.7598036664782508e-05,
      "loss": 0.0619,
      "num_input_tokens_seen": 15871232,
      "step": 24275
    },
    {
      "epoch": 12.725366876310272,
      "grad_norm": 3.321601629257202,
      "learning_rate": 1.7587115826611766e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 15876800,
      "step": 24280
    },
    {
      "epoch": 12.727987421383649,
      "grad_norm": 1.065973162651062,
      "learning_rate": 1.7576196539131828e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 15879520,
      "step": 24285
    },
    {
      "epoch": 12.730607966457024,
      "grad_norm": 1.2069346904754639,
      "learning_rate": 1.7565278804626872e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 15882432,
      "step": 24290
    },
    {
      "epoch": 12.733228511530399,
      "grad_norm": 2.6094846725463867,
      "learning_rate": 1.755436262538078e-05,
      "loss": 0.0633,
      "num_input_tokens_seen": 15887648,
      "step": 24295
    },
    {
      "epoch": 12.735849056603774,
      "grad_norm": 3.5944554805755615,
      "learning_rate": 1.7543448003677082e-05,
      "loss": 0.0462,
      "num_input_tokens_seen": 15892704,
      "step": 24300
    },
    {
      "epoch": 12.73846960167715,
      "grad_norm": 2.9789063930511475,
      "learning_rate": 1.7532534941798985e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 15895840,
      "step": 24305
    },
    {
      "epoch": 12.741090146750524,
      "grad_norm": 4.282928466796875,
      "learning_rate": 1.752162344202939e-05,
      "loss": 0.0699,
      "num_input_tokens_seen": 15899488,
      "step": 24310
    },
    {
      "epoch": 12.7437106918239,
      "grad_norm": 1.2666743993759155,
      "learning_rate": 1.751071350665085e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 15902816,
      "step": 24315
    },
    {
      "epoch": 12.746331236897275,
      "grad_norm": 1.46170175075531,
      "learning_rate": 1.7499805137945602e-05,
      "loss": 0.0477,
      "num_input_tokens_seen": 15905568,
      "step": 24320
    },
    {
      "epoch": 12.74895178197065,
      "grad_norm": 2.012800455093384,
      "learning_rate": 1.7488898338195566e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 15909120,
      "step": 24325
    },
    {
      "epoch": 12.751572327044025,
      "grad_norm": 1.3144168853759766,
      "learning_rate": 1.7477993109682288e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 15911872,
      "step": 24330
    },
    {
      "epoch": 12.7541928721174,
      "grad_norm": 9.221458435058594,
      "learning_rate": 1.7467089454687043e-05,
      "loss": 0.0598,
      "num_input_tokens_seen": 15914784,
      "step": 24335
    },
    {
      "epoch": 12.756813417190775,
      "grad_norm": 3.021181583404541,
      "learning_rate": 1.745618737549074e-05,
      "loss": 0.0818,
      "num_input_tokens_seen": 15918944,
      "step": 24340
    },
    {
      "epoch": 12.75943396226415,
      "grad_norm": 0.19150547683238983,
      "learning_rate": 1.7445286874373974e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 15921312,
      "step": 24345
    },
    {
      "epoch": 12.762054507337526,
      "grad_norm": 1.9078177213668823,
      "learning_rate": 1.743438795361701e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 15924320,
      "step": 24350
    },
    {
      "epoch": 12.764675052410901,
      "grad_norm": 3.4345479011535645,
      "learning_rate": 1.742349061549978e-05,
      "loss": 0.056,
      "num_input_tokens_seen": 15927936,
      "step": 24355
    },
    {
      "epoch": 12.767295597484276,
      "grad_norm": 1.4267618656158447,
      "learning_rate": 1.741259486230186e-05,
      "loss": 0.0997,
      "num_input_tokens_seen": 15931104,
      "step": 24360
    },
    {
      "epoch": 12.769916142557651,
      "grad_norm": 3.792741537094116,
      "learning_rate": 1.7401700696302546e-05,
      "loss": 0.0718,
      "num_input_tokens_seen": 15935008,
      "step": 24365
    },
    {
      "epoch": 12.772536687631026,
      "grad_norm": 2.1220932006835938,
      "learning_rate": 1.739080811978075e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 15937824,
      "step": 24370
    },
    {
      "epoch": 12.775157232704402,
      "grad_norm": 1.4246044158935547,
      "learning_rate": 1.7379917135015093e-05,
      "loss": 0.0353,
      "num_input_tokens_seen": 15941056,
      "step": 24375
    },
    {
      "epoch": 12.777777777777779,
      "grad_norm": 3.3406765460968018,
      "learning_rate": 1.736902774428384e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 15944640,
      "step": 24380
    },
    {
      "epoch": 12.780398322851154,
      "grad_norm": 3.6178040504455566,
      "learning_rate": 1.735813994986491e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 15948384,
      "step": 24385
    },
    {
      "epoch": 12.783018867924529,
      "grad_norm": 1.4714100360870361,
      "learning_rate": 1.7347253754035933e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 15951232,
      "step": 24390
    },
    {
      "epoch": 12.785639412997904,
      "grad_norm": 4.560977935791016,
      "learning_rate": 1.7336369159074154e-05,
      "loss": 0.0643,
      "num_input_tokens_seen": 15954816,
      "step": 24395
    },
    {
      "epoch": 12.78825995807128,
      "grad_norm": 0.3680558204650879,
      "learning_rate": 1.7325486167256522e-05,
      "loss": 0.2279,
      "num_input_tokens_seen": 15958336,
      "step": 24400
    },
    {
      "epoch": 12.790880503144654,
      "grad_norm": 2.381514310836792,
      "learning_rate": 1.7314604780859623e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 15961984,
      "step": 24405
    },
    {
      "epoch": 12.79350104821803,
      "grad_norm": 1.1027793884277344,
      "learning_rate": 1.730372500215973e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 15965056,
      "step": 24410
    },
    {
      "epoch": 12.796121593291405,
      "grad_norm": 3.7821314334869385,
      "learning_rate": 1.7292846833432763e-05,
      "loss": 0.0507,
      "num_input_tokens_seen": 15967904,
      "step": 24415
    },
    {
      "epoch": 12.79874213836478,
      "grad_norm": 2.1970772743225098,
      "learning_rate": 1.7281970276954305e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 15971424,
      "step": 24420
    },
    {
      "epoch": 12.801362683438155,
      "grad_norm": 2.7542107105255127,
      "learning_rate": 1.727109533499961e-05,
      "loss": 0.1404,
      "num_input_tokens_seen": 15974272,
      "step": 24425
    },
    {
      "epoch": 12.80398322851153,
      "grad_norm": 0.28929081559181213,
      "learning_rate": 1.72602220098436e-05,
      "loss": 0.0535,
      "num_input_tokens_seen": 15977600,
      "step": 24430
    },
    {
      "epoch": 12.806603773584905,
      "grad_norm": 1.681494116783142,
      "learning_rate": 1.7249350303760846e-05,
      "loss": 0.0461,
      "num_input_tokens_seen": 15980640,
      "step": 24435
    },
    {
      "epoch": 12.80922431865828,
      "grad_norm": 4.073733806610107,
      "learning_rate": 1.7238480219025593e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 15983520,
      "step": 24440
    },
    {
      "epoch": 12.811844863731656,
      "grad_norm": 1.8787332773208618,
      "learning_rate": 1.7227611757911723e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 15986592,
      "step": 24445
    },
    {
      "epoch": 12.814465408805031,
      "grad_norm": 4.320318698883057,
      "learning_rate": 1.7216744922692794e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 15989600,
      "step": 24450
    },
    {
      "epoch": 12.817085953878406,
      "grad_norm": 1.0710865259170532,
      "learning_rate": 1.7205879715642037e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 15992416,
      "step": 24455
    },
    {
      "epoch": 12.819706498951781,
      "grad_norm": 3.4400134086608887,
      "learning_rate": 1.719501613903232e-05,
      "loss": 0.0456,
      "num_input_tokens_seen": 15994816,
      "step": 24460
    },
    {
      "epoch": 12.822327044025156,
      "grad_norm": 2.0836639404296875,
      "learning_rate": 1.7184154195136192e-05,
      "loss": 0.0466,
      "num_input_tokens_seen": 15997536,
      "step": 24465
    },
    {
      "epoch": 12.824947589098532,
      "grad_norm": 1.1056246757507324,
      "learning_rate": 1.7173293886225837e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 16001344,
      "step": 24470
    },
    {
      "epoch": 12.827568134171909,
      "grad_norm": 2.9362518787384033,
      "learning_rate": 1.71624352145731e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 16005440,
      "step": 24475
    },
    {
      "epoch": 12.830188679245284,
      "grad_norm": 0.8540970087051392,
      "learning_rate": 1.7151578182449508e-05,
      "loss": 0.0451,
      "num_input_tokens_seen": 16008992,
      "step": 24480
    },
    {
      "epoch": 12.832809224318659,
      "grad_norm": 1.584998369216919,
      "learning_rate": 1.7140722792126212e-05,
      "loss": 0.0308,
      "num_input_tokens_seen": 16011712,
      "step": 24485
    },
    {
      "epoch": 12.835429769392034,
      "grad_norm": 1.422204852104187,
      "learning_rate": 1.7129869045874053e-05,
      "loss": 0.0379,
      "num_input_tokens_seen": 16014528,
      "step": 24490
    },
    {
      "epoch": 12.83805031446541,
      "grad_norm": 1.3662127256393433,
      "learning_rate": 1.7119016945963506e-05,
      "loss": 0.029,
      "num_input_tokens_seen": 16016736,
      "step": 24495
    },
    {
      "epoch": 12.840670859538784,
      "grad_norm": 3.4922473430633545,
      "learning_rate": 1.7108166494664694e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 16020128,
      "step": 24500
    },
    {
      "epoch": 12.84329140461216,
      "grad_norm": 2.2805769443511963,
      "learning_rate": 1.7097317694247417e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 16023008,
      "step": 24505
    },
    {
      "epoch": 12.845911949685535,
      "grad_norm": 4.107669353485107,
      "learning_rate": 1.7086470546981122e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 16026656,
      "step": 24510
    },
    {
      "epoch": 12.84853249475891,
      "grad_norm": 2.791775703430176,
      "learning_rate": 1.7075625055134904e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 16029472,
      "step": 24515
    },
    {
      "epoch": 12.851153039832285,
      "grad_norm": 2.000501871109009,
      "learning_rate": 1.706478122097752e-05,
      "loss": 0.0607,
      "num_input_tokens_seen": 16033504,
      "step": 24520
    },
    {
      "epoch": 12.85377358490566,
      "grad_norm": 0.4581514894962311,
      "learning_rate": 1.7053939046777384e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 16036480,
      "step": 24525
    },
    {
      "epoch": 12.856394129979035,
      "grad_norm": 1.8513599634170532,
      "learning_rate": 1.7043098534802533e-05,
      "loss": 0.0661,
      "num_input_tokens_seen": 16039072,
      "step": 24530
    },
    {
      "epoch": 12.85901467505241,
      "grad_norm": 3.2248659133911133,
      "learning_rate": 1.7032259687320695e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 16042144,
      "step": 24535
    },
    {
      "epoch": 12.861635220125786,
      "grad_norm": 0.9494704604148865,
      "learning_rate": 1.7021422506599228e-05,
      "loss": 0.0513,
      "num_input_tokens_seen": 16044672,
      "step": 24540
    },
    {
      "epoch": 12.864255765199161,
      "grad_norm": 2.347648859024048,
      "learning_rate": 1.7010586994905155e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 16046912,
      "step": 24545
    },
    {
      "epoch": 12.866876310272536,
      "grad_norm": 1.5372495651245117,
      "learning_rate": 1.6999753154505123e-05,
      "loss": 0.0427,
      "num_input_tokens_seen": 16049920,
      "step": 24550
    },
    {
      "epoch": 12.869496855345911,
      "grad_norm": 3.794933319091797,
      "learning_rate": 1.6988920987665476e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 16053952,
      "step": 24555
    },
    {
      "epoch": 12.872117400419286,
      "grad_norm": 4.514016151428223,
      "learning_rate": 1.6978090496652156e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 16056352,
      "step": 24560
    },
    {
      "epoch": 12.874737945492662,
      "grad_norm": 4.349162578582764,
      "learning_rate": 1.6967261683730778e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 16060608,
      "step": 24565
    },
    {
      "epoch": 12.877358490566039,
      "grad_norm": 1.8754363059997559,
      "learning_rate": 1.695643455116662e-05,
      "loss": 0.076,
      "num_input_tokens_seen": 16063904,
      "step": 24570
    },
    {
      "epoch": 12.879979035639414,
      "grad_norm": 0.9608133435249329,
      "learning_rate": 1.694560910122458e-05,
      "loss": 0.0384,
      "num_input_tokens_seen": 16067136,
      "step": 24575
    },
    {
      "epoch": 12.882599580712789,
      "grad_norm": 2.7202537059783936,
      "learning_rate": 1.693478533616924e-05,
      "loss": 0.0251,
      "num_input_tokens_seen": 16069696,
      "step": 24580
    },
    {
      "epoch": 12.885220125786164,
      "grad_norm": 1.6853349208831787,
      "learning_rate": 1.692396325826479e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 16072800,
      "step": 24585
    },
    {
      "epoch": 12.88784067085954,
      "grad_norm": 1.9567400217056274,
      "learning_rate": 1.6913142869775088e-05,
      "loss": 0.0523,
      "num_input_tokens_seen": 16075424,
      "step": 24590
    },
    {
      "epoch": 12.890461215932914,
      "grad_norm": 1.9412243366241455,
      "learning_rate": 1.6902324172963635e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 16078400,
      "step": 24595
    },
    {
      "epoch": 12.89308176100629,
      "grad_norm": 1.6821727752685547,
      "learning_rate": 1.6891507170093586e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 16086432,
      "step": 24600
    },
    {
      "epoch": 12.895702306079665,
      "grad_norm": 2.4941887855529785,
      "learning_rate": 1.6880691863427725e-05,
      "loss": 0.0712,
      "num_input_tokens_seen": 16089472,
      "step": 24605
    },
    {
      "epoch": 12.89832285115304,
      "grad_norm": 2.831077814102173,
      "learning_rate": 1.686987825522851e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 16092800,
      "step": 24610
    },
    {
      "epoch": 12.900943396226415,
      "grad_norm": 1.2149615287780762,
      "learning_rate": 1.6859066347758002e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 16100992,
      "step": 24615
    },
    {
      "epoch": 12.90356394129979,
      "grad_norm": 2.89274263381958,
      "learning_rate": 1.6848256143277933e-05,
      "loss": 0.0568,
      "num_input_tokens_seen": 16104320,
      "step": 24620
    },
    {
      "epoch": 12.906184486373165,
      "grad_norm": 5.626150608062744,
      "learning_rate": 1.683744764404968e-05,
      "loss": 0.0784,
      "num_input_tokens_seen": 16108064,
      "step": 24625
    },
    {
      "epoch": 12.90880503144654,
      "grad_norm": 1.919701337814331,
      "learning_rate": 1.6826640852334257e-05,
      "loss": 0.0713,
      "num_input_tokens_seen": 16111008,
      "step": 24630
    },
    {
      "epoch": 12.911425576519916,
      "grad_norm": 5.505278587341309,
      "learning_rate": 1.6815835770392324e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 16114080,
      "step": 24635
    },
    {
      "epoch": 12.914046121593291,
      "grad_norm": 1.8704334497451782,
      "learning_rate": 1.6805032400484184e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 16116832,
      "step": 24640
    },
    {
      "epoch": 12.916666666666666,
      "grad_norm": 1.0658491849899292,
      "learning_rate": 1.6794230744869767e-05,
      "loss": 0.0282,
      "num_input_tokens_seen": 16119872,
      "step": 24645
    },
    {
      "epoch": 12.919287211740041,
      "grad_norm": 3.751432418823242,
      "learning_rate": 1.6783430805808658e-05,
      "loss": 0.0845,
      "num_input_tokens_seen": 16123200,
      "step": 24650
    },
    {
      "epoch": 12.921907756813416,
      "grad_norm": 2.788883924484253,
      "learning_rate": 1.6772632585560084e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 16126048,
      "step": 24655
    },
    {
      "epoch": 12.924528301886792,
      "grad_norm": 3.945524215698242,
      "learning_rate": 1.6761836086382908e-05,
      "loss": 0.0498,
      "num_input_tokens_seen": 16129088,
      "step": 24660
    },
    {
      "epoch": 12.927148846960169,
      "grad_norm": 2.217316150665283,
      "learning_rate": 1.6751041310535653e-05,
      "loss": 0.0797,
      "num_input_tokens_seen": 16131584,
      "step": 24665
    },
    {
      "epoch": 12.929769392033544,
      "grad_norm": 5.893285751342773,
      "learning_rate": 1.6740248260276437e-05,
      "loss": 0.0742,
      "num_input_tokens_seen": 16133984,
      "step": 24670
    },
    {
      "epoch": 12.932389937106919,
      "grad_norm": 1.5548980236053467,
      "learning_rate": 1.6729456937863047e-05,
      "loss": 0.0319,
      "num_input_tokens_seen": 16136480,
      "step": 24675
    },
    {
      "epoch": 12.935010482180294,
      "grad_norm": 0.6029900312423706,
      "learning_rate": 1.6718667345552918e-05,
      "loss": 0.0186,
      "num_input_tokens_seen": 16139616,
      "step": 24680
    },
    {
      "epoch": 12.93763102725367,
      "grad_norm": 4.21518611907959,
      "learning_rate": 1.6707879485603097e-05,
      "loss": 0.0283,
      "num_input_tokens_seen": 16143008,
      "step": 24685
    },
    {
      "epoch": 12.940251572327044,
      "grad_norm": 2.6374263763427734,
      "learning_rate": 1.669709336027029e-05,
      "loss": 0.0574,
      "num_input_tokens_seen": 16146304,
      "step": 24690
    },
    {
      "epoch": 12.94287211740042,
      "grad_norm": 4.241422176361084,
      "learning_rate": 1.6686308971810833e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 16149184,
      "step": 24695
    },
    {
      "epoch": 12.945492662473795,
      "grad_norm": 2.2584550380706787,
      "learning_rate": 1.6675526322480677e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 16151968,
      "step": 24700
    },
    {
      "epoch": 12.94811320754717,
      "grad_norm": 1.5246049165725708,
      "learning_rate": 1.6664745414535454e-05,
      "loss": 0.0682,
      "num_input_tokens_seen": 16154848,
      "step": 24705
    },
    {
      "epoch": 12.950733752620545,
      "grad_norm": 1.1346659660339355,
      "learning_rate": 1.665396625023039e-05,
      "loss": 0.0491,
      "num_input_tokens_seen": 16158336,
      "step": 24710
    },
    {
      "epoch": 12.95335429769392,
      "grad_norm": 2.786837577819824,
      "learning_rate": 1.6643188831820375e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 16161856,
      "step": 24715
    },
    {
      "epoch": 12.955974842767295,
      "grad_norm": 2.310468912124634,
      "learning_rate": 1.6632413161559912e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 16165696,
      "step": 24720
    },
    {
      "epoch": 12.95859538784067,
      "grad_norm": 7.169255256652832,
      "learning_rate": 1.6621639241703162e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 16169440,
      "step": 24725
    },
    {
      "epoch": 12.961215932914046,
      "grad_norm": 8.089010238647461,
      "learning_rate": 1.6610867074503893e-05,
      "loss": 0.0249,
      "num_input_tokens_seen": 16173088,
      "step": 24730
    },
    {
      "epoch": 12.963836477987421,
      "grad_norm": 1.161290168762207,
      "learning_rate": 1.660009666221552e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 16176352,
      "step": 24735
    },
    {
      "epoch": 12.966457023060796,
      "grad_norm": 4.198596000671387,
      "learning_rate": 1.6589328007091093e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 16180128,
      "step": 24740
    },
    {
      "epoch": 12.969077568134171,
      "grad_norm": 2.352010726928711,
      "learning_rate": 1.65785611113833e-05,
      "loss": 0.0838,
      "num_input_tokens_seen": 16182880,
      "step": 24745
    },
    {
      "epoch": 12.971698113207546,
      "grad_norm": 3.1719727516174316,
      "learning_rate": 1.6567795977344447e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 16186880,
      "step": 24750
    },
    {
      "epoch": 12.974318658280922,
      "grad_norm": 11.3502836227417,
      "learning_rate": 1.6557032607226466e-05,
      "loss": 0.0887,
      "num_input_tokens_seen": 16189600,
      "step": 24755
    },
    {
      "epoch": 12.976939203354299,
      "grad_norm": 0.9630809426307678,
      "learning_rate": 1.654627100328095e-05,
      "loss": 0.1328,
      "num_input_tokens_seen": 16194976,
      "step": 24760
    },
    {
      "epoch": 12.979559748427674,
      "grad_norm": 2.4815196990966797,
      "learning_rate": 1.6535511167759085e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 16197536,
      "step": 24765
    },
    {
      "epoch": 12.982180293501049,
      "grad_norm": 2.243436574935913,
      "learning_rate": 1.652475310291172e-05,
      "loss": 0.062,
      "num_input_tokens_seen": 16201440,
      "step": 24770
    },
    {
      "epoch": 12.984800838574424,
      "grad_norm": 1.1426048278808594,
      "learning_rate": 1.6513996810989317e-05,
      "loss": 0.055,
      "num_input_tokens_seen": 16204448,
      "step": 24775
    },
    {
      "epoch": 12.9874213836478,
      "grad_norm": 2.2135939598083496,
      "learning_rate": 1.6503242294241973e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 16207584,
      "step": 24780
    },
    {
      "epoch": 12.990041928721174,
      "grad_norm": 1.1566938161849976,
      "learning_rate": 1.6492489554919404e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 16209920,
      "step": 24785
    },
    {
      "epoch": 12.99266247379455,
      "grad_norm": 1.3133453130722046,
      "learning_rate": 1.6481738595270953e-05,
      "loss": 0.0279,
      "num_input_tokens_seen": 16212736,
      "step": 24790
    },
    {
      "epoch": 12.995283018867925,
      "grad_norm": 2.461709976196289,
      "learning_rate": 1.6470989417545608e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 16216480,
      "step": 24795
    },
    {
      "epoch": 12.9979035639413,
      "grad_norm": 4.909161567687988,
      "learning_rate": 1.646024202399197e-05,
      "loss": 0.0673,
      "num_input_tokens_seen": 16218976,
      "step": 24800
    },
    {
      "epoch": 13.0,
      "eval_loss": 1.1541403532028198,
      "eval_runtime": 11.733,
      "eval_samples_per_second": 72.275,
      "eval_steps_per_second": 18.069,
      "num_input_tokens_seen": 16220808,
      "step": 24804
    },
    {
      "epoch": 13.000524109014675,
      "grad_norm": 0.9362913370132446,
      "learning_rate": 1.6449496416858284e-05,
      "loss": 0.0367,
      "num_input_tokens_seen": 16221160,
      "step": 24805
    },
    {
      "epoch": 13.00314465408805,
      "grad_norm": 1.1197419166564941,
      "learning_rate": 1.64387525983924e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 16223848,
      "step": 24810
    },
    {
      "epoch": 13.005765199161425,
      "grad_norm": 0.5585929751396179,
      "learning_rate": 1.6428010570841788e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 16226856,
      "step": 24815
    },
    {
      "epoch": 13.0083857442348,
      "grad_norm": 0.5406925678253174,
      "learning_rate": 1.6417270336453573e-05,
      "loss": 0.0204,
      "num_input_tokens_seen": 16229736,
      "step": 24820
    },
    {
      "epoch": 13.011006289308176,
      "grad_norm": 2.485931873321533,
      "learning_rate": 1.6406531897474492e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 16233064,
      "step": 24825
    },
    {
      "epoch": 13.01362683438155,
      "grad_norm": 1.0132625102996826,
      "learning_rate": 1.639579525615089e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 16236456,
      "step": 24830
    },
    {
      "epoch": 13.016247379454926,
      "grad_norm": 2.822192907333374,
      "learning_rate": 1.638506041472877e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 16239272,
      "step": 24835
    },
    {
      "epoch": 13.018867924528301,
      "grad_norm": 0.09234961867332458,
      "learning_rate": 1.6374327375453724e-05,
      "loss": 0.0115,
      "num_input_tokens_seen": 16242152,
      "step": 24840
    },
    {
      "epoch": 13.021488469601676,
      "grad_norm": 3.1202573776245117,
      "learning_rate": 1.6363596140570982e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 16245832,
      "step": 24845
    },
    {
      "epoch": 13.024109014675052,
      "grad_norm": 0.8264515399932861,
      "learning_rate": 1.6352866712325395e-05,
      "loss": 0.0193,
      "num_input_tokens_seen": 16249064,
      "step": 24850
    },
    {
      "epoch": 13.026729559748428,
      "grad_norm": 1.8069547414779663,
      "learning_rate": 1.634213909296144e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 16251944,
      "step": 24855
    },
    {
      "epoch": 13.029350104821804,
      "grad_norm": 2.2799344062805176,
      "learning_rate": 1.6331413284723217e-05,
      "loss": 0.0851,
      "num_input_tokens_seen": 16255368,
      "step": 24860
    },
    {
      "epoch": 13.031970649895179,
      "grad_norm": 2.09096360206604,
      "learning_rate": 1.6320689289854442e-05,
      "loss": 0.0529,
      "num_input_tokens_seen": 16258152,
      "step": 24865
    },
    {
      "epoch": 13.034591194968554,
      "grad_norm": 0.7804759740829468,
      "learning_rate": 1.6309967110598438e-05,
      "loss": 0.0385,
      "num_input_tokens_seen": 16261672,
      "step": 24870
    },
    {
      "epoch": 13.03721174004193,
      "grad_norm": 1.0145732164382935,
      "learning_rate": 1.6299246749198183e-05,
      "loss": 0.0224,
      "num_input_tokens_seen": 16265416,
      "step": 24875
    },
    {
      "epoch": 13.039832285115304,
      "grad_norm": 1.7742173671722412,
      "learning_rate": 1.6288528207896236e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 16268008,
      "step": 24880
    },
    {
      "epoch": 13.04245283018868,
      "grad_norm": 2.0130085945129395,
      "learning_rate": 1.627781148893481e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 16271912,
      "step": 24885
    },
    {
      "epoch": 13.045073375262055,
      "grad_norm": 1.9326071739196777,
      "learning_rate": 1.6267096594555703e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 16275304,
      "step": 24890
    },
    {
      "epoch": 13.04769392033543,
      "grad_norm": 1.4128012657165527,
      "learning_rate": 1.6256383527000372e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 16279240,
      "step": 24895
    },
    {
      "epoch": 13.050314465408805,
      "grad_norm": 4.757261276245117,
      "learning_rate": 1.6245672288509843e-05,
      "loss": 0.1096,
      "num_input_tokens_seen": 16282376,
      "step": 24900
    },
    {
      "epoch": 13.05293501048218,
      "grad_norm": 0.994536817073822,
      "learning_rate": 1.62349628813248e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 16285000,
      "step": 24905
    },
    {
      "epoch": 13.055555555555555,
      "grad_norm": 7.765084743499756,
      "learning_rate": 1.6224255307685525e-05,
      "loss": 0.0226,
      "num_input_tokens_seen": 16287976,
      "step": 24910
    },
    {
      "epoch": 13.05817610062893,
      "grad_norm": 0.26982080936431885,
      "learning_rate": 1.6213549569831922e-05,
      "loss": 0.0425,
      "num_input_tokens_seen": 16291080,
      "step": 24915
    },
    {
      "epoch": 13.060796645702306,
      "grad_norm": 3.7797775268554688,
      "learning_rate": 1.62028456700035e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 16294504,
      "step": 24920
    },
    {
      "epoch": 13.06341719077568,
      "grad_norm": 1.2706644535064697,
      "learning_rate": 1.619214361043942e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 16298696,
      "step": 24925
    },
    {
      "epoch": 13.066037735849056,
      "grad_norm": 2.2317488193511963,
      "learning_rate": 1.61814433933784e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 16302536,
      "step": 24930
    },
    {
      "epoch": 13.068658280922431,
      "grad_norm": 0.955058217048645,
      "learning_rate": 1.617074502105881e-05,
      "loss": 0.0249,
      "num_input_tokens_seen": 16305992,
      "step": 24935
    },
    {
      "epoch": 13.071278825995806,
      "grad_norm": 13.149081230163574,
      "learning_rate": 1.616004849571864e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 16309768,
      "step": 24940
    },
    {
      "epoch": 13.073899371069182,
      "grad_norm": 3.5379576683044434,
      "learning_rate": 1.6149353819595463e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 16313160,
      "step": 24945
    },
    {
      "epoch": 13.076519916142558,
      "grad_norm": 1.332486867904663,
      "learning_rate": 1.6138660994926508e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 16315688,
      "step": 24950
    },
    {
      "epoch": 13.079140461215934,
      "grad_norm": 22.611391067504883,
      "learning_rate": 1.612797002394857e-05,
      "loss": 0.035,
      "num_input_tokens_seen": 16319208,
      "step": 24955
    },
    {
      "epoch": 13.081761006289309,
      "grad_norm": 2.266671657562256,
      "learning_rate": 1.611728090889808e-05,
      "loss": 0.0397,
      "num_input_tokens_seen": 16322600,
      "step": 24960
    },
    {
      "epoch": 13.084381551362684,
      "grad_norm": 6.492519855499268,
      "learning_rate": 1.610659365201109e-05,
      "loss": 0.0454,
      "num_input_tokens_seen": 16326568,
      "step": 24965
    },
    {
      "epoch": 13.08700209643606,
      "grad_norm": 5.165815830230713,
      "learning_rate": 1.609590825552324e-05,
      "loss": 0.058,
      "num_input_tokens_seen": 16331336,
      "step": 24970
    },
    {
      "epoch": 13.089622641509434,
      "grad_norm": 2.131074905395508,
      "learning_rate": 1.6085224721669807e-05,
      "loss": 0.057,
      "num_input_tokens_seen": 16334824,
      "step": 24975
    },
    {
      "epoch": 13.09224318658281,
      "grad_norm": 2.660583257675171,
      "learning_rate": 1.607454305268566e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 16337992,
      "step": 24980
    },
    {
      "epoch": 13.094863731656185,
      "grad_norm": 0.8377470374107361,
      "learning_rate": 1.606386325080528e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 16340680,
      "step": 24985
    },
    {
      "epoch": 13.09748427672956,
      "grad_norm": 0.9473900198936462,
      "learning_rate": 1.6053185318262746e-05,
      "loss": 0.0176,
      "num_input_tokens_seen": 16343464,
      "step": 24990
    },
    {
      "epoch": 13.100104821802935,
      "grad_norm": 0.7672808766365051,
      "learning_rate": 1.6042509257291783e-05,
      "loss": 0.0511,
      "num_input_tokens_seen": 16346376,
      "step": 24995
    },
    {
      "epoch": 13.10272536687631,
      "grad_norm": 0.9597046971321106,
      "learning_rate": 1.6031835070125685e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 16349032,
      "step": 25000
    },
    {
      "epoch": 13.105345911949685,
      "grad_norm": 4.761418342590332,
      "learning_rate": 1.6021162758997382e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 16352104,
      "step": 25005
    },
    {
      "epoch": 13.10796645702306,
      "grad_norm": 0.6748215556144714,
      "learning_rate": 1.6010492326139398e-05,
      "loss": 0.0608,
      "num_input_tokens_seen": 16355496,
      "step": 25010
    },
    {
      "epoch": 13.110587002096436,
      "grad_norm": 0.7858273386955261,
      "learning_rate": 1.599982377378385e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 16359080,
      "step": 25015
    },
    {
      "epoch": 13.11320754716981,
      "grad_norm": 0.6117845773696899,
      "learning_rate": 1.5989157104162496e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 16362408,
      "step": 25020
    },
    {
      "epoch": 13.115828092243186,
      "grad_norm": 1.1613712310791016,
      "learning_rate": 1.5978492319506673e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 16365384,
      "step": 25025
    },
    {
      "epoch": 13.118448637316561,
      "grad_norm": 1.31854248046875,
      "learning_rate": 1.596782942204733e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 16368840,
      "step": 25030
    },
    {
      "epoch": 13.121069182389936,
      "grad_norm": 3.709364652633667,
      "learning_rate": 1.595716841401504e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 16372296,
      "step": 25035
    },
    {
      "epoch": 13.123689727463312,
      "grad_norm": 0.12158603221178055,
      "learning_rate": 1.594650929763994e-05,
      "loss": 0.013,
      "num_input_tokens_seen": 16375400,
      "step": 25040
    },
    {
      "epoch": 13.126310272536688,
      "grad_norm": 1.8004200458526611,
      "learning_rate": 1.5935852075151812e-05,
      "loss": 0.0616,
      "num_input_tokens_seen": 16378728,
      "step": 25045
    },
    {
      "epoch": 13.128930817610064,
      "grad_norm": 3.723858118057251,
      "learning_rate": 1.5925196748780016e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 16381672,
      "step": 25050
    },
    {
      "epoch": 13.131551362683439,
      "grad_norm": 1.0013995170593262,
      "learning_rate": 1.5914543320753528e-05,
      "loss": 0.0855,
      "num_input_tokens_seen": 16384744,
      "step": 25055
    },
    {
      "epoch": 13.134171907756814,
      "grad_norm": 3.319044589996338,
      "learning_rate": 1.5903891793300932e-05,
      "loss": 0.1223,
      "num_input_tokens_seen": 16387816,
      "step": 25060
    },
    {
      "epoch": 13.13679245283019,
      "grad_norm": 1.332429051399231,
      "learning_rate": 1.5893242168650403e-05,
      "loss": 0.0216,
      "num_input_tokens_seen": 16390600,
      "step": 25065
    },
    {
      "epoch": 13.139412997903564,
      "grad_norm": 3.2233903408050537,
      "learning_rate": 1.5882594449029705e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 16394280,
      "step": 25070
    },
    {
      "epoch": 13.14203354297694,
      "grad_norm": 1.9252910614013672,
      "learning_rate": 1.587194863666624e-05,
      "loss": 0.0207,
      "num_input_tokens_seen": 16396904,
      "step": 25075
    },
    {
      "epoch": 13.144654088050315,
      "grad_norm": 1.100732445716858,
      "learning_rate": 1.5861304733786974e-05,
      "loss": 0.0998,
      "num_input_tokens_seen": 16400136,
      "step": 25080
    },
    {
      "epoch": 13.14727463312369,
      "grad_norm": 0.7325779795646667,
      "learning_rate": 1.5850662742618505e-05,
      "loss": 0.0086,
      "num_input_tokens_seen": 16402568,
      "step": 25085
    },
    {
      "epoch": 13.149895178197065,
      "grad_norm": 1.1993969678878784,
      "learning_rate": 1.5840022665387004e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 16405832,
      "step": 25090
    },
    {
      "epoch": 13.15251572327044,
      "grad_norm": 4.332397937774658,
      "learning_rate": 1.5829384504318268e-05,
      "loss": 0.0602,
      "num_input_tokens_seen": 16409288,
      "step": 25095
    },
    {
      "epoch": 13.155136268343815,
      "grad_norm": 1.4894251823425293,
      "learning_rate": 1.5818748261637662e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 16412584,
      "step": 25100
    },
    {
      "epoch": 13.15775681341719,
      "grad_norm": 2.231900215148926,
      "learning_rate": 1.5808113939570173e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 16415080,
      "step": 25105
    },
    {
      "epoch": 13.160377358490566,
      "grad_norm": 0.4737462103366852,
      "learning_rate": 1.5797481540340386e-05,
      "loss": 0.0175,
      "num_input_tokens_seen": 16418184,
      "step": 25110
    },
    {
      "epoch": 13.16299790356394,
      "grad_norm": 0.3359353840351105,
      "learning_rate": 1.5786851066172466e-05,
      "loss": 0.0866,
      "num_input_tokens_seen": 16421672,
      "step": 25115
    },
    {
      "epoch": 13.165618448637316,
      "grad_norm": 1.8180298805236816,
      "learning_rate": 1.5776222519290208e-05,
      "loss": 0.021,
      "num_input_tokens_seen": 16425448,
      "step": 25120
    },
    {
      "epoch": 13.168238993710691,
      "grad_norm": 2.1172597408294678,
      "learning_rate": 1.576559590191696e-05,
      "loss": 0.0496,
      "num_input_tokens_seen": 16428872,
      "step": 25125
    },
    {
      "epoch": 13.170859538784066,
      "grad_norm": 17.160297393798828,
      "learning_rate": 1.5754971216275692e-05,
      "loss": 0.1238,
      "num_input_tokens_seen": 16431560,
      "step": 25130
    },
    {
      "epoch": 13.173480083857442,
      "grad_norm": 2.7081565856933594,
      "learning_rate": 1.5744348464588975e-05,
      "loss": 0.0244,
      "num_input_tokens_seen": 16435048,
      "step": 25135
    },
    {
      "epoch": 13.176100628930818,
      "grad_norm": 1.525179386138916,
      "learning_rate": 1.5733727649078968e-05,
      "loss": 0.0211,
      "num_input_tokens_seen": 16438248,
      "step": 25140
    },
    {
      "epoch": 13.178721174004194,
      "grad_norm": 2.1567792892456055,
      "learning_rate": 1.5723108771967414e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 16441384,
      "step": 25145
    },
    {
      "epoch": 13.181341719077569,
      "grad_norm": 3.552480936050415,
      "learning_rate": 1.5712491835475685e-05,
      "loss": 0.093,
      "num_input_tokens_seen": 16445992,
      "step": 25150
    },
    {
      "epoch": 13.183962264150944,
      "grad_norm": 0.959029495716095,
      "learning_rate": 1.5701876841824698e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 16449416,
      "step": 25155
    },
    {
      "epoch": 13.18658280922432,
      "grad_norm": 1.7574979066848755,
      "learning_rate": 1.5691263793234993e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 16454472,
      "step": 25160
    },
    {
      "epoch": 13.189203354297694,
      "grad_norm": 0.6505897045135498,
      "learning_rate": 1.5680652691926706e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 16456808,
      "step": 25165
    },
    {
      "epoch": 13.19182389937107,
      "grad_norm": 1.2843239307403564,
      "learning_rate": 1.5670043540119553e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 16459752,
      "step": 25170
    },
    {
      "epoch": 13.194444444444445,
      "grad_norm": 2.2839553356170654,
      "learning_rate": 1.5659436340032855e-05,
      "loss": 0.0243,
      "num_input_tokens_seen": 16462760,
      "step": 25175
    },
    {
      "epoch": 13.19706498951782,
      "grad_norm": 2.5266289710998535,
      "learning_rate": 1.5648831093885515e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 16466536,
      "step": 25180
    },
    {
      "epoch": 13.199685534591195,
      "grad_norm": 5.750988960266113,
      "learning_rate": 1.5638227803896015e-05,
      "loss": 0.068,
      "num_input_tokens_seen": 16469384,
      "step": 25185
    },
    {
      "epoch": 13.20230607966457,
      "grad_norm": 0.5697463154792786,
      "learning_rate": 1.562762647228246e-05,
      "loss": 0.0149,
      "num_input_tokens_seen": 16472488,
      "step": 25190
    },
    {
      "epoch": 13.204926624737945,
      "grad_norm": 2.3514292240142822,
      "learning_rate": 1.5617027101262515e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 16475528,
      "step": 25195
    },
    {
      "epoch": 13.20754716981132,
      "grad_norm": 1.3555915355682373,
      "learning_rate": 1.5606429693053463e-05,
      "loss": 0.0234,
      "num_input_tokens_seen": 16478216,
      "step": 25200
    },
    {
      "epoch": 13.210167714884696,
      "grad_norm": 2.5339138507843018,
      "learning_rate": 1.5595834249872145e-05,
      "loss": 0.0834,
      "num_input_tokens_seen": 16481480,
      "step": 25205
    },
    {
      "epoch": 13.21278825995807,
      "grad_norm": 2.6307597160339355,
      "learning_rate": 1.558524077393502e-05,
      "loss": 0.032,
      "num_input_tokens_seen": 16485768,
      "step": 25210
    },
    {
      "epoch": 13.215408805031446,
      "grad_norm": 1.340245008468628,
      "learning_rate": 1.557464926745811e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 16490152,
      "step": 25215
    },
    {
      "epoch": 13.218029350104821,
      "grad_norm": 0.9008194208145142,
      "learning_rate": 1.5564059732657045e-05,
      "loss": 0.0245,
      "num_input_tokens_seen": 16493096,
      "step": 25220
    },
    {
      "epoch": 13.220649895178196,
      "grad_norm": 6.814749717712402,
      "learning_rate": 1.555347217174703e-05,
      "loss": 0.0387,
      "num_input_tokens_seen": 16495848,
      "step": 25225
    },
    {
      "epoch": 13.223270440251572,
      "grad_norm": 2.8407695293426514,
      "learning_rate": 1.554288658694287e-05,
      "loss": 0.0731,
      "num_input_tokens_seen": 16502888,
      "step": 25230
    },
    {
      "epoch": 13.225890985324948,
      "grad_norm": 1.2651829719543457,
      "learning_rate": 1.553230298045895e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 16505928,
      "step": 25235
    },
    {
      "epoch": 13.228511530398324,
      "grad_norm": 2.9762122631073,
      "learning_rate": 1.5521721354509223e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 16508648,
      "step": 25240
    },
    {
      "epoch": 13.231132075471699,
      "grad_norm": 0.7583364844322205,
      "learning_rate": 1.5511141711307258e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 16512968,
      "step": 25245
    },
    {
      "epoch": 13.233752620545074,
      "grad_norm": 1.19231379032135,
      "learning_rate": 1.5500564053066184e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 16516552,
      "step": 25250
    },
    {
      "epoch": 13.23637316561845,
      "grad_norm": 2.502117872238159,
      "learning_rate": 1.5489988381998743e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 16519528,
      "step": 25255
    },
    {
      "epoch": 13.238993710691824,
      "grad_norm": 3.5446665287017822,
      "learning_rate": 1.5479414700317237e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 16522504,
      "step": 25260
    },
    {
      "epoch": 13.2416142557652,
      "grad_norm": 0.714229941368103,
      "learning_rate": 1.546884301023356e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 16525384,
      "step": 25265
    },
    {
      "epoch": 13.244234800838575,
      "grad_norm": 1.0036113262176514,
      "learning_rate": 1.5458273313959186e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 16528840,
      "step": 25270
    },
    {
      "epoch": 13.24685534591195,
      "grad_norm": 2.4580342769622803,
      "learning_rate": 1.5447705613705177e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 16532648,
      "step": 25275
    },
    {
      "epoch": 13.249475890985325,
      "grad_norm": 3.8049917221069336,
      "learning_rate": 1.5437139911682176e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 16535720,
      "step": 25280
    },
    {
      "epoch": 13.2520964360587,
      "grad_norm": 0.8024536371231079,
      "learning_rate": 1.54265762101004e-05,
      "loss": 0.0149,
      "num_input_tokens_seen": 16538312,
      "step": 25285
    },
    {
      "epoch": 13.254716981132075,
      "grad_norm": 1.5882443189620972,
      "learning_rate": 1.5416014511169663e-05,
      "loss": 0.0864,
      "num_input_tokens_seen": 16541640,
      "step": 25290
    },
    {
      "epoch": 13.25733752620545,
      "grad_norm": 0.1482858508825302,
      "learning_rate": 1.5405454817099363e-05,
      "loss": 0.0774,
      "num_input_tokens_seen": 16545640,
      "step": 25295
    },
    {
      "epoch": 13.259958071278826,
      "grad_norm": 1.8505390882492065,
      "learning_rate": 1.5394897130098446e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 16548488,
      "step": 25300
    },
    {
      "epoch": 13.2625786163522,
      "grad_norm": 0.3038007318973541,
      "learning_rate": 1.5384341452375465e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 16551592,
      "step": 25305
    },
    {
      "epoch": 13.265199161425576,
      "grad_norm": 1.169418454170227,
      "learning_rate": 1.5373787786138556e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 16554600,
      "step": 25310
    },
    {
      "epoch": 13.267819706498951,
      "grad_norm": 0.5646571516990662,
      "learning_rate": 1.5363236133595416e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 16557192,
      "step": 25315
    },
    {
      "epoch": 13.270440251572326,
      "grad_norm": 2.265223741531372,
      "learning_rate": 1.5352686496953344e-05,
      "loss": 0.0352,
      "num_input_tokens_seen": 16560040,
      "step": 25320
    },
    {
      "epoch": 13.273060796645701,
      "grad_norm": 2.845276355743408,
      "learning_rate": 1.534213887841919e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 16563304,
      "step": 25325
    },
    {
      "epoch": 13.275681341719078,
      "grad_norm": 6.7770094871521,
      "learning_rate": 1.5331593280199395e-05,
      "loss": 0.027,
      "num_input_tokens_seen": 16566312,
      "step": 25330
    },
    {
      "epoch": 13.278301886792454,
      "grad_norm": 0.8473243713378906,
      "learning_rate": 1.532104970449999e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 16569736,
      "step": 25335
    },
    {
      "epoch": 13.280922431865829,
      "grad_norm": 0.656802237033844,
      "learning_rate": 1.5310508153526555e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 16572680,
      "step": 25340
    },
    {
      "epoch": 13.283542976939204,
      "grad_norm": 2.6968212127685547,
      "learning_rate": 1.5299968629484276e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 16576168,
      "step": 25345
    },
    {
      "epoch": 13.286163522012579,
      "grad_norm": 0.7672199606895447,
      "learning_rate": 1.5289431134577904e-05,
      "loss": 0.0153,
      "num_input_tokens_seen": 16578504,
      "step": 25350
    },
    {
      "epoch": 13.288784067085954,
      "grad_norm": 3.115903377532959,
      "learning_rate": 1.527889567101174e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 16582472,
      "step": 25355
    },
    {
      "epoch": 13.29140461215933,
      "grad_norm": 0.7378748059272766,
      "learning_rate": 1.5268362240989707e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 16585128,
      "step": 25360
    },
    {
      "epoch": 13.294025157232705,
      "grad_norm": 2.208568811416626,
      "learning_rate": 1.525783084671526e-05,
      "loss": 0.0197,
      "num_input_tokens_seen": 16587592,
      "step": 25365
    },
    {
      "epoch": 13.29664570230608,
      "grad_norm": 3.328939199447632,
      "learning_rate": 1.5247301490391458e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 16594600,
      "step": 25370
    },
    {
      "epoch": 13.299266247379455,
      "grad_norm": 1.3726131916046143,
      "learning_rate": 1.5236774174220922e-05,
      "loss": 0.0105,
      "num_input_tokens_seen": 16598120,
      "step": 25375
    },
    {
      "epoch": 13.30188679245283,
      "grad_norm": 5.050185203552246,
      "learning_rate": 1.5226248900405852e-05,
      "loss": 0.0392,
      "num_input_tokens_seen": 16601768,
      "step": 25380
    },
    {
      "epoch": 13.304507337526205,
      "grad_norm": 1.5612235069274902,
      "learning_rate": 1.5215725671147996e-05,
      "loss": 0.0174,
      "num_input_tokens_seen": 16604776,
      "step": 25385
    },
    {
      "epoch": 13.30712788259958,
      "grad_norm": 0.9914645552635193,
      "learning_rate": 1.5205204488648709e-05,
      "loss": 0.0481,
      "num_input_tokens_seen": 16607688,
      "step": 25390
    },
    {
      "epoch": 13.309748427672956,
      "grad_norm": 1.243024230003357,
      "learning_rate": 1.5194685355108893e-05,
      "loss": 0.0198,
      "num_input_tokens_seen": 16610920,
      "step": 25395
    },
    {
      "epoch": 13.31236897274633,
      "grad_norm": 14.549079895019531,
      "learning_rate": 1.518416827272904e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 16614024,
      "step": 25400
    },
    {
      "epoch": 13.314989517819706,
      "grad_norm": 2.642590284347534,
      "learning_rate": 1.5173653243709207e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 16616936,
      "step": 25405
    },
    {
      "epoch": 13.317610062893081,
      "grad_norm": 2.219639539718628,
      "learning_rate": 1.5163140270249e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 16619432,
      "step": 25410
    },
    {
      "epoch": 13.320230607966456,
      "grad_norm": 2.7209415435791016,
      "learning_rate": 1.5152629354547626e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 16622152,
      "step": 25415
    },
    {
      "epoch": 13.322851153039831,
      "grad_norm": 2.1350650787353516,
      "learning_rate": 1.5142120498803842e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 16624808,
      "step": 25420
    },
    {
      "epoch": 13.325471698113208,
      "grad_norm": 3.4125139713287354,
      "learning_rate": 1.5131613705215991e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 16628744,
      "step": 25425
    },
    {
      "epoch": 13.328092243186584,
      "grad_norm": 2.100674867630005,
      "learning_rate": 1.5121108975981965e-05,
      "loss": 0.089,
      "num_input_tokens_seen": 16631368,
      "step": 25430
    },
    {
      "epoch": 13.330712788259959,
      "grad_norm": 1.0253411531448364,
      "learning_rate": 1.5110606313299248e-05,
      "loss": 0.0119,
      "num_input_tokens_seen": 16634536,
      "step": 25435
    },
    {
      "epoch": 13.333333333333334,
      "grad_norm": 2.537071943283081,
      "learning_rate": 1.5100105719364855e-05,
      "loss": 0.0559,
      "num_input_tokens_seen": 16637544,
      "step": 25440
    },
    {
      "epoch": 13.335953878406709,
      "grad_norm": 0.9287161827087402,
      "learning_rate": 1.50896071963754e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 16641096,
      "step": 25445
    },
    {
      "epoch": 13.338574423480084,
      "grad_norm": 1.7581629753112793,
      "learning_rate": 1.5079110746527055e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 16643688,
      "step": 25450
    },
    {
      "epoch": 13.34119496855346,
      "grad_norm": 5.975470542907715,
      "learning_rate": 1.5068616372015565e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 16646888,
      "step": 25455
    },
    {
      "epoch": 13.343815513626835,
      "grad_norm": 0.6720540523529053,
      "learning_rate": 1.5058124075036223e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 16650312,
      "step": 25460
    },
    {
      "epoch": 13.34643605870021,
      "grad_norm": 0.9916610717773438,
      "learning_rate": 1.5047633857783908e-05,
      "loss": 0.0164,
      "num_input_tokens_seen": 16653896,
      "step": 25465
    },
    {
      "epoch": 13.349056603773585,
      "grad_norm": 0.6511126756668091,
      "learning_rate": 1.5037145722453044e-05,
      "loss": 0.0193,
      "num_input_tokens_seen": 16657320,
      "step": 25470
    },
    {
      "epoch": 13.35167714884696,
      "grad_norm": 1.9307050704956055,
      "learning_rate": 1.5026659671237624e-05,
      "loss": 0.036,
      "num_input_tokens_seen": 16660456,
      "step": 25475
    },
    {
      "epoch": 13.354297693920335,
      "grad_norm": 1.0038673877716064,
      "learning_rate": 1.501617570633123e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 16664424,
      "step": 25480
    },
    {
      "epoch": 13.35691823899371,
      "grad_norm": 1.3369415998458862,
      "learning_rate": 1.500569382992697e-05,
      "loss": 0.0467,
      "num_input_tokens_seen": 16667656,
      "step": 25485
    },
    {
      "epoch": 13.359538784067086,
      "grad_norm": 1.6260207891464233,
      "learning_rate": 1.4995214044217547e-05,
      "loss": 0.044,
      "num_input_tokens_seen": 16671688,
      "step": 25490
    },
    {
      "epoch": 13.36215932914046,
      "grad_norm": 1.4109041690826416,
      "learning_rate": 1.49847363513952e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 16674856,
      "step": 25495
    },
    {
      "epoch": 13.364779874213836,
      "grad_norm": 1.2856217622756958,
      "learning_rate": 1.4974260753651748e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 16677864,
      "step": 25500
    },
    {
      "epoch": 13.367400419287211,
      "grad_norm": 4.607205867767334,
      "learning_rate": 1.4963787253178568e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 16680808,
      "step": 25505
    },
    {
      "epoch": 13.370020964360586,
      "grad_norm": 2.1404740810394287,
      "learning_rate": 1.4953315852166594e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 16683208,
      "step": 25510
    },
    {
      "epoch": 13.372641509433961,
      "grad_norm": 2.5768396854400635,
      "learning_rate": 1.4942846552806322e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 16686952,
      "step": 25515
    },
    {
      "epoch": 13.375262054507338,
      "grad_norm": 1.9776009321212769,
      "learning_rate": 1.4932379357287829e-05,
      "loss": 0.0285,
      "num_input_tokens_seen": 16689576,
      "step": 25520
    },
    {
      "epoch": 13.377882599580714,
      "grad_norm": 2.137551784515381,
      "learning_rate": 1.49219142678007e-05,
      "loss": 0.0364,
      "num_input_tokens_seen": 16693320,
      "step": 25525
    },
    {
      "epoch": 13.380503144654089,
      "grad_norm": 0.846396267414093,
      "learning_rate": 1.4911451286534128e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 16698344,
      "step": 25530
    },
    {
      "epoch": 13.383123689727464,
      "grad_norm": 1.0130767822265625,
      "learning_rate": 1.490099041567686e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 16701576,
      "step": 25535
    },
    {
      "epoch": 13.385744234800839,
      "grad_norm": 3.191086530685425,
      "learning_rate": 1.4890531657417176e-05,
      "loss": 0.0233,
      "num_input_tokens_seen": 16704904,
      "step": 25540
    },
    {
      "epoch": 13.388364779874214,
      "grad_norm": 1.8702199459075928,
      "learning_rate": 1.4880075013942938e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 16707304,
      "step": 25545
    },
    {
      "epoch": 13.39098532494759,
      "grad_norm": 0.4184364676475525,
      "learning_rate": 1.4869620487441558e-05,
      "loss": 0.088,
      "num_input_tokens_seen": 16710120,
      "step": 25550
    },
    {
      "epoch": 13.393605870020965,
      "grad_norm": 1.582487940788269,
      "learning_rate": 1.4859168080099988e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 16713736,
      "step": 25555
    },
    {
      "epoch": 13.39622641509434,
      "grad_norm": 1.8269081115722656,
      "learning_rate": 1.484871779410477e-05,
      "loss": 0.0292,
      "num_input_tokens_seen": 16716616,
      "step": 25560
    },
    {
      "epoch": 13.398846960167715,
      "grad_norm": 2.6536877155303955,
      "learning_rate": 1.4838269631641971e-05,
      "loss": 0.0701,
      "num_input_tokens_seen": 16719432,
      "step": 25565
    },
    {
      "epoch": 13.40146750524109,
      "grad_norm": 1.1640722751617432,
      "learning_rate": 1.482782359489724e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 16723176,
      "step": 25570
    },
    {
      "epoch": 13.404088050314465,
      "grad_norm": 1.874040126800537,
      "learning_rate": 1.4817379686055754e-05,
      "loss": 0.0252,
      "num_input_tokens_seen": 16725608,
      "step": 25575
    },
    {
      "epoch": 13.40670859538784,
      "grad_norm": 0.9781525731086731,
      "learning_rate": 1.4806937907302282e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 16729320,
      "step": 25580
    },
    {
      "epoch": 13.409329140461216,
      "grad_norm": 2.0787546634674072,
      "learning_rate": 1.4796498260821103e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 16731976,
      "step": 25585
    },
    {
      "epoch": 13.41194968553459,
      "grad_norm": 2.6291286945343018,
      "learning_rate": 1.4786060748796076e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 16734792,
      "step": 25590
    },
    {
      "epoch": 13.414570230607966,
      "grad_norm": 1.2167129516601562,
      "learning_rate": 1.4775625373410617e-05,
      "loss": 0.0366,
      "num_input_tokens_seen": 16737608,
      "step": 25595
    },
    {
      "epoch": 13.417190775681341,
      "grad_norm": 0.6064425706863403,
      "learning_rate": 1.4765192136847685e-05,
      "loss": 0.0396,
      "num_input_tokens_seen": 16741000,
      "step": 25600
    },
    {
      "epoch": 13.419811320754716,
      "grad_norm": 3.337883234024048,
      "learning_rate": 1.47547610412898e-05,
      "loss": 0.0226,
      "num_input_tokens_seen": 16743944,
      "step": 25605
    },
    {
      "epoch": 13.422431865828091,
      "grad_norm": 3.802692174911499,
      "learning_rate": 1.4744332088919011e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 16747176,
      "step": 25610
    },
    {
      "epoch": 13.425052410901468,
      "grad_norm": 3.353125810623169,
      "learning_rate": 1.4733905281916954e-05,
      "loss": 0.0597,
      "num_input_tokens_seen": 16749928,
      "step": 25615
    },
    {
      "epoch": 13.427672955974844,
      "grad_norm": 0.5073763728141785,
      "learning_rate": 1.472348062246478e-05,
      "loss": 0.0118,
      "num_input_tokens_seen": 16753128,
      "step": 25620
    },
    {
      "epoch": 13.430293501048219,
      "grad_norm": 1.0294833183288574,
      "learning_rate": 1.4713058112743228e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 16756104,
      "step": 25625
    },
    {
      "epoch": 13.432914046121594,
      "grad_norm": 1.4420723915100098,
      "learning_rate": 1.4702637754932553e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 16758824,
      "step": 25630
    },
    {
      "epoch": 13.435534591194969,
      "grad_norm": 6.171036720275879,
      "learning_rate": 1.4692219551212587e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 16761224,
      "step": 25635
    },
    {
      "epoch": 13.438155136268344,
      "grad_norm": 3.3706178665161133,
      "learning_rate": 1.4681803503762695e-05,
      "loss": 0.0343,
      "num_input_tokens_seen": 16764072,
      "step": 25640
    },
    {
      "epoch": 13.44077568134172,
      "grad_norm": 2.094705581665039,
      "learning_rate": 1.4671389614761788e-05,
      "loss": 0.0362,
      "num_input_tokens_seen": 16767560,
      "step": 25645
    },
    {
      "epoch": 13.443396226415095,
      "grad_norm": 2.083425283432007,
      "learning_rate": 1.4660977886388339e-05,
      "loss": 0.0176,
      "num_input_tokens_seen": 16770504,
      "step": 25650
    },
    {
      "epoch": 13.44601677148847,
      "grad_norm": 1.8171372413635254,
      "learning_rate": 1.4650568320820362e-05,
      "loss": 0.0441,
      "num_input_tokens_seen": 16773224,
      "step": 25655
    },
    {
      "epoch": 13.448637316561845,
      "grad_norm": 2.096881628036499,
      "learning_rate": 1.4640160920235423e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 16775880,
      "step": 25660
    },
    {
      "epoch": 13.45125786163522,
      "grad_norm": 8.078121185302734,
      "learning_rate": 1.4629755686810631e-05,
      "loss": 0.0342,
      "num_input_tokens_seen": 16778920,
      "step": 25665
    },
    {
      "epoch": 13.453878406708595,
      "grad_norm": 2.880236864089966,
      "learning_rate": 1.4619352622722632e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 16781544,
      "step": 25670
    },
    {
      "epoch": 13.45649895178197,
      "grad_norm": 2.236726999282837,
      "learning_rate": 1.4608951730147635e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 16784264,
      "step": 25675
    },
    {
      "epoch": 13.459119496855346,
      "grad_norm": 0.7651851177215576,
      "learning_rate": 1.4598553011261395e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 16787176,
      "step": 25680
    },
    {
      "epoch": 13.46174004192872,
      "grad_norm": 2.1246097087860107,
      "learning_rate": 1.4588156468239192e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 16790376,
      "step": 25685
    },
    {
      "epoch": 13.464360587002096,
      "grad_norm": 0.6452711820602417,
      "learning_rate": 1.4577762103255882e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 16793608,
      "step": 25690
    },
    {
      "epoch": 13.466981132075471,
      "grad_norm": 1.7807680368423462,
      "learning_rate": 1.4567369918485828e-05,
      "loss": 0.0412,
      "num_input_tokens_seen": 16796552,
      "step": 25695
    },
    {
      "epoch": 13.469601677148846,
      "grad_norm": 0.9557117819786072,
      "learning_rate": 1.4556979916102976e-05,
      "loss": 0.0265,
      "num_input_tokens_seen": 16799432,
      "step": 25700
    },
    {
      "epoch": 13.472222222222221,
      "grad_norm": 0.3468354344367981,
      "learning_rate": 1.454659209828077e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 16802184,
      "step": 25705
    },
    {
      "epoch": 13.474842767295598,
      "grad_norm": 14.60519027709961,
      "learning_rate": 1.4536206467192243e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 16807208,
      "step": 25710
    },
    {
      "epoch": 13.477463312368974,
      "grad_norm": 2.7594566345214844,
      "learning_rate": 1.4525823025009944e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 16810856,
      "step": 25715
    },
    {
      "epoch": 13.480083857442349,
      "grad_norm": 0.4441141188144684,
      "learning_rate": 1.4515441773905983e-05,
      "loss": 0.0189,
      "num_input_tokens_seen": 16813736,
      "step": 25720
    },
    {
      "epoch": 13.482704402515724,
      "grad_norm": 2.3401148319244385,
      "learning_rate": 1.4505062716051988e-05,
      "loss": 0.0332,
      "num_input_tokens_seen": 16817928,
      "step": 25725
    },
    {
      "epoch": 13.485324947589099,
      "grad_norm": 1.6995402574539185,
      "learning_rate": 1.4494685853619133e-05,
      "loss": 0.0182,
      "num_input_tokens_seen": 16820872,
      "step": 25730
    },
    {
      "epoch": 13.487945492662474,
      "grad_norm": 0.4762003719806671,
      "learning_rate": 1.4484311188778143e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 16823464,
      "step": 25735
    },
    {
      "epoch": 13.49056603773585,
      "grad_norm": 2.1019372940063477,
      "learning_rate": 1.4473938723699285e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 16826632,
      "step": 25740
    },
    {
      "epoch": 13.493186582809225,
      "grad_norm": 4.3179097175598145,
      "learning_rate": 1.4463568460552368e-05,
      "loss": 0.029,
      "num_input_tokens_seen": 16829640,
      "step": 25745
    },
    {
      "epoch": 13.4958071278826,
      "grad_norm": 0.659760594367981,
      "learning_rate": 1.4453200401506723e-05,
      "loss": 0.0174,
      "num_input_tokens_seen": 16832456,
      "step": 25750
    },
    {
      "epoch": 13.498427672955975,
      "grad_norm": 1.4689395427703857,
      "learning_rate": 1.4442834548731218e-05,
      "loss": 0.0164,
      "num_input_tokens_seen": 16835208,
      "step": 25755
    },
    {
      "epoch": 13.50104821802935,
      "grad_norm": 2.5735645294189453,
      "learning_rate": 1.4432470904394285e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 16838632,
      "step": 25760
    },
    {
      "epoch": 13.503668763102725,
      "grad_norm": 1.2059539556503296,
      "learning_rate": 1.4422109470663878e-05,
      "loss": 0.0177,
      "num_input_tokens_seen": 16842088,
      "step": 25765
    },
    {
      "epoch": 13.5062893081761,
      "grad_norm": 2.233794689178467,
      "learning_rate": 1.441175024970749e-05,
      "loss": 0.0201,
      "num_input_tokens_seen": 16844488,
      "step": 25770
    },
    {
      "epoch": 13.508909853249476,
      "grad_norm": 15.868612289428711,
      "learning_rate": 1.4401393243692163e-05,
      "loss": 0.0647,
      "num_input_tokens_seen": 16848712,
      "step": 25775
    },
    {
      "epoch": 13.51153039832285,
      "grad_norm": 2.6892216205596924,
      "learning_rate": 1.4391038454784455e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 16852232,
      "step": 25780
    },
    {
      "epoch": 13.514150943396226,
      "grad_norm": 1.5353896617889404,
      "learning_rate": 1.4380685885150458e-05,
      "loss": 0.0558,
      "num_input_tokens_seen": 16855368,
      "step": 25785
    },
    {
      "epoch": 13.516771488469601,
      "grad_norm": 1.628306269645691,
      "learning_rate": 1.437033553695582e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 16858536,
      "step": 25790
    },
    {
      "epoch": 13.519392033542976,
      "grad_norm": 0.8507859706878662,
      "learning_rate": 1.4359987412365721e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 16862152,
      "step": 25795
    },
    {
      "epoch": 13.522012578616351,
      "grad_norm": 5.549081802368164,
      "learning_rate": 1.4349641513544876e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 16864904,
      "step": 25800
    },
    {
      "epoch": 13.524633123689728,
      "grad_norm": 1.5153846740722656,
      "learning_rate": 1.4339297842657523e-05,
      "loss": 0.0537,
      "num_input_tokens_seen": 16868136,
      "step": 25805
    },
    {
      "epoch": 13.527253668763104,
      "grad_norm": 0.8212335705757141,
      "learning_rate": 1.4328956401867427e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 16871432,
      "step": 25810
    },
    {
      "epoch": 13.529874213836479,
      "grad_norm": 1.1932374238967896,
      "learning_rate": 1.4318617193337908e-05,
      "loss": 0.0587,
      "num_input_tokens_seen": 16874120,
      "step": 25815
    },
    {
      "epoch": 13.532494758909854,
      "grad_norm": 1.7933629751205444,
      "learning_rate": 1.4308280219231812e-05,
      "loss": 0.0564,
      "num_input_tokens_seen": 16878088,
      "step": 25820
    },
    {
      "epoch": 13.535115303983229,
      "grad_norm": 2.706962823867798,
      "learning_rate": 1.4297945481711527e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 16880776,
      "step": 25825
    },
    {
      "epoch": 13.537735849056604,
      "grad_norm": 1.2647219896316528,
      "learning_rate": 1.4287612982938946e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 16884360,
      "step": 25830
    },
    {
      "epoch": 13.54035639412998,
      "grad_norm": 0.6885787844657898,
      "learning_rate": 1.4277282725075523e-05,
      "loss": 0.1038,
      "num_input_tokens_seen": 16887080,
      "step": 25835
    },
    {
      "epoch": 13.542976939203355,
      "grad_norm": 3.455132007598877,
      "learning_rate": 1.426695471028221e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 16890536,
      "step": 25840
    },
    {
      "epoch": 13.54559748427673,
      "grad_norm": 0.4089815616607666,
      "learning_rate": 1.4256628940719524e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 16893992,
      "step": 25845
    },
    {
      "epoch": 13.548218029350105,
      "grad_norm": 3.2913784980773926,
      "learning_rate": 1.42463054185475e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 16896904,
      "step": 25850
    },
    {
      "epoch": 13.55083857442348,
      "grad_norm": 2.1504018306732178,
      "learning_rate": 1.4235984145925705e-05,
      "loss": 0.0168,
      "num_input_tokens_seen": 16899880,
      "step": 25855
    },
    {
      "epoch": 13.553459119496855,
      "grad_norm": 1.1251418590545654,
      "learning_rate": 1.4225665125013224e-05,
      "loss": 0.026,
      "num_input_tokens_seen": 16903528,
      "step": 25860
    },
    {
      "epoch": 13.55607966457023,
      "grad_norm": 5.220943927764893,
      "learning_rate": 1.4215348357968669e-05,
      "loss": 0.0429,
      "num_input_tokens_seen": 16905736,
      "step": 25865
    },
    {
      "epoch": 13.558700209643606,
      "grad_norm": 2.0383753776550293,
      "learning_rate": 1.42050338469502e-05,
      "loss": 0.0322,
      "num_input_tokens_seen": 16909256,
      "step": 25870
    },
    {
      "epoch": 13.56132075471698,
      "grad_norm": 0.4763370454311371,
      "learning_rate": 1.4194721594115495e-05,
      "loss": 0.0248,
      "num_input_tokens_seen": 16911976,
      "step": 25875
    },
    {
      "epoch": 13.563941299790356,
      "grad_norm": 1.3410446643829346,
      "learning_rate": 1.4184411601621769e-05,
      "loss": 0.0134,
      "num_input_tokens_seen": 16915592,
      "step": 25880
    },
    {
      "epoch": 13.566561844863731,
      "grad_norm": 1.8261661529541016,
      "learning_rate": 1.4174103871625738e-05,
      "loss": 0.0329,
      "num_input_tokens_seen": 16919080,
      "step": 25885
    },
    {
      "epoch": 13.569182389937106,
      "grad_norm": 1.0807758569717407,
      "learning_rate": 1.4163798406283674e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 16922632,
      "step": 25890
    },
    {
      "epoch": 13.571802935010481,
      "grad_norm": 1.7637455463409424,
      "learning_rate": 1.415349520775135e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 16925736,
      "step": 25895
    },
    {
      "epoch": 13.574423480083858,
      "grad_norm": 1.8634058237075806,
      "learning_rate": 1.4143194278184086e-05,
      "loss": 0.0198,
      "num_input_tokens_seen": 16929448,
      "step": 25900
    },
    {
      "epoch": 13.577044025157234,
      "grad_norm": 2.4902846813201904,
      "learning_rate": 1.4132895619736724e-05,
      "loss": 0.0792,
      "num_input_tokens_seen": 16932584,
      "step": 25905
    },
    {
      "epoch": 13.579664570230609,
      "grad_norm": 2.7390849590301514,
      "learning_rate": 1.4122599234563613e-05,
      "loss": 0.0734,
      "num_input_tokens_seen": 16936072,
      "step": 25910
    },
    {
      "epoch": 13.582285115303984,
      "grad_norm": 1.2388471364974976,
      "learning_rate": 1.4112305124818647e-05,
      "loss": 0.0171,
      "num_input_tokens_seen": 16939464,
      "step": 25915
    },
    {
      "epoch": 13.584905660377359,
      "grad_norm": 2.77535343170166,
      "learning_rate": 1.4102013292655242e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 16943912,
      "step": 25920
    },
    {
      "epoch": 13.587526205450734,
      "grad_norm": 7.7995524406433105,
      "learning_rate": 1.409172374022632e-05,
      "loss": 0.0486,
      "num_input_tokens_seen": 16947464,
      "step": 25925
    },
    {
      "epoch": 13.59014675052411,
      "grad_norm": 1.0866880416870117,
      "learning_rate": 1.4081436469684339e-05,
      "loss": 0.0237,
      "num_input_tokens_seen": 16950504,
      "step": 25930
    },
    {
      "epoch": 13.592767295597485,
      "grad_norm": 1.3038674592971802,
      "learning_rate": 1.4071151483181294e-05,
      "loss": 0.0612,
      "num_input_tokens_seen": 16953672,
      "step": 25935
    },
    {
      "epoch": 13.59538784067086,
      "grad_norm": 3.4848744869232178,
      "learning_rate": 1.4060868782868669e-05,
      "loss": 0.0148,
      "num_input_tokens_seen": 16956520,
      "step": 25940
    },
    {
      "epoch": 13.598008385744235,
      "grad_norm": 4.216692924499512,
      "learning_rate": 1.405058837089749e-05,
      "loss": 0.0865,
      "num_input_tokens_seen": 16958888,
      "step": 25945
    },
    {
      "epoch": 13.60062893081761,
      "grad_norm": 4.634052276611328,
      "learning_rate": 1.404031024941832e-05,
      "loss": 0.0294,
      "num_input_tokens_seen": 16962216,
      "step": 25950
    },
    {
      "epoch": 13.603249475890985,
      "grad_norm": 1.5218143463134766,
      "learning_rate": 1.40300344205812e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 16966152,
      "step": 25955
    },
    {
      "epoch": 13.60587002096436,
      "grad_norm": 0.3762449026107788,
      "learning_rate": 1.4019760886535738e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 16969096,
      "step": 25960
    },
    {
      "epoch": 13.608490566037736,
      "grad_norm": 0.43541213870048523,
      "learning_rate": 1.4009489649431018e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 16971592,
      "step": 25965
    },
    {
      "epoch": 13.61111111111111,
      "grad_norm": 3.22739577293396,
      "learning_rate": 1.3999220711415675e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 16974056,
      "step": 25970
    },
    {
      "epoch": 13.613731656184486,
      "grad_norm": 0.5044981837272644,
      "learning_rate": 1.3988954074637867e-05,
      "loss": 0.0109,
      "num_input_tokens_seen": 16976008,
      "step": 25975
    },
    {
      "epoch": 13.616352201257861,
      "grad_norm": 1.8394826650619507,
      "learning_rate": 1.3978689741245238e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 16979080,
      "step": 25980
    },
    {
      "epoch": 13.618972746331236,
      "grad_norm": 2.0819170475006104,
      "learning_rate": 1.3968427713384986e-05,
      "loss": 0.0508,
      "num_input_tokens_seen": 16982216,
      "step": 25985
    },
    {
      "epoch": 13.621593291404611,
      "grad_norm": 1.5592259168624878,
      "learning_rate": 1.3958167993203791e-05,
      "loss": 0.0304,
      "num_input_tokens_seen": 16985288,
      "step": 25990
    },
    {
      "epoch": 13.624213836477988,
      "grad_norm": 1.022223949432373,
      "learning_rate": 1.3947910582847879e-05,
      "loss": 0.015,
      "num_input_tokens_seen": 16988264,
      "step": 25995
    },
    {
      "epoch": 13.626834381551364,
      "grad_norm": 0.8750319480895996,
      "learning_rate": 1.3937655484462985e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 16990920,
      "step": 26000
    },
    {
      "epoch": 13.629454926624739,
      "grad_norm": 1.4362722635269165,
      "learning_rate": 1.392740270019437e-05,
      "loss": 0.0702,
      "num_input_tokens_seen": 16993768,
      "step": 26005
    },
    {
      "epoch": 13.632075471698114,
      "grad_norm": 0.9389566779136658,
      "learning_rate": 1.3917152232186776e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 16996584,
      "step": 26010
    },
    {
      "epoch": 13.634696016771489,
      "grad_norm": 3.8254082202911377,
      "learning_rate": 1.3906904082584504e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 17000040,
      "step": 26015
    },
    {
      "epoch": 13.637316561844864,
      "grad_norm": 2.2203028202056885,
      "learning_rate": 1.3896658253531334e-05,
      "loss": 0.0234,
      "num_input_tokens_seen": 17002696,
      "step": 26020
    },
    {
      "epoch": 13.63993710691824,
      "grad_norm": 1.0953950881958008,
      "learning_rate": 1.3886414747170581e-05,
      "loss": 0.0298,
      "num_input_tokens_seen": 17005992,
      "step": 26025
    },
    {
      "epoch": 13.642557651991615,
      "grad_norm": 4.2027740478515625,
      "learning_rate": 1.3876173565645072e-05,
      "loss": 0.0478,
      "num_input_tokens_seen": 17008904,
      "step": 26030
    },
    {
      "epoch": 13.64517819706499,
      "grad_norm": 0.17848290503025055,
      "learning_rate": 1.3865934711097156e-05,
      "loss": 0.0262,
      "num_input_tokens_seen": 17011880,
      "step": 26035
    },
    {
      "epoch": 13.647798742138365,
      "grad_norm": 3.781497001647949,
      "learning_rate": 1.3855698185668675e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 17015208,
      "step": 26040
    },
    {
      "epoch": 13.65041928721174,
      "grad_norm": 1.162235140800476,
      "learning_rate": 1.3845463991500985e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 17018664,
      "step": 26045
    },
    {
      "epoch": 13.653039832285115,
      "grad_norm": 2.396284341812134,
      "learning_rate": 1.3835232130734968e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 17022280,
      "step": 26050
    },
    {
      "epoch": 13.65566037735849,
      "grad_norm": 15.173179626464844,
      "learning_rate": 1.3825002605511017e-05,
      "loss": 0.0527,
      "num_input_tokens_seen": 17026344,
      "step": 26055
    },
    {
      "epoch": 13.658280922431866,
      "grad_norm": 1.6031303405761719,
      "learning_rate": 1.3814775417969037e-05,
      "loss": 0.0189,
      "num_input_tokens_seen": 17030024,
      "step": 26060
    },
    {
      "epoch": 13.66090146750524,
      "grad_norm": 4.391366958618164,
      "learning_rate": 1.3804550570248431e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 17033992,
      "step": 26065
    },
    {
      "epoch": 13.663522012578616,
      "grad_norm": 0.7702339291572571,
      "learning_rate": 1.3794328064488116e-05,
      "loss": 0.0409,
      "num_input_tokens_seen": 17040936,
      "step": 26070
    },
    {
      "epoch": 13.666142557651991,
      "grad_norm": 1.122131109237671,
      "learning_rate": 1.3784107902826523e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 17043432,
      "step": 26075
    },
    {
      "epoch": 13.668763102725366,
      "grad_norm": 3.9865710735321045,
      "learning_rate": 1.3773890087401603e-05,
      "loss": 0.0391,
      "num_input_tokens_seen": 17046184,
      "step": 26080
    },
    {
      "epoch": 13.671383647798741,
      "grad_norm": 2.3764808177948,
      "learning_rate": 1.3763674620350802e-05,
      "loss": 0.0351,
      "num_input_tokens_seen": 17049160,
      "step": 26085
    },
    {
      "epoch": 13.674004192872118,
      "grad_norm": 0.1559046506881714,
      "learning_rate": 1.3753461503811089e-05,
      "loss": 0.0423,
      "num_input_tokens_seen": 17052584,
      "step": 26090
    },
    {
      "epoch": 13.676624737945493,
      "grad_norm": 0.9907568097114563,
      "learning_rate": 1.3743250739918922e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 17055624,
      "step": 26095
    },
    {
      "epoch": 13.679245283018869,
      "grad_norm": 6.035277843475342,
      "learning_rate": 1.3733042330810265e-05,
      "loss": 0.0434,
      "num_input_tokens_seen": 17058280,
      "step": 26100
    },
    {
      "epoch": 13.681865828092244,
      "grad_norm": 2.063601016998291,
      "learning_rate": 1.3722836278620616e-05,
      "loss": 0.0328,
      "num_input_tokens_seen": 17061256,
      "step": 26105
    },
    {
      "epoch": 13.684486373165619,
      "grad_norm": 3.4588747024536133,
      "learning_rate": 1.3712632585484957e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 17065160,
      "step": 26110
    },
    {
      "epoch": 13.687106918238994,
      "grad_norm": 1.052006483078003,
      "learning_rate": 1.37024312535378e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 17067816,
      "step": 26115
    },
    {
      "epoch": 13.68972746331237,
      "grad_norm": 1.5551178455352783,
      "learning_rate": 1.3692232284913133e-05,
      "loss": 0.0383,
      "num_input_tokens_seen": 17070952,
      "step": 26120
    },
    {
      "epoch": 13.692348008385745,
      "grad_norm": 0.5466210246086121,
      "learning_rate": 1.3682035681744459e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 17073928,
      "step": 26125
    },
    {
      "epoch": 13.69496855345912,
      "grad_norm": 0.7759193778038025,
      "learning_rate": 1.3671841446164795e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 17077192,
      "step": 26130
    },
    {
      "epoch": 13.697589098532495,
      "grad_norm": 2.2707762718200684,
      "learning_rate": 1.3661649580306669e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 17080360,
      "step": 26135
    },
    {
      "epoch": 13.70020964360587,
      "grad_norm": 1.4674474000930786,
      "learning_rate": 1.3651460086302087e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 17083304,
      "step": 26140
    },
    {
      "epoch": 13.702830188679245,
      "grad_norm": 2.114197254180908,
      "learning_rate": 1.3641272966282609e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 17086952,
      "step": 26145
    },
    {
      "epoch": 13.70545073375262,
      "grad_norm": 1.18375825881958,
      "learning_rate": 1.3631088222379218e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 17091656,
      "step": 26150
    },
    {
      "epoch": 13.708071278825996,
      "grad_norm": 1.841254711151123,
      "learning_rate": 1.3620905856722468e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 17094888,
      "step": 26155
    },
    {
      "epoch": 13.71069182389937,
      "grad_norm": 1.9800059795379639,
      "learning_rate": 1.3610725871442392e-05,
      "loss": 0.0625,
      "num_input_tokens_seen": 17098856,
      "step": 26160
    },
    {
      "epoch": 13.713312368972746,
      "grad_norm": 0.21648840606212616,
      "learning_rate": 1.3600548268668532e-05,
      "loss": 0.0198,
      "num_input_tokens_seen": 17101992,
      "step": 26165
    },
    {
      "epoch": 13.715932914046121,
      "grad_norm": 0.889519453048706,
      "learning_rate": 1.3590373050529927e-05,
      "loss": 0.0446,
      "num_input_tokens_seen": 17104520,
      "step": 26170
    },
    {
      "epoch": 13.718553459119496,
      "grad_norm": 0.8964023590087891,
      "learning_rate": 1.3580200219155116e-05,
      "loss": 0.1029,
      "num_input_tokens_seen": 17107944,
      "step": 26175
    },
    {
      "epoch": 13.721174004192871,
      "grad_norm": 0.8207173347473145,
      "learning_rate": 1.3570029776672122e-05,
      "loss": 0.0372,
      "num_input_tokens_seen": 17111016,
      "step": 26180
    },
    {
      "epoch": 13.723794549266248,
      "grad_norm": 2.5317201614379883,
      "learning_rate": 1.35598617252085e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 17113544,
      "step": 26185
    },
    {
      "epoch": 13.726415094339622,
      "grad_norm": 2.324080467224121,
      "learning_rate": 1.354969606689129e-05,
      "loss": 0.0285,
      "num_input_tokens_seen": 17116168,
      "step": 26190
    },
    {
      "epoch": 13.729035639412999,
      "grad_norm": 1.524329423904419,
      "learning_rate": 1.353953280384704e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 17120264,
      "step": 26195
    },
    {
      "epoch": 13.731656184486374,
      "grad_norm": 1.0235782861709595,
      "learning_rate": 1.352937193820177e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 17123080,
      "step": 26200
    },
    {
      "epoch": 13.734276729559749,
      "grad_norm": 0.16657979786396027,
      "learning_rate": 1.3519213472081038e-05,
      "loss": 0.0405,
      "num_input_tokens_seen": 17125960,
      "step": 26205
    },
    {
      "epoch": 13.736897274633124,
      "grad_norm": 1.3625409603118896,
      "learning_rate": 1.3509057407609854e-05,
      "loss": 0.0345,
      "num_input_tokens_seen": 17129544,
      "step": 26210
    },
    {
      "epoch": 13.7395178197065,
      "grad_norm": 0.684516966342926,
      "learning_rate": 1.3498903746912769e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 17133384,
      "step": 26215
    },
    {
      "epoch": 13.742138364779874,
      "grad_norm": 4.418755531311035,
      "learning_rate": 1.3488752492113807e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 17136008,
      "step": 26220
    },
    {
      "epoch": 13.74475890985325,
      "grad_norm": 0.6167290210723877,
      "learning_rate": 1.3478603645336508e-05,
      "loss": 0.0679,
      "num_input_tokens_seen": 17139144,
      "step": 26225
    },
    {
      "epoch": 13.747379454926625,
      "grad_norm": 0.41780373454093933,
      "learning_rate": 1.3468457208703877e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 17142280,
      "step": 26230
    },
    {
      "epoch": 13.75,
      "grad_norm": 1.489955186843872,
      "learning_rate": 1.3458313184338437e-05,
      "loss": 0.0655,
      "num_input_tokens_seen": 17146664,
      "step": 26235
    },
    {
      "epoch": 13.752620545073375,
      "grad_norm": 1.3961831331253052,
      "learning_rate": 1.3448171574362201e-05,
      "loss": 0.0442,
      "num_input_tokens_seen": 17149384,
      "step": 26240
    },
    {
      "epoch": 13.75524109014675,
      "grad_norm": 1.5862265825271606,
      "learning_rate": 1.3438032380896681e-05,
      "loss": 0.0538,
      "num_input_tokens_seen": 17152328,
      "step": 26245
    },
    {
      "epoch": 13.757861635220126,
      "grad_norm": 0.2930346131324768,
      "learning_rate": 1.342789560606289e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 17155624,
      "step": 26250
    },
    {
      "epoch": 13.7604821802935,
      "grad_norm": 1.6525534391403198,
      "learning_rate": 1.341776125198131e-05,
      "loss": 0.0265,
      "num_input_tokens_seen": 17158376,
      "step": 26255
    },
    {
      "epoch": 13.763102725366876,
      "grad_norm": 0.4574817717075348,
      "learning_rate": 1.3407629320771948e-05,
      "loss": 0.0422,
      "num_input_tokens_seen": 17161320,
      "step": 26260
    },
    {
      "epoch": 13.765723270440251,
      "grad_norm": 0.6034703254699707,
      "learning_rate": 1.3397499814554266e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 17165288,
      "step": 26265
    },
    {
      "epoch": 13.768343815513626,
      "grad_norm": 2.348598003387451,
      "learning_rate": 1.3387372735447256e-05,
      "loss": 0.0295,
      "num_input_tokens_seen": 17168456,
      "step": 26270
    },
    {
      "epoch": 13.770964360587001,
      "grad_norm": 0.8198205828666687,
      "learning_rate": 1.3377248085569399e-05,
      "loss": 0.0315,
      "num_input_tokens_seen": 17171496,
      "step": 26275
    },
    {
      "epoch": 13.773584905660378,
      "grad_norm": 2.676492691040039,
      "learning_rate": 1.336712586703863e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 17174824,
      "step": 26280
    },
    {
      "epoch": 13.776205450733752,
      "grad_norm": 2.8982112407684326,
      "learning_rate": 1.3357006081972414e-05,
      "loss": 0.0863,
      "num_input_tokens_seen": 17177064,
      "step": 26285
    },
    {
      "epoch": 13.778825995807129,
      "grad_norm": 0.8292302489280701,
      "learning_rate": 1.3346888732487706e-05,
      "loss": 0.0091,
      "num_input_tokens_seen": 17179752,
      "step": 26290
    },
    {
      "epoch": 13.781446540880504,
      "grad_norm": 2.4145216941833496,
      "learning_rate": 1.3336773820700921e-05,
      "loss": 0.0447,
      "num_input_tokens_seen": 17184232,
      "step": 26295
    },
    {
      "epoch": 13.784067085953879,
      "grad_norm": 2.7946863174438477,
      "learning_rate": 1.3326661348727992e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 17187624,
      "step": 26300
    },
    {
      "epoch": 13.786687631027254,
      "grad_norm": 0.4546522796154022,
      "learning_rate": 1.3316551318684342e-05,
      "loss": 0.0165,
      "num_input_tokens_seen": 17190088,
      "step": 26305
    },
    {
      "epoch": 13.78930817610063,
      "grad_norm": 1.3911076784133911,
      "learning_rate": 1.3306443732684854e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 17194216,
      "step": 26310
    },
    {
      "epoch": 13.791928721174004,
      "grad_norm": 2.9421467781066895,
      "learning_rate": 1.3296338592843929e-05,
      "loss": 0.0311,
      "num_input_tokens_seen": 17197544,
      "step": 26315
    },
    {
      "epoch": 13.79454926624738,
      "grad_norm": 1.0172170400619507,
      "learning_rate": 1.3286235901275457e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 17200488,
      "step": 26320
    },
    {
      "epoch": 13.797169811320755,
      "grad_norm": 0.9140568971633911,
      "learning_rate": 1.327613566009279e-05,
      "loss": 0.0171,
      "num_input_tokens_seen": 17202824,
      "step": 26325
    },
    {
      "epoch": 13.79979035639413,
      "grad_norm": 0.3562050759792328,
      "learning_rate": 1.3266037871408798e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 17206632,
      "step": 26330
    },
    {
      "epoch": 13.802410901467505,
      "grad_norm": 0.40996384620666504,
      "learning_rate": 1.3255942537335805e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 17212648,
      "step": 26335
    },
    {
      "epoch": 13.80503144654088,
      "grad_norm": 2.8840925693511963,
      "learning_rate": 1.3245849659985652e-05,
      "loss": 0.0175,
      "num_input_tokens_seen": 17215176,
      "step": 26340
    },
    {
      "epoch": 13.807651991614255,
      "grad_norm": 2.535845994949341,
      "learning_rate": 1.3235759241469658e-05,
      "loss": 0.0146,
      "num_input_tokens_seen": 17217672,
      "step": 26345
    },
    {
      "epoch": 13.81027253668763,
      "grad_norm": 0.9351651668548584,
      "learning_rate": 1.322567128389861e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 17221096,
      "step": 26350
    },
    {
      "epoch": 13.812893081761006,
      "grad_norm": 3.3509483337402344,
      "learning_rate": 1.3215585789382812e-05,
      "loss": 0.0226,
      "num_input_tokens_seen": 17223784,
      "step": 26355
    },
    {
      "epoch": 13.815513626834381,
      "grad_norm": 13.284460067749023,
      "learning_rate": 1.320550276003201e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 17227496,
      "step": 26360
    },
    {
      "epoch": 13.818134171907756,
      "grad_norm": 3.395928382873535,
      "learning_rate": 1.3195422197955473e-05,
      "loss": 0.041,
      "num_input_tokens_seen": 17230728,
      "step": 26365
    },
    {
      "epoch": 13.820754716981131,
      "grad_norm": 0.3501248061656952,
      "learning_rate": 1.3185344105261937e-05,
      "loss": 0.0259,
      "num_input_tokens_seen": 17233800,
      "step": 26370
    },
    {
      "epoch": 13.823375262054507,
      "grad_norm": 0.6161431074142456,
      "learning_rate": 1.3175268484059639e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 17236744,
      "step": 26375
    },
    {
      "epoch": 13.825995807127882,
      "grad_norm": 2.214280605316162,
      "learning_rate": 1.316519533645626e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 17241224,
      "step": 26380
    },
    {
      "epoch": 13.828616352201259,
      "grad_norm": 4.612048149108887,
      "learning_rate": 1.3155124664559004e-05,
      "loss": 0.0289,
      "num_input_tokens_seen": 17244808,
      "step": 26385
    },
    {
      "epoch": 13.831236897274634,
      "grad_norm": 2.459592342376709,
      "learning_rate": 1.3145056470474525e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 17248168,
      "step": 26390
    },
    {
      "epoch": 13.833857442348009,
      "grad_norm": 0.8272463083267212,
      "learning_rate": 1.313499075630899e-05,
      "loss": 0.0321,
      "num_input_tokens_seen": 17252136,
      "step": 26395
    },
    {
      "epoch": 13.836477987421384,
      "grad_norm": 3.007523536682129,
      "learning_rate": 1.312492752416802e-05,
      "loss": 0.0519,
      "num_input_tokens_seen": 17255880,
      "step": 26400
    },
    {
      "epoch": 13.83909853249476,
      "grad_norm": 2.5277156829833984,
      "learning_rate": 1.3114866776156748e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 17258728,
      "step": 26405
    },
    {
      "epoch": 13.841719077568134,
      "grad_norm": 7.791179180145264,
      "learning_rate": 1.3104808514379748e-05,
      "loss": 0.045,
      "num_input_tokens_seen": 17261992,
      "step": 26410
    },
    {
      "epoch": 13.84433962264151,
      "grad_norm": 1.5386936664581299,
      "learning_rate": 1.3094752740941094e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 17265896,
      "step": 26415
    },
    {
      "epoch": 13.846960167714885,
      "grad_norm": 2.159574270248413,
      "learning_rate": 1.3084699457944344e-05,
      "loss": 0.0514,
      "num_input_tokens_seen": 17269032,
      "step": 26420
    },
    {
      "epoch": 13.84958071278826,
      "grad_norm": 8.246274948120117,
      "learning_rate": 1.3074648667492528e-05,
      "loss": 0.051,
      "num_input_tokens_seen": 17272456,
      "step": 26425
    },
    {
      "epoch": 13.852201257861635,
      "grad_norm": 3.5440218448638916,
      "learning_rate": 1.3064600371688171e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 17275496,
      "step": 26430
    },
    {
      "epoch": 13.85482180293501,
      "grad_norm": 2.967829942703247,
      "learning_rate": 1.3054554572633249e-05,
      "loss": 0.0185,
      "num_input_tokens_seen": 17277672,
      "step": 26435
    },
    {
      "epoch": 13.857442348008385,
      "grad_norm": 0.06134989112615585,
      "learning_rate": 1.3044511272429222e-05,
      "loss": 0.0504,
      "num_input_tokens_seen": 17281768,
      "step": 26440
    },
    {
      "epoch": 13.86006289308176,
      "grad_norm": 0.8896481990814209,
      "learning_rate": 1.3034470473177038e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 17284616,
      "step": 26445
    },
    {
      "epoch": 13.862683438155136,
      "grad_norm": 1.991835355758667,
      "learning_rate": 1.3024432176977125e-05,
      "loss": 0.0541,
      "num_input_tokens_seen": 17287848,
      "step": 26450
    },
    {
      "epoch": 13.865303983228511,
      "grad_norm": 2.7909657955169678,
      "learning_rate": 1.3014396385929372e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 17294312,
      "step": 26455
    },
    {
      "epoch": 13.867924528301886,
      "grad_norm": 5.405220985412598,
      "learning_rate": 1.3004363102133165e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 17297672,
      "step": 26460
    },
    {
      "epoch": 13.870545073375261,
      "grad_norm": 2.443969249725342,
      "learning_rate": 1.2994332327687348e-05,
      "loss": 0.0291,
      "num_input_tokens_seen": 17301032,
      "step": 26465
    },
    {
      "epoch": 13.873165618448636,
      "grad_norm": 1.7090953588485718,
      "learning_rate": 1.2984304064690228e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 17304200,
      "step": 26470
    },
    {
      "epoch": 13.875786163522012,
      "grad_norm": 0.7316974401473999,
      "learning_rate": 1.2974278315239619e-05,
      "loss": 0.0285,
      "num_input_tokens_seen": 17307208,
      "step": 26475
    },
    {
      "epoch": 13.878406708595389,
      "grad_norm": 0.8648298382759094,
      "learning_rate": 1.2964255081432786e-05,
      "loss": 0.0214,
      "num_input_tokens_seen": 17310056,
      "step": 26480
    },
    {
      "epoch": 13.881027253668764,
      "grad_norm": 2.0818848609924316,
      "learning_rate": 1.2954234365366488e-05,
      "loss": 0.0413,
      "num_input_tokens_seen": 17313960,
      "step": 26485
    },
    {
      "epoch": 13.883647798742139,
      "grad_norm": 5.6104207038879395,
      "learning_rate": 1.2944216169136936e-05,
      "loss": 0.0229,
      "num_input_tokens_seen": 17316712,
      "step": 26490
    },
    {
      "epoch": 13.886268343815514,
      "grad_norm": 1.7537617683410645,
      "learning_rate": 1.2934200494839815e-05,
      "loss": 0.0299,
      "num_input_tokens_seen": 17321096,
      "step": 26495
    },
    {
      "epoch": 13.88888888888889,
      "grad_norm": 1.4065557718276978,
      "learning_rate": 1.2924187344570293e-05,
      "loss": 0.047,
      "num_input_tokens_seen": 17324168,
      "step": 26500
    },
    {
      "epoch": 13.891509433962264,
      "grad_norm": 1.2358378171920776,
      "learning_rate": 1.2914176720423011e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 17327016,
      "step": 26505
    },
    {
      "epoch": 13.89412997903564,
      "grad_norm": 0.5800265669822693,
      "learning_rate": 1.2904168624492088e-05,
      "loss": 0.0323,
      "num_input_tokens_seen": 17330088,
      "step": 26510
    },
    {
      "epoch": 13.896750524109015,
      "grad_norm": 0.5139709115028381,
      "learning_rate": 1.2894163058871092e-05,
      "loss": 0.0107,
      "num_input_tokens_seen": 17332328,
      "step": 26515
    },
    {
      "epoch": 13.89937106918239,
      "grad_norm": 9.134119987487793,
      "learning_rate": 1.2884160025653064e-05,
      "loss": 0.0346,
      "num_input_tokens_seen": 17336584,
      "step": 26520
    },
    {
      "epoch": 13.901991614255765,
      "grad_norm": 1.5468230247497559,
      "learning_rate": 1.2874159526930532e-05,
      "loss": 0.0358,
      "num_input_tokens_seen": 17339240,
      "step": 26525
    },
    {
      "epoch": 13.90461215932914,
      "grad_norm": 2.5092411041259766,
      "learning_rate": 1.286416156479549e-05,
      "loss": 0.0563,
      "num_input_tokens_seen": 17343688,
      "step": 26530
    },
    {
      "epoch": 13.907232704402515,
      "grad_norm": 1.3647631406784058,
      "learning_rate": 1.2854166141339394e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 17347336,
      "step": 26535
    },
    {
      "epoch": 13.90985324947589,
      "grad_norm": 0.7444062829017639,
      "learning_rate": 1.2844173258653183e-05,
      "loss": 0.0258,
      "num_input_tokens_seen": 17350952,
      "step": 26540
    },
    {
      "epoch": 13.912473794549266,
      "grad_norm": 1.1677268743515015,
      "learning_rate": 1.2834182918827247e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 17354376,
      "step": 26545
    },
    {
      "epoch": 13.915094339622641,
      "grad_norm": 0.31712865829467773,
      "learning_rate": 1.282419512395144e-05,
      "loss": 0.0373,
      "num_input_tokens_seen": 17358216,
      "step": 26550
    },
    {
      "epoch": 13.917714884696016,
      "grad_norm": 1.5809006690979004,
      "learning_rate": 1.2814209876115102e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 17361224,
      "step": 26555
    },
    {
      "epoch": 13.920335429769391,
      "grad_norm": 1.2272874116897583,
      "learning_rate": 1.2804227177407038e-05,
      "loss": 0.0474,
      "num_input_tokens_seen": 17364488,
      "step": 26560
    },
    {
      "epoch": 13.922955974842766,
      "grad_norm": 5.440149784088135,
      "learning_rate": 1.279424702991552e-05,
      "loss": 0.0576,
      "num_input_tokens_seen": 17368936,
      "step": 26565
    },
    {
      "epoch": 13.925576519916142,
      "grad_norm": 3.1510016918182373,
      "learning_rate": 1.2784269435728264e-05,
      "loss": 0.0433,
      "num_input_tokens_seen": 17371912,
      "step": 26570
    },
    {
      "epoch": 13.928197064989519,
      "grad_norm": 1.3853731155395508,
      "learning_rate": 1.2774294396932488e-05,
      "loss": 0.0252,
      "num_input_tokens_seen": 17374856,
      "step": 26575
    },
    {
      "epoch": 13.930817610062894,
      "grad_norm": 0.5390335321426392,
      "learning_rate": 1.2764321915614835e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 17377832,
      "step": 26580
    },
    {
      "epoch": 13.933438155136269,
      "grad_norm": 2.5663790702819824,
      "learning_rate": 1.2754351993861444e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 17381064,
      "step": 26585
    },
    {
      "epoch": 13.936058700209644,
      "grad_norm": 0.5206699371337891,
      "learning_rate": 1.2744384633757922e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 17384968,
      "step": 26590
    },
    {
      "epoch": 13.93867924528302,
      "grad_norm": 1.696744441986084,
      "learning_rate": 1.2734419837389306e-05,
      "loss": 0.046,
      "num_input_tokens_seen": 17388104,
      "step": 26595
    },
    {
      "epoch": 13.941299790356394,
      "grad_norm": 1.40324068069458,
      "learning_rate": 1.2724457606840123e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 17391880,
      "step": 26600
    },
    {
      "epoch": 13.94392033542977,
      "grad_norm": 0.6756210923194885,
      "learning_rate": 1.2714497944194376e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 17395240,
      "step": 26605
    },
    {
      "epoch": 13.946540880503145,
      "grad_norm": 3.538236379623413,
      "learning_rate": 1.2704540851535484e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 17400680,
      "step": 26610
    },
    {
      "epoch": 13.94916142557652,
      "grad_norm": 2.8540947437286377,
      "learning_rate": 1.2694586330946376e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 17403080,
      "step": 26615
    },
    {
      "epoch": 13.951781970649895,
      "grad_norm": 1.705734133720398,
      "learning_rate": 1.2684634384509427e-05,
      "loss": 0.0271,
      "num_input_tokens_seen": 17406600,
      "step": 26620
    },
    {
      "epoch": 13.95440251572327,
      "grad_norm": 1.1363980770111084,
      "learning_rate": 1.2674685014306456e-05,
      "loss": 0.0404,
      "num_input_tokens_seen": 17409320,
      "step": 26625
    },
    {
      "epoch": 13.957023060796645,
      "grad_norm": 13.079869270324707,
      "learning_rate": 1.2664738222418776e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 17412616,
      "step": 26630
    },
    {
      "epoch": 13.95964360587002,
      "grad_norm": 1.0042282342910767,
      "learning_rate": 1.2654794010927118e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 17415912,
      "step": 26635
    },
    {
      "epoch": 13.962264150943396,
      "grad_norm": 2.8437671661376953,
      "learning_rate": 1.2644852381911715e-05,
      "loss": 0.0223,
      "num_input_tokens_seen": 17419656,
      "step": 26640
    },
    {
      "epoch": 13.964884696016771,
      "grad_norm": 0.3214322328567505,
      "learning_rate": 1.2634913337452248e-05,
      "loss": 0.0257,
      "num_input_tokens_seen": 17421768,
      "step": 26645
    },
    {
      "epoch": 13.967505241090146,
      "grad_norm": 0.9472206830978394,
      "learning_rate": 1.2624976879627832e-05,
      "loss": 0.0116,
      "num_input_tokens_seen": 17424968,
      "step": 26650
    },
    {
      "epoch": 13.970125786163521,
      "grad_norm": 3.3674285411834717,
      "learning_rate": 1.2615043010517069e-05,
      "loss": 0.0307,
      "num_input_tokens_seen": 17427912,
      "step": 26655
    },
    {
      "epoch": 13.972746331236896,
      "grad_norm": 5.722716808319092,
      "learning_rate": 1.2605111732198027e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 17431272,
      "step": 26660
    },
    {
      "epoch": 13.975366876310272,
      "grad_norm": 0.6813350319862366,
      "learning_rate": 1.2595183046748188e-05,
      "loss": 0.0191,
      "num_input_tokens_seen": 17434728,
      "step": 26665
    },
    {
      "epoch": 13.977987421383649,
      "grad_norm": 0.30123892426490784,
      "learning_rate": 1.2585256956244545e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 17438216,
      "step": 26670
    },
    {
      "epoch": 13.980607966457024,
      "grad_norm": 11.657934188842773,
      "learning_rate": 1.2575333462763506e-05,
      "loss": 0.0617,
      "num_input_tokens_seen": 17441384,
      "step": 26675
    },
    {
      "epoch": 13.983228511530399,
      "grad_norm": 1.9204010963439941,
      "learning_rate": 1.2565412568380957e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 17444168,
      "step": 26680
    },
    {
      "epoch": 13.985849056603774,
      "grad_norm": 0.4415286183357239,
      "learning_rate": 1.2555494275172237e-05,
      "loss": 0.0398,
      "num_input_tokens_seen": 17448168,
      "step": 26685
    },
    {
      "epoch": 13.98846960167715,
      "grad_norm": 3.2816574573516846,
      "learning_rate": 1.2545578585212148e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 17450824,
      "step": 26690
    },
    {
      "epoch": 13.991090146750524,
      "grad_norm": 1.6157660484313965,
      "learning_rate": 1.2535665500574922e-05,
      "loss": 0.0218,
      "num_input_tokens_seen": 17453832,
      "step": 26695
    },
    {
      "epoch": 13.9937106918239,
      "grad_norm": 0.6565406918525696,
      "learning_rate": 1.2525755023334285e-05,
      "loss": 0.0283,
      "num_input_tokens_seen": 17457704,
      "step": 26700
    },
    {
      "epoch": 13.996331236897275,
      "grad_norm": 4.986877918243408,
      "learning_rate": 1.251584715556337e-05,
      "loss": 0.0437,
      "num_input_tokens_seen": 17460680,
      "step": 26705
    },
    {
      "epoch": 13.99895178197065,
      "grad_norm": 3.326524019241333,
      "learning_rate": 1.2505941899334805e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 17464072,
      "step": 26710
    },
    {
      "epoch": 14.0,
      "eval_loss": 1.2786608934402466,
      "eval_runtime": 11.7919,
      "eval_samples_per_second": 71.914,
      "eval_steps_per_second": 17.978,
      "num_input_tokens_seen": 17464792,
      "step": 26712
    },
    {
      "epoch": 14.001572327044025,
      "grad_norm": 0.2785208225250244,
      "learning_rate": 1.2496039256720665e-05,
      "loss": 0.0145,
      "num_input_tokens_seen": 17466744,
      "step": 26715
    },
    {
      "epoch": 14.0041928721174,
      "grad_norm": 1.7697343826293945,
      "learning_rate": 1.2486139229792452e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 17470072,
      "step": 26720
    },
    {
      "epoch": 14.006813417190775,
      "grad_norm": 0.33831706643104553,
      "learning_rate": 1.2476241820621152e-05,
      "loss": 0.0078,
      "num_input_tokens_seen": 17473432,
      "step": 26725
    },
    {
      "epoch": 14.00943396226415,
      "grad_norm": 0.7265195250511169,
      "learning_rate": 1.2466347031277176e-05,
      "loss": 0.0195,
      "num_input_tokens_seen": 17476568,
      "step": 26730
    },
    {
      "epoch": 14.012054507337526,
      "grad_norm": 0.4536689221858978,
      "learning_rate": 1.2456454863830411e-05,
      "loss": 0.0097,
      "num_input_tokens_seen": 17479608,
      "step": 26735
    },
    {
      "epoch": 14.014675052410901,
      "grad_norm": 3.355975389480591,
      "learning_rate": 1.2446565320350185e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 17482904,
      "step": 26740
    },
    {
      "epoch": 14.017295597484276,
      "grad_norm": 0.8056796193122864,
      "learning_rate": 1.2436678402905283e-05,
      "loss": 0.0071,
      "num_input_tokens_seen": 17485432,
      "step": 26745
    },
    {
      "epoch": 14.019916142557651,
      "grad_norm": 0.6640539765357971,
      "learning_rate": 1.2426794113563933e-05,
      "loss": 0.0074,
      "num_input_tokens_seen": 17487864,
      "step": 26750
    },
    {
      "epoch": 14.022536687631026,
      "grad_norm": 1.9270209074020386,
      "learning_rate": 1.24169124543938e-05,
      "loss": 0.0378,
      "num_input_tokens_seen": 17490424,
      "step": 26755
    },
    {
      "epoch": 14.025157232704403,
      "grad_norm": 0.4714427590370178,
      "learning_rate": 1.2407033427462026e-05,
      "loss": 0.006,
      "num_input_tokens_seen": 17492888,
      "step": 26760
    },
    {
      "epoch": 14.027777777777779,
      "grad_norm": 1.1537845134735107,
      "learning_rate": 1.2397157034835192e-05,
      "loss": 0.0141,
      "num_input_tokens_seen": 17495448,
      "step": 26765
    },
    {
      "epoch": 14.030398322851154,
      "grad_norm": 0.20526573061943054,
      "learning_rate": 1.2387283278579325e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 17499160,
      "step": 26770
    },
    {
      "epoch": 14.033018867924529,
      "grad_norm": 1.5535504817962646,
      "learning_rate": 1.237741216075991e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 17501880,
      "step": 26775
    },
    {
      "epoch": 14.035639412997904,
      "grad_norm": 9.188496589660645,
      "learning_rate": 1.2367543683441868e-05,
      "loss": 0.038,
      "num_input_tokens_seen": 17504696,
      "step": 26780
    },
    {
      "epoch": 14.03825995807128,
      "grad_norm": 1.9759305715560913,
      "learning_rate": 1.2357677848689558e-05,
      "loss": 0.0233,
      "num_input_tokens_seen": 17507672,
      "step": 26785
    },
    {
      "epoch": 14.040880503144654,
      "grad_norm": 2.361889123916626,
      "learning_rate": 1.2347814658566811e-05,
      "loss": 0.0112,
      "num_input_tokens_seen": 17510296,
      "step": 26790
    },
    {
      "epoch": 14.04350104821803,
      "grad_norm": 0.5493770241737366,
      "learning_rate": 1.2337954115136894e-05,
      "loss": 0.0274,
      "num_input_tokens_seen": 17512920,
      "step": 26795
    },
    {
      "epoch": 14.046121593291405,
      "grad_norm": 0.8946231007575989,
      "learning_rate": 1.2328096220462529e-05,
      "loss": 0.0212,
      "num_input_tokens_seen": 17516888,
      "step": 26800
    },
    {
      "epoch": 14.04874213836478,
      "grad_norm": 1.7697114944458008,
      "learning_rate": 1.2318240976605865e-05,
      "loss": 0.0215,
      "num_input_tokens_seen": 17520792,
      "step": 26805
    },
    {
      "epoch": 14.051362683438155,
      "grad_norm": 7.6079421043396,
      "learning_rate": 1.23083883856285e-05,
      "loss": 0.0163,
      "num_input_tokens_seen": 17523896,
      "step": 26810
    },
    {
      "epoch": 14.05398322851153,
      "grad_norm": 2.0132923126220703,
      "learning_rate": 1.229853844959149e-05,
      "loss": 0.0222,
      "num_input_tokens_seen": 17527352,
      "step": 26815
    },
    {
      "epoch": 14.056603773584905,
      "grad_norm": 0.4468654692173004,
      "learning_rate": 1.228869117055533e-05,
      "loss": 0.0059,
      "num_input_tokens_seen": 17530872,
      "step": 26820
    },
    {
      "epoch": 14.05922431865828,
      "grad_norm": 0.8625711798667908,
      "learning_rate": 1.227884655057997e-05,
      "loss": 0.0368,
      "num_input_tokens_seen": 17533528,
      "step": 26825
    },
    {
      "epoch": 14.061844863731656,
      "grad_norm": 1.753871202468872,
      "learning_rate": 1.2269004591724777e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 17537368,
      "step": 26830
    },
    {
      "epoch": 14.064465408805031,
      "grad_norm": 0.73787921667099,
      "learning_rate": 1.2259165296048575e-05,
      "loss": 0.0106,
      "num_input_tokens_seen": 17541080,
      "step": 26835
    },
    {
      "epoch": 14.067085953878406,
      "grad_norm": 2.258514642715454,
      "learning_rate": 1.2249328665609638e-05,
      "loss": 0.0216,
      "num_input_tokens_seen": 17545112,
      "step": 26840
    },
    {
      "epoch": 14.069706498951781,
      "grad_norm": 1.5859088897705078,
      "learning_rate": 1.2239494702465676e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 17547960,
      "step": 26845
    },
    {
      "epoch": 14.072327044025156,
      "grad_norm": 0.8688575029373169,
      "learning_rate": 1.2229663408673842e-05,
      "loss": 0.0152,
      "num_input_tokens_seen": 17551416,
      "step": 26850
    },
    {
      "epoch": 14.074947589098532,
      "grad_norm": 2.9731035232543945,
      "learning_rate": 1.221983478629074e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 17555096,
      "step": 26855
    },
    {
      "epoch": 14.077568134171909,
      "grad_norm": 1.5972825288772583,
      "learning_rate": 1.22100088373724e-05,
      "loss": 0.0363,
      "num_input_tokens_seen": 17559448,
      "step": 26860
    },
    {
      "epoch": 14.080188679245284,
      "grad_norm": 1.2208991050720215,
      "learning_rate": 1.2200185563974283e-05,
      "loss": 0.0187,
      "num_input_tokens_seen": 17562584,
      "step": 26865
    },
    {
      "epoch": 14.082809224318659,
      "grad_norm": 1.3942922353744507,
      "learning_rate": 1.219036496815132e-05,
      "loss": 0.0108,
      "num_input_tokens_seen": 17566776,
      "step": 26870
    },
    {
      "epoch": 14.085429769392034,
      "grad_norm": 1.7994723320007324,
      "learning_rate": 1.2180547051957863e-05,
      "loss": 0.0233,
      "num_input_tokens_seen": 17569368,
      "step": 26875
    },
    {
      "epoch": 14.08805031446541,
      "grad_norm": 1.4795646667480469,
      "learning_rate": 1.2170731817447723e-05,
      "loss": 0.0124,
      "num_input_tokens_seen": 17571928,
      "step": 26880
    },
    {
      "epoch": 14.090670859538784,
      "grad_norm": 0.9582589864730835,
      "learning_rate": 1.216091926667412e-05,
      "loss": 0.0097,
      "num_input_tokens_seen": 17574392,
      "step": 26885
    },
    {
      "epoch": 14.09329140461216,
      "grad_norm": 5.734708786010742,
      "learning_rate": 1.2151109401689723e-05,
      "loss": 0.0386,
      "num_input_tokens_seen": 17578104,
      "step": 26890
    },
    {
      "epoch": 14.095911949685535,
      "grad_norm": 0.4100213944911957,
      "learning_rate": 1.214130222454665e-05,
      "loss": 0.0116,
      "num_input_tokens_seen": 17581368,
      "step": 26895
    },
    {
      "epoch": 14.09853249475891,
      "grad_norm": 2.492410659790039,
      "learning_rate": 1.2131497737296452e-05,
      "loss": 0.0103,
      "num_input_tokens_seen": 17584024,
      "step": 26900
    },
    {
      "epoch": 14.101153039832285,
      "grad_norm": 0.4432511329650879,
      "learning_rate": 1.2121695941990124e-05,
      "loss": 0.0068,
      "num_input_tokens_seen": 17586168,
      "step": 26905
    },
    {
      "epoch": 14.10377358490566,
      "grad_norm": 2.5933468341827393,
      "learning_rate": 1.2111896840678075e-05,
      "loss": 0.0203,
      "num_input_tokens_seen": 17591096,
      "step": 26910
    },
    {
      "epoch": 14.106394129979035,
      "grad_norm": 5.076572895050049,
      "learning_rate": 1.210210043541018e-05,
      "loss": 0.0191,
      "num_input_tokens_seen": 17594264,
      "step": 26915
    },
    {
      "epoch": 14.10901467505241,
      "grad_norm": 0.5464150905609131,
      "learning_rate": 1.2092306728235717e-05,
      "loss": 0.0124,
      "num_input_tokens_seen": 17597944,
      "step": 26920
    },
    {
      "epoch": 14.111635220125786,
      "grad_norm": 1.5530449151992798,
      "learning_rate": 1.2082515721203427e-05,
      "loss": 0.0188,
      "num_input_tokens_seen": 17600728,
      "step": 26925
    },
    {
      "epoch": 14.114255765199161,
      "grad_norm": 1.5906902551651,
      "learning_rate": 1.2072727416361479e-05,
      "loss": 0.0245,
      "num_input_tokens_seen": 17603576,
      "step": 26930
    },
    {
      "epoch": 14.116876310272536,
      "grad_norm": 2.780156373977661,
      "learning_rate": 1.2062941815757484e-05,
      "loss": 0.0198,
      "num_input_tokens_seen": 17606392,
      "step": 26935
    },
    {
      "epoch": 14.119496855345911,
      "grad_norm": 0.9692513346672058,
      "learning_rate": 1.2053158921438454e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 17609944,
      "step": 26940
    },
    {
      "epoch": 14.122117400419286,
      "grad_norm": 0.6561049222946167,
      "learning_rate": 1.2043378735450885e-05,
      "loss": 0.0464,
      "num_input_tokens_seen": 17614424,
      "step": 26945
    },
    {
      "epoch": 14.124737945492662,
      "grad_norm": 5.306742191314697,
      "learning_rate": 1.2033601259840657e-05,
      "loss": 0.0174,
      "num_input_tokens_seen": 17617368,
      "step": 26950
    },
    {
      "epoch": 14.127358490566039,
      "grad_norm": 8.193394660949707,
      "learning_rate": 1.2023826496653118e-05,
      "loss": 0.0313,
      "num_input_tokens_seen": 17620792,
      "step": 26955
    },
    {
      "epoch": 14.129979035639414,
      "grad_norm": 1.7352972030639648,
      "learning_rate": 1.2014054447933044e-05,
      "loss": 0.016,
      "num_input_tokens_seen": 17624152,
      "step": 26960
    },
    {
      "epoch": 14.132599580712789,
      "grad_norm": 0.9178063869476318,
      "learning_rate": 1.200428511572462e-05,
      "loss": 0.0066,
      "num_input_tokens_seen": 17627064,
      "step": 26965
    },
    {
      "epoch": 14.135220125786164,
      "grad_norm": 0.6664779186248779,
      "learning_rate": 1.1994518502071484e-05,
      "loss": 0.0377,
      "num_input_tokens_seen": 17629720,
      "step": 26970
    },
    {
      "epoch": 14.13784067085954,
      "grad_norm": 2.5016448497772217,
      "learning_rate": 1.1984754609016716e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 17633048,
      "step": 26975
    },
    {
      "epoch": 14.140461215932914,
      "grad_norm": 0.819560170173645,
      "learning_rate": 1.1974993438602785e-05,
      "loss": 0.0698,
      "num_input_tokens_seen": 17636472,
      "step": 26980
    },
    {
      "epoch": 14.14308176100629,
      "grad_norm": 3.389216184616089,
      "learning_rate": 1.1965234992871641e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 17640120,
      "step": 26985
    },
    {
      "epoch": 14.145702306079665,
      "grad_norm": 1.4039415121078491,
      "learning_rate": 1.195547927386462e-05,
      "loss": 0.0522,
      "num_input_tokens_seen": 17643704,
      "step": 26990
    },
    {
      "epoch": 14.14832285115304,
      "grad_norm": 2.395064353942871,
      "learning_rate": 1.1945726283622514e-05,
      "loss": 0.0463,
      "num_input_tokens_seen": 17646968,
      "step": 26995
    },
    {
      "epoch": 14.150943396226415,
      "grad_norm": 0.21547746658325195,
      "learning_rate": 1.1935976024185547e-05,
      "loss": 0.013,
      "num_input_tokens_seen": 17650552,
      "step": 27000
    },
    {
      "epoch": 14.15356394129979,
      "grad_norm": 4.101755619049072,
      "learning_rate": 1.1926228497593341e-05,
      "loss": 0.0314,
      "num_input_tokens_seen": 17653528,
      "step": 27005
    },
    {
      "epoch": 14.156184486373165,
      "grad_norm": 2.535196304321289,
      "learning_rate": 1.1916483705884985e-05,
      "loss": 0.0174,
      "num_input_tokens_seen": 17656536,
      "step": 27010
    },
    {
      "epoch": 14.15880503144654,
      "grad_norm": 0.6196778416633606,
      "learning_rate": 1.1906741651098983e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 17662200,
      "step": 27015
    },
    {
      "epoch": 14.161425576519916,
      "grad_norm": 0.3725185692310333,
      "learning_rate": 1.1897002335273239e-05,
      "loss": 0.0254,
      "num_input_tokens_seen": 17665592,
      "step": 27020
    },
    {
      "epoch": 14.164046121593291,
      "grad_norm": 1.6537138223648071,
      "learning_rate": 1.1887265760445121e-05,
      "loss": 0.0158,
      "num_input_tokens_seen": 17668376,
      "step": 27025
    },
    {
      "epoch": 14.166666666666666,
      "grad_norm": 0.48560309410095215,
      "learning_rate": 1.1877531928651417e-05,
      "loss": 0.0151,
      "num_input_tokens_seen": 17672760,
      "step": 27030
    },
    {
      "epoch": 14.169287211740041,
      "grad_norm": 0.8256524205207825,
      "learning_rate": 1.1867800841928313e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 17675480,
      "step": 27035
    },
    {
      "epoch": 14.171907756813416,
      "grad_norm": 1.8114168643951416,
      "learning_rate": 1.1858072502311465e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 17678360,
      "step": 27040
    },
    {
      "epoch": 14.174528301886792,
      "grad_norm": 0.7373493909835815,
      "learning_rate": 1.1848346911835908e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 17682040,
      "step": 27045
    },
    {
      "epoch": 14.177148846960169,
      "grad_norm": 0.8281897902488708,
      "learning_rate": 1.1838624072536137e-05,
      "loss": 0.0303,
      "num_input_tokens_seen": 17685432,
      "step": 27050
    },
    {
      "epoch": 14.179769392033544,
      "grad_norm": 1.2451562881469727,
      "learning_rate": 1.1828903986446055e-05,
      "loss": 0.0114,
      "num_input_tokens_seen": 17688728,
      "step": 27055
    },
    {
      "epoch": 14.182389937106919,
      "grad_norm": 0.32127201557159424,
      "learning_rate": 1.1819186655599007e-05,
      "loss": 0.0471,
      "num_input_tokens_seen": 17692216,
      "step": 27060
    },
    {
      "epoch": 14.185010482180294,
      "grad_norm": 0.5618938207626343,
      "learning_rate": 1.1809472082027734e-05,
      "loss": 0.0082,
      "num_input_tokens_seen": 17696184,
      "step": 27065
    },
    {
      "epoch": 14.18763102725367,
      "grad_norm": 8.589691162109375,
      "learning_rate": 1.1799760267764407e-05,
      "loss": 0.0115,
      "num_input_tokens_seen": 17698808,
      "step": 27070
    },
    {
      "epoch": 14.190251572327044,
      "grad_norm": 0.4465773105621338,
      "learning_rate": 1.1790051214840634e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 17702648,
      "step": 27075
    },
    {
      "epoch": 14.19287211740042,
      "grad_norm": 0.532798707485199,
      "learning_rate": 1.1780344925287443e-05,
      "loss": 0.0135,
      "num_input_tokens_seen": 17705464,
      "step": 27080
    },
    {
      "epoch": 14.195492662473795,
      "grad_norm": 1.7082411050796509,
      "learning_rate": 1.1770641401135282e-05,
      "loss": 0.0141,
      "num_input_tokens_seen": 17707864,
      "step": 27085
    },
    {
      "epoch": 14.19811320754717,
      "grad_norm": 2.786546468734741,
      "learning_rate": 1.1760940644414003e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 17711736,
      "step": 27090
    },
    {
      "epoch": 14.200733752620545,
      "grad_norm": 0.7103698253631592,
      "learning_rate": 1.1751242657152914e-05,
      "loss": 0.0084,
      "num_input_tokens_seen": 17715992,
      "step": 27095
    },
    {
      "epoch": 14.20335429769392,
      "grad_norm": 0.17147362232208252,
      "learning_rate": 1.1741547441380698e-05,
      "loss": 0.0071,
      "num_input_tokens_seen": 17718616,
      "step": 27100
    },
    {
      "epoch": 14.205974842767295,
      "grad_norm": 4.713693141937256,
      "learning_rate": 1.17318549991255e-05,
      "loss": 0.0238,
      "num_input_tokens_seen": 17721880,
      "step": 27105
    },
    {
      "epoch": 14.20859538784067,
      "grad_norm": 0.508185625076294,
      "learning_rate": 1.1722165332414863e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 17724376,
      "step": 27110
    },
    {
      "epoch": 14.211215932914046,
      "grad_norm": 3.6486892700195312,
      "learning_rate": 1.1712478443275768e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 17727448,
      "step": 27115
    },
    {
      "epoch": 14.213836477987421,
      "grad_norm": 1.3935470581054688,
      "learning_rate": 1.170279433373459e-05,
      "loss": 0.0152,
      "num_input_tokens_seen": 17730808,
      "step": 27120
    },
    {
      "epoch": 14.216457023060796,
      "grad_norm": 1.6732544898986816,
      "learning_rate": 1.1693113005817125e-05,
      "loss": 0.0197,
      "num_input_tokens_seen": 17733464,
      "step": 27125
    },
    {
      "epoch": 14.219077568134171,
      "grad_norm": 3.0496232509613037,
      "learning_rate": 1.1683434461548606e-05,
      "loss": 0.0136,
      "num_input_tokens_seen": 17736600,
      "step": 27130
    },
    {
      "epoch": 14.221698113207546,
      "grad_norm": 1.5926767587661743,
      "learning_rate": 1.1673758702953675e-05,
      "loss": 0.024,
      "num_input_tokens_seen": 17739608,
      "step": 27135
    },
    {
      "epoch": 14.224318658280922,
      "grad_norm": 2.790508270263672,
      "learning_rate": 1.1664085732056398e-05,
      "loss": 0.0117,
      "num_input_tokens_seen": 17742840,
      "step": 27140
    },
    {
      "epoch": 14.226939203354299,
      "grad_norm": 4.8341875076293945,
      "learning_rate": 1.1654415550880243e-05,
      "loss": 0.0151,
      "num_input_tokens_seen": 17744824,
      "step": 27145
    },
    {
      "epoch": 14.229559748427674,
      "grad_norm": 4.5698652267456055,
      "learning_rate": 1.1644748161448088e-05,
      "loss": 0.0104,
      "num_input_tokens_seen": 17747544,
      "step": 27150
    },
    {
      "epoch": 14.232180293501049,
      "grad_norm": 0.15237492322921753,
      "learning_rate": 1.1635083565782257e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 17750840,
      "step": 27155
    },
    {
      "epoch": 14.234800838574424,
      "grad_norm": 6.612332820892334,
      "learning_rate": 1.1625421765904465e-05,
      "loss": 0.0844,
      "num_input_tokens_seen": 17753432,
      "step": 27160
    },
    {
      "epoch": 14.2374213836478,
      "grad_norm": 4.070837020874023,
      "learning_rate": 1.1615762763835858e-05,
      "loss": 0.0274,
      "num_input_tokens_seen": 17756536,
      "step": 27165
    },
    {
      "epoch": 14.240041928721174,
      "grad_norm": 1.6768549680709839,
      "learning_rate": 1.1606106561596993e-05,
      "loss": 0.0168,
      "num_input_tokens_seen": 17759896,
      "step": 27170
    },
    {
      "epoch": 14.24266247379455,
      "grad_norm": 3.2090625762939453,
      "learning_rate": 1.1596453161207829e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 17763352,
      "step": 27175
    },
    {
      "epoch": 14.245283018867925,
      "grad_norm": 0.5625814199447632,
      "learning_rate": 1.1586802564687741e-05,
      "loss": 0.0108,
      "num_input_tokens_seen": 17765816,
      "step": 27180
    },
    {
      "epoch": 14.2479035639413,
      "grad_norm": 0.8665987253189087,
      "learning_rate": 1.1577154774055528e-05,
      "loss": 0.0062,
      "num_input_tokens_seen": 17768056,
      "step": 27185
    },
    {
      "epoch": 14.250524109014675,
      "grad_norm": 0.8021190166473389,
      "learning_rate": 1.1567509791329401e-05,
      "loss": 0.0113,
      "num_input_tokens_seen": 17771896,
      "step": 27190
    },
    {
      "epoch": 14.25314465408805,
      "grad_norm": 0.4399612843990326,
      "learning_rate": 1.155786761852699e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 17775096,
      "step": 27195
    },
    {
      "epoch": 14.255765199161425,
      "grad_norm": 1.6568918228149414,
      "learning_rate": 1.1548228257665317e-05,
      "loss": 0.014,
      "num_input_tokens_seen": 17778392,
      "step": 27200
    },
    {
      "epoch": 14.2583857442348,
      "grad_norm": 2.014737606048584,
      "learning_rate": 1.1538591710760815e-05,
      "loss": 0.0087,
      "num_input_tokens_seen": 17781176,
      "step": 27205
    },
    {
      "epoch": 14.261006289308176,
      "grad_norm": 1.2317862510681152,
      "learning_rate": 1.1528957979829355e-05,
      "loss": 0.0727,
      "num_input_tokens_seen": 17785272,
      "step": 27210
    },
    {
      "epoch": 14.26362683438155,
      "grad_norm": 0.9397438764572144,
      "learning_rate": 1.1519327066886198e-05,
      "loss": 0.0361,
      "num_input_tokens_seen": 17788280,
      "step": 27215
    },
    {
      "epoch": 14.266247379454926,
      "grad_norm": 2.6926565170288086,
      "learning_rate": 1.1509698973946032e-05,
      "loss": 0.0178,
      "num_input_tokens_seen": 17791064,
      "step": 27220
    },
    {
      "epoch": 14.268867924528301,
      "grad_norm": 1.4773449897766113,
      "learning_rate": 1.1500073703022926e-05,
      "loss": 0.0066,
      "num_input_tokens_seen": 17793656,
      "step": 27225
    },
    {
      "epoch": 14.271488469601676,
      "grad_norm": 0.5150721669197083,
      "learning_rate": 1.1490451256130394e-05,
      "loss": 0.0134,
      "num_input_tokens_seen": 17796536,
      "step": 27230
    },
    {
      "epoch": 14.274109014675052,
      "grad_norm": 2.9816782474517822,
      "learning_rate": 1.1480831635281328e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 17801304,
      "step": 27235
    },
    {
      "epoch": 14.276729559748428,
      "grad_norm": 3.5818428993225098,
      "learning_rate": 1.1471214842488046e-05,
      "loss": 0.0304,
      "num_input_tokens_seen": 17804472,
      "step": 27240
    },
    {
      "epoch": 14.279350104821804,
      "grad_norm": 2.3372700214385986,
      "learning_rate": 1.1461600879762277e-05,
      "loss": 0.027,
      "num_input_tokens_seen": 17807928,
      "step": 27245
    },
    {
      "epoch": 14.281970649895179,
      "grad_norm": 1.231039047241211,
      "learning_rate": 1.1451989749115157e-05,
      "loss": 0.0269,
      "num_input_tokens_seen": 17810680,
      "step": 27250
    },
    {
      "epoch": 14.284591194968554,
      "grad_norm": 1.1937040090560913,
      "learning_rate": 1.1442381452557219e-05,
      "loss": 0.0132,
      "num_input_tokens_seen": 17814008,
      "step": 27255
    },
    {
      "epoch": 14.28721174004193,
      "grad_norm": 16.548158645629883,
      "learning_rate": 1.1432775992098397e-05,
      "loss": 0.0133,
      "num_input_tokens_seen": 17816440,
      "step": 27260
    },
    {
      "epoch": 14.289832285115304,
      "grad_norm": 1.0088536739349365,
      "learning_rate": 1.1423173369748056e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 17819256,
      "step": 27265
    },
    {
      "epoch": 14.29245283018868,
      "grad_norm": 1.2743010520935059,
      "learning_rate": 1.1413573587514952e-05,
      "loss": 0.0193,
      "num_input_tokens_seen": 17822232,
      "step": 27270
    },
    {
      "epoch": 14.295073375262055,
      "grad_norm": 4.226559638977051,
      "learning_rate": 1.140397664740726e-05,
      "loss": 0.0172,
      "num_input_tokens_seen": 17825432,
      "step": 27275
    },
    {
      "epoch": 14.29769392033543,
      "grad_norm": 1.0903308391571045,
      "learning_rate": 1.1394382551432534e-05,
      "loss": 0.0502,
      "num_input_tokens_seen": 17828952,
      "step": 27280
    },
    {
      "epoch": 14.300314465408805,
      "grad_norm": 1.800482153892517,
      "learning_rate": 1.1384791301597767e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 17831672,
      "step": 27285
    },
    {
      "epoch": 14.30293501048218,
      "grad_norm": 1.2506507635116577,
      "learning_rate": 1.1375202899909317e-05,
      "loss": 0.033,
      "num_input_tokens_seen": 17835384,
      "step": 27290
    },
    {
      "epoch": 14.305555555555555,
      "grad_norm": 1.2336983680725098,
      "learning_rate": 1.1365617348372984e-05,
      "loss": 0.0272,
      "num_input_tokens_seen": 17838552,
      "step": 27295
    },
    {
      "epoch": 14.30817610062893,
      "grad_norm": 0.6500644087791443,
      "learning_rate": 1.135603464899396e-05,
      "loss": 0.0221,
      "num_input_tokens_seen": 17841080,
      "step": 27300
    },
    {
      "epoch": 14.310796645702306,
      "grad_norm": 1.5088804960250854,
      "learning_rate": 1.1346454803776821e-05,
      "loss": 0.0355,
      "num_input_tokens_seen": 17847224,
      "step": 27305
    },
    {
      "epoch": 14.31341719077568,
      "grad_norm": 1.201853632926941,
      "learning_rate": 1.1336877814725566e-05,
      "loss": 0.016,
      "num_input_tokens_seen": 17850840,
      "step": 27310
    },
    {
      "epoch": 14.316037735849056,
      "grad_norm": 5.4351420402526855,
      "learning_rate": 1.1327303683843607e-05,
      "loss": 0.0333,
      "num_input_tokens_seen": 17853176,
      "step": 27315
    },
    {
      "epoch": 14.318658280922431,
      "grad_norm": 1.823990821838379,
      "learning_rate": 1.1317732413133723e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 17856280,
      "step": 27320
    },
    {
      "epoch": 14.321278825995806,
      "grad_norm": 1.6396938562393188,
      "learning_rate": 1.1308164004598118e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 17859064,
      "step": 27325
    },
    {
      "epoch": 14.323899371069182,
      "grad_norm": 7.447174549102783,
      "learning_rate": 1.129859846023841e-05,
      "loss": 0.0479,
      "num_input_tokens_seen": 17861880,
      "step": 27330
    },
    {
      "epoch": 14.326519916142558,
      "grad_norm": 2.737119197845459,
      "learning_rate": 1.1289035782055582e-05,
      "loss": 0.0693,
      "num_input_tokens_seen": 17868120,
      "step": 27335
    },
    {
      "epoch": 14.329140461215934,
      "grad_norm": 0.5365535020828247,
      "learning_rate": 1.1279475972050044e-05,
      "loss": 0.1162,
      "num_input_tokens_seen": 17871896,
      "step": 27340
    },
    {
      "epoch": 14.331761006289309,
      "grad_norm": 1.1316657066345215,
      "learning_rate": 1.1269919032221612e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 17875512,
      "step": 27345
    },
    {
      "epoch": 14.334381551362684,
      "grad_norm": 1.6757792234420776,
      "learning_rate": 1.1260364964569466e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 17879000,
      "step": 27350
    },
    {
      "epoch": 14.33700209643606,
      "grad_norm": 0.7734015583992004,
      "learning_rate": 1.125081377109223e-05,
      "loss": 0.0164,
      "num_input_tokens_seen": 17881368,
      "step": 27355
    },
    {
      "epoch": 14.339622641509434,
      "grad_norm": 4.158113479614258,
      "learning_rate": 1.1241265453787882e-05,
      "loss": 0.0756,
      "num_input_tokens_seen": 17884408,
      "step": 27360
    },
    {
      "epoch": 14.34224318658281,
      "grad_norm": 1.3230953216552734,
      "learning_rate": 1.1231720014653837e-05,
      "loss": 0.025,
      "num_input_tokens_seen": 17887224,
      "step": 27365
    },
    {
      "epoch": 14.344863731656185,
      "grad_norm": 11.52800464630127,
      "learning_rate": 1.1222177455686896e-05,
      "loss": 0.014,
      "num_input_tokens_seen": 17889720,
      "step": 27370
    },
    {
      "epoch": 14.34748427672956,
      "grad_norm": 1.2765685319900513,
      "learning_rate": 1.1212637778883237e-05,
      "loss": 0.0101,
      "num_input_tokens_seen": 17892888,
      "step": 27375
    },
    {
      "epoch": 14.350104821802935,
      "grad_norm": 3.4244301319122314,
      "learning_rate": 1.1203100986238468e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 17896056,
      "step": 27380
    },
    {
      "epoch": 14.35272536687631,
      "grad_norm": 0.7022636532783508,
      "learning_rate": 1.1193567079747561e-05,
      "loss": 0.0067,
      "num_input_tokens_seen": 17898840,
      "step": 27385
    },
    {
      "epoch": 14.355345911949685,
      "grad_norm": 2.749631881713867,
      "learning_rate": 1.1184036061404907e-05,
      "loss": 0.0161,
      "num_input_tokens_seen": 17901560,
      "step": 27390
    },
    {
      "epoch": 14.35796645702306,
      "grad_norm": 1.6101574897766113,
      "learning_rate": 1.117450793320429e-05,
      "loss": 0.0126,
      "num_input_tokens_seen": 17905432,
      "step": 27395
    },
    {
      "epoch": 14.360587002096436,
      "grad_norm": 0.13372193276882172,
      "learning_rate": 1.1164982697138893e-05,
      "loss": 0.0615,
      "num_input_tokens_seen": 17908088,
      "step": 27400
    },
    {
      "epoch": 14.36320754716981,
      "grad_norm": 0.17856325209140778,
      "learning_rate": 1.1155460355201267e-05,
      "loss": 0.0167,
      "num_input_tokens_seen": 17910616,
      "step": 27405
    },
    {
      "epoch": 14.365828092243186,
      "grad_norm": 1.3321821689605713,
      "learning_rate": 1.1145940909383399e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 17913400,
      "step": 27410
    },
    {
      "epoch": 14.368448637316561,
      "grad_norm": 3.6059417724609375,
      "learning_rate": 1.1136424361676626e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 17916600,
      "step": 27415
    },
    {
      "epoch": 14.371069182389936,
      "grad_norm": 0.9277543425559998,
      "learning_rate": 1.1126910714071711e-05,
      "loss": 0.0225,
      "num_input_tokens_seen": 17919160,
      "step": 27420
    },
    {
      "epoch": 14.373689727463312,
      "grad_norm": 1.561408281326294,
      "learning_rate": 1.1117399968558806e-05,
      "loss": 0.0173,
      "num_input_tokens_seen": 17921912,
      "step": 27425
    },
    {
      "epoch": 14.376310272536688,
      "grad_norm": 2.330360174179077,
      "learning_rate": 1.1107892127127454e-05,
      "loss": 0.0251,
      "num_input_tokens_seen": 17926040,
      "step": 27430
    },
    {
      "epoch": 14.378930817610064,
      "grad_norm": 1.0686217546463013,
      "learning_rate": 1.109838719176658e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 17929656,
      "step": 27435
    },
    {
      "epoch": 14.381551362683439,
      "grad_norm": 1.0752770900726318,
      "learning_rate": 1.1088885164464497e-05,
      "loss": 0.0126,
      "num_input_tokens_seen": 17935704,
      "step": 27440
    },
    {
      "epoch": 14.384171907756814,
      "grad_norm": 0.669145941734314,
      "learning_rate": 1.1079386047208932e-05,
      "loss": 0.0101,
      "num_input_tokens_seen": 17938328,
      "step": 27445
    },
    {
      "epoch": 14.38679245283019,
      "grad_norm": 0.17508019506931305,
      "learning_rate": 1.1069889841986991e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 17941240,
      "step": 27450
    },
    {
      "epoch": 14.389412997903564,
      "grad_norm": 2.760629653930664,
      "learning_rate": 1.1060396550785182e-05,
      "loss": 0.0348,
      "num_input_tokens_seen": 17944728,
      "step": 27455
    },
    {
      "epoch": 14.39203354297694,
      "grad_norm": 3.027651309967041,
      "learning_rate": 1.1050906175589382e-05,
      "loss": 0.0207,
      "num_input_tokens_seen": 17948536,
      "step": 27460
    },
    {
      "epoch": 14.394654088050315,
      "grad_norm": 9.902047157287598,
      "learning_rate": 1.1041418718384863e-05,
      "loss": 0.0181,
      "num_input_tokens_seen": 17951448,
      "step": 27465
    },
    {
      "epoch": 14.39727463312369,
      "grad_norm": 1.1337387561798096,
      "learning_rate": 1.1031934181156298e-05,
      "loss": 0.0108,
      "num_input_tokens_seen": 17954776,
      "step": 27470
    },
    {
      "epoch": 14.399895178197065,
      "grad_norm": 1.0775526762008667,
      "learning_rate": 1.1022452565887747e-05,
      "loss": 0.0172,
      "num_input_tokens_seen": 17958712,
      "step": 27475
    },
    {
      "epoch": 14.40251572327044,
      "grad_norm": 0.3756118714809418,
      "learning_rate": 1.1012973874562657e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 17961464,
      "step": 27480
    },
    {
      "epoch": 14.405136268343815,
      "grad_norm": 0.5541787147521973,
      "learning_rate": 1.1003498109163865e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 17965176,
      "step": 27485
    },
    {
      "epoch": 14.40775681341719,
      "grad_norm": 1.5338720083236694,
      "learning_rate": 1.0994025271673592e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 17968440,
      "step": 27490
    },
    {
      "epoch": 14.410377358490566,
      "grad_norm": 2.4432482719421387,
      "learning_rate": 1.0984555364073431e-05,
      "loss": 0.0123,
      "num_input_tokens_seen": 17971416,
      "step": 27495
    },
    {
      "epoch": 14.41299790356394,
      "grad_norm": 5.0735297203063965,
      "learning_rate": 1.0975088388344393e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 17974360,
      "step": 27500
    },
    {
      "epoch": 14.415618448637316,
      "grad_norm": 1.2899364233016968,
      "learning_rate": 1.096562434646686e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 17977912,
      "step": 27505
    },
    {
      "epoch": 14.418238993710691,
      "grad_norm": 1.3824970722198486,
      "learning_rate": 1.0956163240420611e-05,
      "loss": 0.0138,
      "num_input_tokens_seen": 17980792,
      "step": 27510
    },
    {
      "epoch": 14.420859538784066,
      "grad_norm": 0.39060553908348083,
      "learning_rate": 1.094670507218479e-05,
      "loss": 0.0093,
      "num_input_tokens_seen": 17983416,
      "step": 27515
    },
    {
      "epoch": 14.423480083857442,
      "grad_norm": 0.8790339231491089,
      "learning_rate": 1.093724984373793e-05,
      "loss": 0.0103,
      "num_input_tokens_seen": 17986328,
      "step": 27520
    },
    {
      "epoch": 14.426100628930818,
      "grad_norm": 2.4898734092712402,
      "learning_rate": 1.0927797557057968e-05,
      "loss": 0.0164,
      "num_input_tokens_seen": 17990584,
      "step": 27525
    },
    {
      "epoch": 14.428721174004194,
      "grad_norm": 1.8371996879577637,
      "learning_rate": 1.0918348214122209e-05,
      "loss": 0.0144,
      "num_input_tokens_seen": 17993848,
      "step": 27530
    },
    {
      "epoch": 14.431341719077569,
      "grad_norm": 2.4434452056884766,
      "learning_rate": 1.0908901816907366e-05,
      "loss": 0.025,
      "num_input_tokens_seen": 17997208,
      "step": 27535
    },
    {
      "epoch": 14.433962264150944,
      "grad_norm": 0.49829018115997314,
      "learning_rate": 1.08994583673895e-05,
      "loss": 0.0091,
      "num_input_tokens_seen": 18000056,
      "step": 27540
    },
    {
      "epoch": 14.43658280922432,
      "grad_norm": 0.915554940700531,
      "learning_rate": 1.0890017867544068e-05,
      "loss": 0.0495,
      "num_input_tokens_seen": 18002968,
      "step": 27545
    },
    {
      "epoch": 14.439203354297694,
      "grad_norm": 1.453652024269104,
      "learning_rate": 1.088058031934592e-05,
      "loss": 0.0118,
      "num_input_tokens_seen": 18006104,
      "step": 27550
    },
    {
      "epoch": 14.44182389937107,
      "grad_norm": 0.9078636765480042,
      "learning_rate": 1.087114572476929e-05,
      "loss": 0.0901,
      "num_input_tokens_seen": 18010040,
      "step": 27555
    },
    {
      "epoch": 14.444444444444445,
      "grad_norm": 2.419827699661255,
      "learning_rate": 1.0861714085787783e-05,
      "loss": 0.0204,
      "num_input_tokens_seen": 18013720,
      "step": 27560
    },
    {
      "epoch": 14.44706498951782,
      "grad_norm": 1.0825248956680298,
      "learning_rate": 1.0852285404374402e-05,
      "loss": 0.0126,
      "num_input_tokens_seen": 18016696,
      "step": 27565
    },
    {
      "epoch": 14.449685534591195,
      "grad_norm": 1.076959490776062,
      "learning_rate": 1.0842859682501505e-05,
      "loss": 0.0995,
      "num_input_tokens_seen": 18020024,
      "step": 27570
    },
    {
      "epoch": 14.45230607966457,
      "grad_norm": 3.4605977535247803,
      "learning_rate": 1.0833436922140841e-05,
      "loss": 0.0276,
      "num_input_tokens_seen": 18023352,
      "step": 27575
    },
    {
      "epoch": 14.454926624737945,
      "grad_norm": 2.3019042015075684,
      "learning_rate": 1.0824017125263555e-05,
      "loss": 0.0178,
      "num_input_tokens_seen": 18026648,
      "step": 27580
    },
    {
      "epoch": 14.45754716981132,
      "grad_norm": 3.4832842350006104,
      "learning_rate": 1.0814600293840154e-05,
      "loss": 0.0268,
      "num_input_tokens_seen": 18029624,
      "step": 27585
    },
    {
      "epoch": 14.460167714884696,
      "grad_norm": 1.2857807874679565,
      "learning_rate": 1.080518642984055e-05,
      "loss": 0.0609,
      "num_input_tokens_seen": 18032376,
      "step": 27590
    },
    {
      "epoch": 14.46278825995807,
      "grad_norm": 1.1619068384170532,
      "learning_rate": 1.0795775535233988e-05,
      "loss": 0.03,
      "num_input_tokens_seen": 18035448,
      "step": 27595
    },
    {
      "epoch": 14.465408805031446,
      "grad_norm": 0.6818233728408813,
      "learning_rate": 1.0786367611989145e-05,
      "loss": 0.0165,
      "num_input_tokens_seen": 18039576,
      "step": 27600
    },
    {
      "epoch": 14.468029350104821,
      "grad_norm": 3.8659002780914307,
      "learning_rate": 1.0776962662074031e-05,
      "loss": 0.013,
      "num_input_tokens_seen": 18044952,
      "step": 27605
    },
    {
      "epoch": 14.470649895178196,
      "grad_norm": 6.622477054595947,
      "learning_rate": 1.0767560687456062e-05,
      "loss": 0.029,
      "num_input_tokens_seen": 18048472,
      "step": 27610
    },
    {
      "epoch": 14.473270440251572,
      "grad_norm": 1.1002482175827026,
      "learning_rate": 1.0758161690102036e-05,
      "loss": 0.0118,
      "num_input_tokens_seen": 18052088,
      "step": 27615
    },
    {
      "epoch": 14.475890985324948,
      "grad_norm": 4.299016952514648,
      "learning_rate": 1.0748765671978095e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 18055000,
      "step": 27620
    },
    {
      "epoch": 14.478511530398324,
      "grad_norm": 1.4036054611206055,
      "learning_rate": 1.0739372635049797e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 18058712,
      "step": 27625
    },
    {
      "epoch": 14.481132075471699,
      "grad_norm": 0.6153357028961182,
      "learning_rate": 1.0729982581282042e-05,
      "loss": 0.0061,
      "num_input_tokens_seen": 18062616,
      "step": 27630
    },
    {
      "epoch": 14.483752620545074,
      "grad_norm": 2.047748327255249,
      "learning_rate": 1.0720595512639129e-05,
      "loss": 0.0147,
      "num_input_tokens_seen": 18066360,
      "step": 27635
    },
    {
      "epoch": 14.48637316561845,
      "grad_norm": 1.5533647537231445,
      "learning_rate": 1.0711211431084728e-05,
      "loss": 0.0108,
      "num_input_tokens_seen": 18069848,
      "step": 27640
    },
    {
      "epoch": 14.488993710691824,
      "grad_norm": 1.474660038948059,
      "learning_rate": 1.0701830338581887e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 18072504,
      "step": 27645
    },
    {
      "epoch": 14.4916142557652,
      "grad_norm": 0.40907683968544006,
      "learning_rate": 1.069245223709301e-05,
      "loss": 0.0284,
      "num_input_tokens_seen": 18075960,
      "step": 27650
    },
    {
      "epoch": 14.494234800838575,
      "grad_norm": 1.3741710186004639,
      "learning_rate": 1.0683077128579904e-05,
      "loss": 0.0247,
      "num_input_tokens_seen": 18079608,
      "step": 27655
    },
    {
      "epoch": 14.49685534591195,
      "grad_norm": 1.5076024532318115,
      "learning_rate": 1.0673705015003718e-05,
      "loss": 0.0237,
      "num_input_tokens_seen": 18082232,
      "step": 27660
    },
    {
      "epoch": 14.499475890985325,
      "grad_norm": 1.0527702569961548,
      "learning_rate": 1.0664335898325002e-05,
      "loss": 0.013,
      "num_input_tokens_seen": 18085432,
      "step": 27665
    },
    {
      "epoch": 14.5020964360587,
      "grad_norm": 0.520197331905365,
      "learning_rate": 1.065496978050367e-05,
      "loss": 0.0143,
      "num_input_tokens_seen": 18089176,
      "step": 27670
    },
    {
      "epoch": 14.504716981132075,
      "grad_norm": 5.811574459075928,
      "learning_rate": 1.0645606663499e-05,
      "loss": 0.0249,
      "num_input_tokens_seen": 18092600,
      "step": 27675
    },
    {
      "epoch": 14.50733752620545,
      "grad_norm": 4.54005241394043,
      "learning_rate": 1.0636246549269649e-05,
      "loss": 0.0567,
      "num_input_tokens_seen": 18096056,
      "step": 27680
    },
    {
      "epoch": 14.509958071278826,
      "grad_norm": 1.4094005823135376,
      "learning_rate": 1.062688943977366e-05,
      "loss": 0.0604,
      "num_input_tokens_seen": 18099192,
      "step": 27685
    },
    {
      "epoch": 14.5125786163522,
      "grad_norm": 0.7125228047370911,
      "learning_rate": 1.0617535336968413e-05,
      "loss": 0.0552,
      "num_input_tokens_seen": 18106744,
      "step": 27690
    },
    {
      "epoch": 14.515199161425576,
      "grad_norm": 0.7931041121482849,
      "learning_rate": 1.0608184242810698e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 18109592,
      "step": 27695
    },
    {
      "epoch": 14.517819706498951,
      "grad_norm": 1.0745691061019897,
      "learning_rate": 1.0598836159256642e-05,
      "loss": 0.0085,
      "num_input_tokens_seen": 18112632,
      "step": 27700
    },
    {
      "epoch": 14.520440251572326,
      "grad_norm": 1.5852411985397339,
      "learning_rate": 1.0589491088261763e-05,
      "loss": 0.0443,
      "num_input_tokens_seen": 18116824,
      "step": 27705
    },
    {
      "epoch": 14.523060796645701,
      "grad_norm": 6.015468597412109,
      "learning_rate": 1.0580149031780945e-05,
      "loss": 0.0407,
      "num_input_tokens_seen": 18120504,
      "step": 27710
    },
    {
      "epoch": 14.525681341719078,
      "grad_norm": 1.006323218345642,
      "learning_rate": 1.0570809991768447e-05,
      "loss": 0.0144,
      "num_input_tokens_seen": 18123768,
      "step": 27715
    },
    {
      "epoch": 14.528301886792454,
      "grad_norm": 0.8534802198410034,
      "learning_rate": 1.0561473970177873e-05,
      "loss": 0.0246,
      "num_input_tokens_seen": 18126584,
      "step": 27720
    },
    {
      "epoch": 14.530922431865829,
      "grad_norm": 1.7570191621780396,
      "learning_rate": 1.0552140968962232e-05,
      "loss": 0.0312,
      "num_input_tokens_seen": 18131480,
      "step": 27725
    },
    {
      "epoch": 14.533542976939204,
      "grad_norm": 1.6914054155349731,
      "learning_rate": 1.054281099007386e-05,
      "loss": 0.0389,
      "num_input_tokens_seen": 18135224,
      "step": 27730
    },
    {
      "epoch": 14.536163522012579,
      "grad_norm": 1.1765108108520508,
      "learning_rate": 1.0533484035464492e-05,
      "loss": 0.0231,
      "num_input_tokens_seen": 18138232,
      "step": 27735
    },
    {
      "epoch": 14.538784067085954,
      "grad_norm": 1.462307333946228,
      "learning_rate": 1.0524160107085231e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 18142040,
      "step": 27740
    },
    {
      "epoch": 14.54140461215933,
      "grad_norm": 4.1231689453125,
      "learning_rate": 1.0514839206886518e-05,
      "loss": 0.0287,
      "num_input_tokens_seen": 18144888,
      "step": 27745
    },
    {
      "epoch": 14.544025157232705,
      "grad_norm": 2.2301933765411377,
      "learning_rate": 1.0505521336818195e-05,
      "loss": 0.015,
      "num_input_tokens_seen": 18147672,
      "step": 27750
    },
    {
      "epoch": 14.54664570230608,
      "grad_norm": 2.479201316833496,
      "learning_rate": 1.0496206498829438e-05,
      "loss": 0.0075,
      "num_input_tokens_seen": 18150808,
      "step": 27755
    },
    {
      "epoch": 14.549266247379455,
      "grad_norm": 1.3953481912612915,
      "learning_rate": 1.0486894694868814e-05,
      "loss": 0.017,
      "num_input_tokens_seen": 18153400,
      "step": 27760
    },
    {
      "epoch": 14.55188679245283,
      "grad_norm": 0.7540873289108276,
      "learning_rate": 1.0477585926884246e-05,
      "loss": 0.0131,
      "num_input_tokens_seen": 18157432,
      "step": 27765
    },
    {
      "epoch": 14.554507337526205,
      "grad_norm": 0.8095356822013855,
      "learning_rate": 1.046828019682303e-05,
      "loss": 0.0177,
      "num_input_tokens_seen": 18160472,
      "step": 27770
    },
    {
      "epoch": 14.55712788259958,
      "grad_norm": 0.8317137360572815,
      "learning_rate": 1.0458977506631808e-05,
      "loss": 0.0178,
      "num_input_tokens_seen": 18163608,
      "step": 27775
    },
    {
      "epoch": 14.559748427672956,
      "grad_norm": 2.185107707977295,
      "learning_rate": 1.0449677858256593e-05,
      "loss": 0.0165,
      "num_input_tokens_seen": 18167736,
      "step": 27780
    },
    {
      "epoch": 14.56236897274633,
      "grad_norm": 0.4142165780067444,
      "learning_rate": 1.0440381253642773e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 18171160,
      "step": 27785
    },
    {
      "epoch": 14.564989517819706,
      "grad_norm": 0.8261743187904358,
      "learning_rate": 1.0431087694735087e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 18174168,
      "step": 27790
    },
    {
      "epoch": 14.567610062893081,
      "grad_norm": 0.2186216115951538,
      "learning_rate": 1.0421797183477644e-05,
      "loss": 0.0371,
      "num_input_tokens_seen": 18177080,
      "step": 27795
    },
    {
      "epoch": 14.570230607966456,
      "grad_norm": 3.5375406742095947,
      "learning_rate": 1.041250972181392e-05,
      "loss": 0.0531,
      "num_input_tokens_seen": 18181976,
      "step": 27800
    },
    {
      "epoch": 14.572851153039831,
      "grad_norm": 1.3109173774719238,
      "learning_rate": 1.0403225311686743e-05,
      "loss": 0.0438,
      "num_input_tokens_seen": 18186168,
      "step": 27805
    },
    {
      "epoch": 14.575471698113208,
      "grad_norm": 0.847129225730896,
      "learning_rate": 1.0393943955038293e-05,
      "loss": 0.0199,
      "num_input_tokens_seen": 18189656,
      "step": 27810
    },
    {
      "epoch": 14.578092243186584,
      "grad_norm": 0.16079740226268768,
      "learning_rate": 1.038466565381013e-05,
      "loss": 0.0209,
      "num_input_tokens_seen": 18191768,
      "step": 27815
    },
    {
      "epoch": 14.580712788259959,
      "grad_norm": 2.1183464527130127,
      "learning_rate": 1.0375390409943175e-05,
      "loss": 0.0115,
      "num_input_tokens_seen": 18194296,
      "step": 27820
    },
    {
      "epoch": 14.583333333333334,
      "grad_norm": 5.045507907867432,
      "learning_rate": 1.0366118225377708e-05,
      "loss": 0.0167,
      "num_input_tokens_seen": 18197528,
      "step": 27825
    },
    {
      "epoch": 14.585953878406709,
      "grad_norm": 1.3487520217895508,
      "learning_rate": 1.0356849102053353e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 18200600,
      "step": 27830
    },
    {
      "epoch": 14.588574423480084,
      "grad_norm": 0.9448147416114807,
      "learning_rate": 1.0347583041909104e-05,
      "loss": 0.0251,
      "num_input_tokens_seen": 18203864,
      "step": 27835
    },
    {
      "epoch": 14.59119496855346,
      "grad_norm": 0.6061326861381531,
      "learning_rate": 1.0338320046883315e-05,
      "loss": 0.021,
      "num_input_tokens_seen": 18207192,
      "step": 27840
    },
    {
      "epoch": 14.593815513626835,
      "grad_norm": 2.529022693634033,
      "learning_rate": 1.0329060118913705e-05,
      "loss": 0.0193,
      "num_input_tokens_seen": 18211352,
      "step": 27845
    },
    {
      "epoch": 14.59643605870021,
      "grad_norm": 3.0535218715667725,
      "learning_rate": 1.0319803259937353e-05,
      "loss": 0.0207,
      "num_input_tokens_seen": 18214200,
      "step": 27850
    },
    {
      "epoch": 14.599056603773585,
      "grad_norm": 0.8195034861564636,
      "learning_rate": 1.0310549471890676e-05,
      "loss": 0.011,
      "num_input_tokens_seen": 18216920,
      "step": 27855
    },
    {
      "epoch": 14.60167714884696,
      "grad_norm": 0.8985005021095276,
      "learning_rate": 1.0301298756709454e-05,
      "loss": 0.0325,
      "num_input_tokens_seen": 18220184,
      "step": 27860
    },
    {
      "epoch": 14.604297693920335,
      "grad_norm": 1.6437039375305176,
      "learning_rate": 1.0292051116328844e-05,
      "loss": 0.0359,
      "num_input_tokens_seen": 18224280,
      "step": 27865
    },
    {
      "epoch": 14.60691823899371,
      "grad_norm": 0.09424322098493576,
      "learning_rate": 1.028280655268334e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 18226744,
      "step": 27870
    },
    {
      "epoch": 14.609538784067086,
      "grad_norm": 1.9730799198150635,
      "learning_rate": 1.0273565067706806e-05,
      "loss": 0.0301,
      "num_input_tokens_seen": 18230232,
      "step": 27875
    },
    {
      "epoch": 14.61215932914046,
      "grad_norm": 1.6140916347503662,
      "learning_rate": 1.0264326663332457e-05,
      "loss": 0.039,
      "num_input_tokens_seen": 18234040,
      "step": 27880
    },
    {
      "epoch": 14.614779874213836,
      "grad_norm": 0.8956087231636047,
      "learning_rate": 1.0255091341492859e-05,
      "loss": 0.0034,
      "num_input_tokens_seen": 18236632,
      "step": 27885
    },
    {
      "epoch": 14.617400419287211,
      "grad_norm": 1.6873764991760254,
      "learning_rate": 1.0245859104119923e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 18239640,
      "step": 27890
    },
    {
      "epoch": 14.620020964360586,
      "grad_norm": 0.7225280404090881,
      "learning_rate": 1.0236629953144938e-05,
      "loss": 0.0161,
      "num_input_tokens_seen": 18242648,
      "step": 27895
    },
    {
      "epoch": 14.622641509433961,
      "grad_norm": 1.9347418546676636,
      "learning_rate": 1.0227403890498537e-05,
      "loss": 0.0234,
      "num_input_tokens_seen": 18246584,
      "step": 27900
    },
    {
      "epoch": 14.625262054507338,
      "grad_norm": 0.581731915473938,
      "learning_rate": 1.021818091811072e-05,
      "loss": 0.0645,
      "num_input_tokens_seen": 18249400,
      "step": 27905
    },
    {
      "epoch": 14.627882599580714,
      "grad_norm": 0.9897846579551697,
      "learning_rate": 1.0208961037910813e-05,
      "loss": 0.0202,
      "num_input_tokens_seen": 18252952,
      "step": 27910
    },
    {
      "epoch": 14.630503144654089,
      "grad_norm": 1.0393372774124146,
      "learning_rate": 1.0199744251827503e-05,
      "loss": 0.0583,
      "num_input_tokens_seen": 18257176,
      "step": 27915
    },
    {
      "epoch": 14.633123689727464,
      "grad_norm": 1.1668686866760254,
      "learning_rate": 1.0190530561788845e-05,
      "loss": 0.0364,
      "num_input_tokens_seen": 18260280,
      "step": 27920
    },
    {
      "epoch": 14.635744234800839,
      "grad_norm": 0.6664900183677673,
      "learning_rate": 1.0181319969722236e-05,
      "loss": 0.0198,
      "num_input_tokens_seen": 18263480,
      "step": 27925
    },
    {
      "epoch": 14.638364779874214,
      "grad_norm": 2.079829216003418,
      "learning_rate": 1.0172112477554441e-05,
      "loss": 0.025,
      "num_input_tokens_seen": 18267128,
      "step": 27930
    },
    {
      "epoch": 14.64098532494759,
      "grad_norm": 0.9155235886573792,
      "learning_rate": 1.016290808721154e-05,
      "loss": 0.0287,
      "num_input_tokens_seen": 18270968,
      "step": 27935
    },
    {
      "epoch": 14.643605870020965,
      "grad_norm": 3.3411128520965576,
      "learning_rate": 1.0153706800619007e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 18273720,
      "step": 27940
    },
    {
      "epoch": 14.64622641509434,
      "grad_norm": 3.5408990383148193,
      "learning_rate": 1.0144508619701623e-05,
      "loss": 0.0183,
      "num_input_tokens_seen": 18276920,
      "step": 27945
    },
    {
      "epoch": 14.648846960167715,
      "grad_norm": 1.4632407426834106,
      "learning_rate": 1.013531354638356e-05,
      "loss": 0.0271,
      "num_input_tokens_seen": 18280952,
      "step": 27950
    },
    {
      "epoch": 14.65146750524109,
      "grad_norm": 0.7618055939674377,
      "learning_rate": 1.0126121582588316e-05,
      "loss": 0.0708,
      "num_input_tokens_seen": 18284120,
      "step": 27955
    },
    {
      "epoch": 14.654088050314465,
      "grad_norm": 1.9405392408370972,
      "learning_rate": 1.0116932730238754e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 18288120,
      "step": 27960
    },
    {
      "epoch": 14.65670859538784,
      "grad_norm": 1.4926246404647827,
      "learning_rate": 1.0107746991257063e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 18291224,
      "step": 27965
    },
    {
      "epoch": 14.659329140461216,
      "grad_norm": 1.1160269975662231,
      "learning_rate": 1.0098564367564811e-05,
      "loss": 0.0128,
      "num_input_tokens_seen": 18294520,
      "step": 27970
    },
    {
      "epoch": 14.66194968553459,
      "grad_norm": 1.458040475845337,
      "learning_rate": 1.0089384861082876e-05,
      "loss": 0.0123,
      "num_input_tokens_seen": 18298296,
      "step": 27975
    },
    {
      "epoch": 14.664570230607966,
      "grad_norm": 1.8105204105377197,
      "learning_rate": 1.0080208473731523e-05,
      "loss": 0.0122,
      "num_input_tokens_seen": 18301016,
      "step": 27980
    },
    {
      "epoch": 14.667190775681341,
      "grad_norm": 1.6641277074813843,
      "learning_rate": 1.0071035207430352e-05,
      "loss": 0.0192,
      "num_input_tokens_seen": 18304728,
      "step": 27985
    },
    {
      "epoch": 14.669811320754716,
      "grad_norm": 1.4573036432266235,
      "learning_rate": 1.0061865064098285e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 18307736,
      "step": 27990
    },
    {
      "epoch": 14.672431865828091,
      "grad_norm": 4.177162170410156,
      "learning_rate": 1.0052698045653625e-05,
      "loss": 0.0175,
      "num_input_tokens_seen": 18310744,
      "step": 27995
    },
    {
      "epoch": 14.675052410901468,
      "grad_norm": 0.05129533261060715,
      "learning_rate": 1.0043534154014015e-05,
      "loss": 0.0213,
      "num_input_tokens_seen": 18313336,
      "step": 28000
    },
    {
      "epoch": 14.677672955974844,
      "grad_norm": 2.5853567123413086,
      "learning_rate": 1.003437339109642e-05,
      "loss": 0.04,
      "num_input_tokens_seen": 18316120,
      "step": 28005
    },
    {
      "epoch": 14.680293501048219,
      "grad_norm": 1.9580156803131104,
      "learning_rate": 1.0025215758817183e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 18319480,
      "step": 28010
    },
    {
      "epoch": 14.682914046121594,
      "grad_norm": 1.0581215620040894,
      "learning_rate": 1.0016061259091958e-05,
      "loss": 0.0136,
      "num_input_tokens_seen": 18323288,
      "step": 28015
    },
    {
      "epoch": 14.685534591194969,
      "grad_norm": 1.0847879648208618,
      "learning_rate": 1.000690989383577e-05,
      "loss": 0.0148,
      "num_input_tokens_seen": 18325752,
      "step": 28020
    },
    {
      "epoch": 14.688155136268344,
      "grad_norm": 1.81784188747406,
      "learning_rate": 9.997761664962993e-06,
      "loss": 0.0249,
      "num_input_tokens_seen": 18330424,
      "step": 28025
    },
    {
      "epoch": 14.69077568134172,
      "grad_norm": 0.2938399612903595,
      "learning_rate": 9.988616574387314e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 18333336,
      "step": 28030
    },
    {
      "epoch": 14.693396226415095,
      "grad_norm": 0.9932610392570496,
      "learning_rate": 9.979474624021787e-06,
      "loss": 0.0226,
      "num_input_tokens_seen": 18336760,
      "step": 28035
    },
    {
      "epoch": 14.69601677148847,
      "grad_norm": 1.3778831958770752,
      "learning_rate": 9.970335815778815e-06,
      "loss": 0.0139,
      "num_input_tokens_seen": 18339288,
      "step": 28040
    },
    {
      "epoch": 14.698637316561845,
      "grad_norm": 4.632768630981445,
      "learning_rate": 9.96120015157011e-06,
      "loss": 0.0695,
      "num_input_tokens_seen": 18343064,
      "step": 28045
    },
    {
      "epoch": 14.70125786163522,
      "grad_norm": 5.739060878753662,
      "learning_rate": 9.952067633306767e-06,
      "loss": 0.0283,
      "num_input_tokens_seen": 18346584,
      "step": 28050
    },
    {
      "epoch": 14.703878406708595,
      "grad_norm": 0.6928592920303345,
      "learning_rate": 9.942938262899204e-06,
      "loss": 0.0187,
      "num_input_tokens_seen": 18349752,
      "step": 28055
    },
    {
      "epoch": 14.70649895178197,
      "grad_norm": 0.832679033279419,
      "learning_rate": 9.93381204225717e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 18352632,
      "step": 28060
    },
    {
      "epoch": 14.709119496855346,
      "grad_norm": 1.4742854833602905,
      "learning_rate": 9.924688973289782e-06,
      "loss": 0.0476,
      "num_input_tokens_seen": 18356280,
      "step": 28065
    },
    {
      "epoch": 14.71174004192872,
      "grad_norm": 0.6743787527084351,
      "learning_rate": 9.915569057905463e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 18359512,
      "step": 28070
    },
    {
      "epoch": 14.714360587002096,
      "grad_norm": 7.732043743133545,
      "learning_rate": 9.906452298012003e-06,
      "loss": 0.0249,
      "num_input_tokens_seen": 18362072,
      "step": 28075
    },
    {
      "epoch": 14.716981132075471,
      "grad_norm": 1.2542988061904907,
      "learning_rate": 9.897338695516529e-06,
      "loss": 0.0259,
      "num_input_tokens_seen": 18365400,
      "step": 28080
    },
    {
      "epoch": 14.719601677148846,
      "grad_norm": 1.8762843608856201,
      "learning_rate": 9.888228252325507e-06,
      "loss": 0.0331,
      "num_input_tokens_seen": 18369176,
      "step": 28085
    },
    {
      "epoch": 14.722222222222221,
      "grad_norm": 1.864773154258728,
      "learning_rate": 9.87912097034473e-06,
      "loss": 0.0097,
      "num_input_tokens_seen": 18372472,
      "step": 28090
    },
    {
      "epoch": 14.724842767295598,
      "grad_norm": 1.1690119504928589,
      "learning_rate": 9.870016851479327e-06,
      "loss": 0.0342,
      "num_input_tokens_seen": 18375608,
      "step": 28095
    },
    {
      "epoch": 14.727463312368974,
      "grad_norm": 3.4170515537261963,
      "learning_rate": 9.860915897633791e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 18378424,
      "step": 28100
    },
    {
      "epoch": 14.730083857442349,
      "grad_norm": 2.3757097721099854,
      "learning_rate": 9.85181811071193e-06,
      "loss": 0.057,
      "num_input_tokens_seen": 18383544,
      "step": 28105
    },
    {
      "epoch": 14.732704402515724,
      "grad_norm": 1.4446423053741455,
      "learning_rate": 9.842723492616913e-06,
      "loss": 0.02,
      "num_input_tokens_seen": 18386488,
      "step": 28110
    },
    {
      "epoch": 14.735324947589099,
      "grad_norm": 0.3762805461883545,
      "learning_rate": 9.833632045251207e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 18389272,
      "step": 28115
    },
    {
      "epoch": 14.737945492662474,
      "grad_norm": 0.566523015499115,
      "learning_rate": 9.824543770516656e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 18393240,
      "step": 28120
    },
    {
      "epoch": 14.74056603773585,
      "grad_norm": 9.580687522888184,
      "learning_rate": 9.815458670314412e-06,
      "loss": 0.0235,
      "num_input_tokens_seen": 18396504,
      "step": 28125
    },
    {
      "epoch": 14.743186582809225,
      "grad_norm": 0.14979398250579834,
      "learning_rate": 9.80637674654498e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 18399096,
      "step": 28130
    },
    {
      "epoch": 14.7458071278826,
      "grad_norm": 0.39949485659599304,
      "learning_rate": 9.797298001108193e-06,
      "loss": 0.018,
      "num_input_tokens_seen": 18402936,
      "step": 28135
    },
    {
      "epoch": 14.748427672955975,
      "grad_norm": 0.639892041683197,
      "learning_rate": 9.788222435903232e-06,
      "loss": 0.0181,
      "num_input_tokens_seen": 18406008,
      "step": 28140
    },
    {
      "epoch": 14.75104821802935,
      "grad_norm": 0.41376689076423645,
      "learning_rate": 9.779150052828596e-06,
      "loss": 0.0528,
      "num_input_tokens_seen": 18409752,
      "step": 28145
    },
    {
      "epoch": 14.753668763102725,
      "grad_norm": 6.764246463775635,
      "learning_rate": 9.770080853782108e-06,
      "loss": 0.0184,
      "num_input_tokens_seen": 18412408,
      "step": 28150
    },
    {
      "epoch": 14.7562893081761,
      "grad_norm": 2.933867931365967,
      "learning_rate": 9.76101484066095e-06,
      "loss": 0.0466,
      "num_input_tokens_seen": 18415672,
      "step": 28155
    },
    {
      "epoch": 14.758909853249476,
      "grad_norm": 2.885996103286743,
      "learning_rate": 9.751952015361637e-06,
      "loss": 0.039,
      "num_input_tokens_seen": 18418904,
      "step": 28160
    },
    {
      "epoch": 14.76153039832285,
      "grad_norm": 0.9931508898735046,
      "learning_rate": 9.742892379780003e-06,
      "loss": 0.0173,
      "num_input_tokens_seen": 18421816,
      "step": 28165
    },
    {
      "epoch": 14.764150943396226,
      "grad_norm": 1.5434621572494507,
      "learning_rate": 9.733835935811237e-06,
      "loss": 0.0869,
      "num_input_tokens_seen": 18424696,
      "step": 28170
    },
    {
      "epoch": 14.766771488469601,
      "grad_norm": 0.7092440128326416,
      "learning_rate": 9.724782685349809e-06,
      "loss": 0.0471,
      "num_input_tokens_seen": 18427864,
      "step": 28175
    },
    {
      "epoch": 14.769392033542976,
      "grad_norm": 1.606374979019165,
      "learning_rate": 9.715732630289576e-06,
      "loss": 0.0304,
      "num_input_tokens_seen": 18430872,
      "step": 28180
    },
    {
      "epoch": 14.772012578616351,
      "grad_norm": 1.4828917980194092,
      "learning_rate": 9.706685772523701e-06,
      "loss": 0.0142,
      "num_input_tokens_seen": 18433400,
      "step": 28185
    },
    {
      "epoch": 14.774633123689728,
      "grad_norm": 2.5786900520324707,
      "learning_rate": 9.697642113944692e-06,
      "loss": 0.0326,
      "num_input_tokens_seen": 18436536,
      "step": 28190
    },
    {
      "epoch": 14.777253668763104,
      "grad_norm": 1.1768791675567627,
      "learning_rate": 9.688601656444381e-06,
      "loss": 0.0229,
      "num_input_tokens_seen": 18439896,
      "step": 28195
    },
    {
      "epoch": 14.779874213836479,
      "grad_norm": 1.5763459205627441,
      "learning_rate": 9.679564401913921e-06,
      "loss": 0.0222,
      "num_input_tokens_seen": 18445112,
      "step": 28200
    },
    {
      "epoch": 14.782494758909854,
      "grad_norm": 0.5930331349372864,
      "learning_rate": 9.670530352243798e-06,
      "loss": 0.013,
      "num_input_tokens_seen": 18448056,
      "step": 28205
    },
    {
      "epoch": 14.785115303983229,
      "grad_norm": 0.0841803029179573,
      "learning_rate": 9.661499509323835e-06,
      "loss": 0.0246,
      "num_input_tokens_seen": 18451320,
      "step": 28210
    },
    {
      "epoch": 14.787735849056604,
      "grad_norm": 0.5225451588630676,
      "learning_rate": 9.652471875043182e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 18454168,
      "step": 28215
    },
    {
      "epoch": 14.79035639412998,
      "grad_norm": 0.6267397999763489,
      "learning_rate": 9.64344745129033e-06,
      "loss": 0.0269,
      "num_input_tokens_seen": 18457144,
      "step": 28220
    },
    {
      "epoch": 14.792976939203355,
      "grad_norm": 1.4381718635559082,
      "learning_rate": 9.634426239953073e-06,
      "loss": 0.0154,
      "num_input_tokens_seen": 18460088,
      "step": 28225
    },
    {
      "epoch": 14.79559748427673,
      "grad_norm": 0.1553766280412674,
      "learning_rate": 9.62540824291854e-06,
      "loss": 0.0513,
      "num_input_tokens_seen": 18463800,
      "step": 28230
    },
    {
      "epoch": 14.798218029350105,
      "grad_norm": 2.1629374027252197,
      "learning_rate": 9.616393462073197e-06,
      "loss": 0.0282,
      "num_input_tokens_seen": 18466744,
      "step": 28235
    },
    {
      "epoch": 14.80083857442348,
      "grad_norm": 0.7387337684631348,
      "learning_rate": 9.607381899302839e-06,
      "loss": 0.0457,
      "num_input_tokens_seen": 18469848,
      "step": 28240
    },
    {
      "epoch": 14.803459119496855,
      "grad_norm": 1.190223217010498,
      "learning_rate": 9.59837355649258e-06,
      "loss": 0.032,
      "num_input_tokens_seen": 18473496,
      "step": 28245
    },
    {
      "epoch": 14.80607966457023,
      "grad_norm": 1.7248284816741943,
      "learning_rate": 9.589368435526868e-06,
      "loss": 0.021,
      "num_input_tokens_seen": 18477048,
      "step": 28250
    },
    {
      "epoch": 14.808700209643606,
      "grad_norm": 2.5546956062316895,
      "learning_rate": 9.580366538289468e-06,
      "loss": 0.0302,
      "num_input_tokens_seen": 18479640,
      "step": 28255
    },
    {
      "epoch": 14.81132075471698,
      "grad_norm": 1.7872364521026611,
      "learning_rate": 9.57136786666346e-06,
      "loss": 0.033,
      "num_input_tokens_seen": 18482808,
      "step": 28260
    },
    {
      "epoch": 14.813941299790356,
      "grad_norm": 0.7557271122932434,
      "learning_rate": 9.562372422531276e-06,
      "loss": 0.0151,
      "num_input_tokens_seen": 18485496,
      "step": 28265
    },
    {
      "epoch": 14.816561844863731,
      "grad_norm": 1.7166101932525635,
      "learning_rate": 9.553380207774662e-06,
      "loss": 0.0215,
      "num_input_tokens_seen": 18488568,
      "step": 28270
    },
    {
      "epoch": 14.819182389937106,
      "grad_norm": 1.6323484182357788,
      "learning_rate": 9.54439122427469e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 18491256,
      "step": 28275
    },
    {
      "epoch": 14.821802935010481,
      "grad_norm": 4.936087608337402,
      "learning_rate": 9.535405473911748e-06,
      "loss": 0.0165,
      "num_input_tokens_seen": 18494424,
      "step": 28280
    },
    {
      "epoch": 14.824423480083858,
      "grad_norm": 1.485011339187622,
      "learning_rate": 9.526422958565546e-06,
      "loss": 0.0262,
      "num_input_tokens_seen": 18497496,
      "step": 28285
    },
    {
      "epoch": 14.827044025157234,
      "grad_norm": 2.064101457595825,
      "learning_rate": 9.517443680115124e-06,
      "loss": 0.0233,
      "num_input_tokens_seen": 18501304,
      "step": 28290
    },
    {
      "epoch": 14.829664570230609,
      "grad_norm": 2.513773202896118,
      "learning_rate": 9.50846764043885e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 18505048,
      "step": 28295
    },
    {
      "epoch": 14.832285115303984,
      "grad_norm": 0.710281252861023,
      "learning_rate": 9.499494841414422e-06,
      "loss": 0.0298,
      "num_input_tokens_seen": 18508216,
      "step": 28300
    },
    {
      "epoch": 14.834905660377359,
      "grad_norm": 1.3603289127349854,
      "learning_rate": 9.490525284918822e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 18511192,
      "step": 28305
    },
    {
      "epoch": 14.837526205450734,
      "grad_norm": 1.914388656616211,
      "learning_rate": 9.4815589728284e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 18514296,
      "step": 28310
    },
    {
      "epoch": 14.84014675052411,
      "grad_norm": 1.3680388927459717,
      "learning_rate": 9.472595907018788e-06,
      "loss": 0.0577,
      "num_input_tokens_seen": 18517560,
      "step": 28315
    },
    {
      "epoch": 14.842767295597485,
      "grad_norm": 0.830146312713623,
      "learning_rate": 9.463636089364968e-06,
      "loss": 0.0447,
      "num_input_tokens_seen": 18520888,
      "step": 28320
    },
    {
      "epoch": 14.84538784067086,
      "grad_norm": 1.6242562532424927,
      "learning_rate": 9.454679521741239e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 18525016,
      "step": 28325
    },
    {
      "epoch": 14.848008385744235,
      "grad_norm": 0.7385802268981934,
      "learning_rate": 9.445726206021194e-06,
      "loss": 0.0258,
      "num_input_tokens_seen": 18529112,
      "step": 28330
    },
    {
      "epoch": 14.85062893081761,
      "grad_norm": 0.9625027775764465,
      "learning_rate": 9.436776144077772e-06,
      "loss": 0.0225,
      "num_input_tokens_seen": 18532344,
      "step": 28335
    },
    {
      "epoch": 14.853249475890985,
      "grad_norm": 2.254093885421753,
      "learning_rate": 9.427829337783237e-06,
      "loss": 0.0116,
      "num_input_tokens_seen": 18535608,
      "step": 28340
    },
    {
      "epoch": 14.85587002096436,
      "grad_norm": 1.0847538709640503,
      "learning_rate": 9.418885789009138e-06,
      "loss": 0.0221,
      "num_input_tokens_seen": 18538200,
      "step": 28345
    },
    {
      "epoch": 14.858490566037736,
      "grad_norm": 1.7986754179000854,
      "learning_rate": 9.409945499626374e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 18540600,
      "step": 28350
    },
    {
      "epoch": 14.86111111111111,
      "grad_norm": 0.9477250576019287,
      "learning_rate": 9.40100847150516e-06,
      "loss": 0.0091,
      "num_input_tokens_seen": 18543192,
      "step": 28355
    },
    {
      "epoch": 14.863731656184486,
      "grad_norm": 1.6878414154052734,
      "learning_rate": 9.392074706515003e-06,
      "loss": 0.022,
      "num_input_tokens_seen": 18546488,
      "step": 28360
    },
    {
      "epoch": 14.866352201257861,
      "grad_norm": 1.0872855186462402,
      "learning_rate": 9.383144206524753e-06,
      "loss": 0.0161,
      "num_input_tokens_seen": 18549752,
      "step": 28365
    },
    {
      "epoch": 14.868972746331236,
      "grad_norm": 2.3680577278137207,
      "learning_rate": 9.374216973402575e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 18552696,
      "step": 28370
    },
    {
      "epoch": 14.871593291404611,
      "grad_norm": 7.8024001121521,
      "learning_rate": 9.365293009015926e-06,
      "loss": 0.0304,
      "num_input_tokens_seen": 18555416,
      "step": 28375
    },
    {
      "epoch": 14.874213836477988,
      "grad_norm": 1.0287858247756958,
      "learning_rate": 9.356372315231623e-06,
      "loss": 0.0468,
      "num_input_tokens_seen": 18558712,
      "step": 28380
    },
    {
      "epoch": 14.876834381551364,
      "grad_norm": 1.5076217651367188,
      "learning_rate": 9.347454893915749e-06,
      "loss": 0.0165,
      "num_input_tokens_seen": 18562072,
      "step": 28385
    },
    {
      "epoch": 14.879454926624739,
      "grad_norm": 0.6259467601776123,
      "learning_rate": 9.338540746933741e-06,
      "loss": 0.0184,
      "num_input_tokens_seen": 18565208,
      "step": 28390
    },
    {
      "epoch": 14.882075471698114,
      "grad_norm": 3.8484911918640137,
      "learning_rate": 9.329629876150339e-06,
      "loss": 0.0333,
      "num_input_tokens_seen": 18568568,
      "step": 28395
    },
    {
      "epoch": 14.884696016771489,
      "grad_norm": 0.395988792181015,
      "learning_rate": 9.320722283429581e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 18571384,
      "step": 28400
    },
    {
      "epoch": 14.887316561844864,
      "grad_norm": 1.075134515762329,
      "learning_rate": 9.311817970634854e-06,
      "loss": 0.0142,
      "num_input_tokens_seen": 18574680,
      "step": 28405
    },
    {
      "epoch": 14.88993710691824,
      "grad_norm": 0.34610259532928467,
      "learning_rate": 9.302916939628814e-06,
      "loss": 0.0142,
      "num_input_tokens_seen": 18577464,
      "step": 28410
    },
    {
      "epoch": 14.892557651991615,
      "grad_norm": 0.8868510723114014,
      "learning_rate": 9.29401919227347e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 18580024,
      "step": 28415
    },
    {
      "epoch": 14.89517819706499,
      "grad_norm": 2.2293403148651123,
      "learning_rate": 9.28512473043013e-06,
      "loss": 0.0305,
      "num_input_tokens_seen": 18582840,
      "step": 28420
    },
    {
      "epoch": 14.897798742138365,
      "grad_norm": 5.346724033355713,
      "learning_rate": 9.276233555959419e-06,
      "loss": 0.0748,
      "num_input_tokens_seen": 18585880,
      "step": 28425
    },
    {
      "epoch": 14.90041928721174,
      "grad_norm": 6.170387268066406,
      "learning_rate": 9.267345670721254e-06,
      "loss": 0.0224,
      "num_input_tokens_seen": 18588920,
      "step": 28430
    },
    {
      "epoch": 14.903039832285115,
      "grad_norm": 1.087564468383789,
      "learning_rate": 9.258461076574895e-06,
      "loss": 0.0169,
      "num_input_tokens_seen": 18591736,
      "step": 28435
    },
    {
      "epoch": 14.90566037735849,
      "grad_norm": 3.813788890838623,
      "learning_rate": 9.24957977537888e-06,
      "loss": 0.0293,
      "num_input_tokens_seen": 18594392,
      "step": 28440
    },
    {
      "epoch": 14.908280922431866,
      "grad_norm": 0.9880952835083008,
      "learning_rate": 9.240701768991086e-06,
      "loss": 0.0273,
      "num_input_tokens_seen": 18598424,
      "step": 28445
    },
    {
      "epoch": 14.91090146750524,
      "grad_norm": 2.0746090412139893,
      "learning_rate": 9.231827059268692e-06,
      "loss": 0.0179,
      "num_input_tokens_seen": 18601976,
      "step": 28450
    },
    {
      "epoch": 14.913522012578616,
      "grad_norm": 0.9182659983634949,
      "learning_rate": 9.222955648068193e-06,
      "loss": 0.0181,
      "num_input_tokens_seen": 18604568,
      "step": 28455
    },
    {
      "epoch": 14.916142557651991,
      "grad_norm": 1.0153746604919434,
      "learning_rate": 9.214087537245378e-06,
      "loss": 0.0124,
      "num_input_tokens_seen": 18607544,
      "step": 28460
    },
    {
      "epoch": 14.918763102725366,
      "grad_norm": 1.4195510149002075,
      "learning_rate": 9.205222728655344e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 18611352,
      "step": 28465
    },
    {
      "epoch": 14.921383647798741,
      "grad_norm": 0.5334834456443787,
      "learning_rate": 9.196361224152522e-06,
      "loss": 0.0144,
      "num_input_tokens_seen": 18616920,
      "step": 28470
    },
    {
      "epoch": 14.924004192872118,
      "grad_norm": 0.15966035425662994,
      "learning_rate": 9.187503025590632e-06,
      "loss": 0.0505,
      "num_input_tokens_seen": 18619544,
      "step": 28475
    },
    {
      "epoch": 14.926624737945493,
      "grad_norm": 2.3318300247192383,
      "learning_rate": 9.17864813482272e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 18622072,
      "step": 28480
    },
    {
      "epoch": 14.929245283018869,
      "grad_norm": 1.0442644357681274,
      "learning_rate": 9.169796553701122e-06,
      "loss": 0.0517,
      "num_input_tokens_seen": 18624952,
      "step": 28485
    },
    {
      "epoch": 14.931865828092244,
      "grad_norm": 7.6955485343933105,
      "learning_rate": 9.160948284077473e-06,
      "loss": 0.0726,
      "num_input_tokens_seen": 18627672,
      "step": 28490
    },
    {
      "epoch": 14.934486373165619,
      "grad_norm": 1.774337887763977,
      "learning_rate": 9.152103327802738e-06,
      "loss": 0.0203,
      "num_input_tokens_seen": 18631160,
      "step": 28495
    },
    {
      "epoch": 14.937106918238994,
      "grad_norm": 1.5196112394332886,
      "learning_rate": 9.143261686727189e-06,
      "loss": 0.0351,
      "num_input_tokens_seen": 18634168,
      "step": 28500
    },
    {
      "epoch": 14.93972746331237,
      "grad_norm": 0.7768208980560303,
      "learning_rate": 9.134423362700392e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 18636984,
      "step": 28505
    },
    {
      "epoch": 14.942348008385745,
      "grad_norm": 2.321002244949341,
      "learning_rate": 9.12558835757123e-06,
      "loss": 0.0306,
      "num_input_tokens_seen": 18639864,
      "step": 28510
    },
    {
      "epoch": 14.94496855345912,
      "grad_norm": 0.8064241409301758,
      "learning_rate": 9.116756673187878e-06,
      "loss": 0.0206,
      "num_input_tokens_seen": 18642264,
      "step": 28515
    },
    {
      "epoch": 14.947589098532495,
      "grad_norm": 2.520841360092163,
      "learning_rate": 9.107928311397817e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 18646072,
      "step": 28520
    },
    {
      "epoch": 14.95020964360587,
      "grad_norm": 1.5904268026351929,
      "learning_rate": 9.09910327404785e-06,
      "loss": 0.0135,
      "num_input_tokens_seen": 18648696,
      "step": 28525
    },
    {
      "epoch": 14.952830188679245,
      "grad_norm": 10.025676727294922,
      "learning_rate": 9.090281562984068e-06,
      "loss": 0.0233,
      "num_input_tokens_seen": 18651608,
      "step": 28530
    },
    {
      "epoch": 14.95545073375262,
      "grad_norm": 0.3531889319419861,
      "learning_rate": 9.081463180051883e-06,
      "loss": 0.0224,
      "num_input_tokens_seen": 18655064,
      "step": 28535
    },
    {
      "epoch": 14.958071278825996,
      "grad_norm": 2.1177639961242676,
      "learning_rate": 9.072648127095996e-06,
      "loss": 0.0181,
      "num_input_tokens_seen": 18658040,
      "step": 28540
    },
    {
      "epoch": 14.96069182389937,
      "grad_norm": 2.689056634902954,
      "learning_rate": 9.0638364059604e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 18660472,
      "step": 28545
    },
    {
      "epoch": 14.963312368972746,
      "grad_norm": 0.329384982585907,
      "learning_rate": 9.055028018488417e-06,
      "loss": 0.0193,
      "num_input_tokens_seen": 18664216,
      "step": 28550
    },
    {
      "epoch": 14.965932914046121,
      "grad_norm": 8.259618759155273,
      "learning_rate": 9.046222966522665e-06,
      "loss": 0.0158,
      "num_input_tokens_seen": 18667224,
      "step": 28555
    },
    {
      "epoch": 14.968553459119496,
      "grad_norm": 0.37917423248291016,
      "learning_rate": 9.037421251905059e-06,
      "loss": 0.0265,
      "num_input_tokens_seen": 18670008,
      "step": 28560
    },
    {
      "epoch": 14.971174004192871,
      "grad_norm": 2.391714096069336,
      "learning_rate": 9.028622876476833e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 18672888,
      "step": 28565
    },
    {
      "epoch": 14.973794549266248,
      "grad_norm": 0.8858758211135864,
      "learning_rate": 9.019827842078474e-06,
      "loss": 0.015,
      "num_input_tokens_seen": 18676440,
      "step": 28570
    },
    {
      "epoch": 14.976415094339622,
      "grad_norm": 11.69606876373291,
      "learning_rate": 9.011036150549817e-06,
      "loss": 0.1736,
      "num_input_tokens_seen": 18679480,
      "step": 28575
    },
    {
      "epoch": 14.979035639412999,
      "grad_norm": 0.9434432983398438,
      "learning_rate": 9.002247803729987e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 18683096,
      "step": 28580
    },
    {
      "epoch": 14.981656184486374,
      "grad_norm": 4.414154529571533,
      "learning_rate": 8.993462803457404e-06,
      "loss": 0.0316,
      "num_input_tokens_seen": 18685304,
      "step": 28585
    },
    {
      "epoch": 14.984276729559749,
      "grad_norm": 1.5358695983886719,
      "learning_rate": 8.984681151569799e-06,
      "loss": 0.0242,
      "num_input_tokens_seen": 18687640,
      "step": 28590
    },
    {
      "epoch": 14.986897274633124,
      "grad_norm": 0.8015395402908325,
      "learning_rate": 8.975902849904183e-06,
      "loss": 0.0157,
      "num_input_tokens_seen": 18691832,
      "step": 28595
    },
    {
      "epoch": 14.9895178197065,
      "grad_norm": 0.691697895526886,
      "learning_rate": 8.967127900296874e-06,
      "loss": 0.038,
      "num_input_tokens_seen": 18694744,
      "step": 28600
    },
    {
      "epoch": 14.992138364779874,
      "grad_norm": 1.9739075899124146,
      "learning_rate": 8.95835630458349e-06,
      "loss": 0.0366,
      "num_input_tokens_seen": 18699096,
      "step": 28605
    },
    {
      "epoch": 14.99475890985325,
      "grad_norm": 1.8816375732421875,
      "learning_rate": 8.949588064598958e-06,
      "loss": 0.0295,
      "num_input_tokens_seen": 18702168,
      "step": 28610
    },
    {
      "epoch": 14.997379454926625,
      "grad_norm": 0.40476444363594055,
      "learning_rate": 8.940823182177499e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 18704920,
      "step": 28615
    },
    {
      "epoch": 15.0,
      "grad_norm": 1.991878628730774,
      "learning_rate": 8.932061659152607e-06,
      "loss": 0.0165,
      "num_input_tokens_seen": 18706976,
      "step": 28620
    },
    {
      "epoch": 15.0,
      "eval_loss": 1.3870089054107666,
      "eval_runtime": 11.6984,
      "eval_samples_per_second": 72.489,
      "eval_steps_per_second": 18.122,
      "num_input_tokens_seen": 18706976,
      "step": 28620
    },
    {
      "epoch": 15.002620545073375,
      "grad_norm": 0.4931572675704956,
      "learning_rate": 8.92330349735711e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 18710720,
      "step": 28625
    },
    {
      "epoch": 15.00524109014675,
      "grad_norm": 0.6515786051750183,
      "learning_rate": 8.914548698623104e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 18714528,
      "step": 28630
    },
    {
      "epoch": 15.007861635220126,
      "grad_norm": 0.9285647869110107,
      "learning_rate": 8.905797264781995e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 18717120,
      "step": 28635
    },
    {
      "epoch": 15.0104821802935,
      "grad_norm": 0.5981734991073608,
      "learning_rate": 8.89704919766449e-06,
      "loss": 0.0232,
      "num_input_tokens_seen": 18719776,
      "step": 28640
    },
    {
      "epoch": 15.013102725366876,
      "grad_norm": 0.5951062440872192,
      "learning_rate": 8.888304499100586e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 18722976,
      "step": 28645
    },
    {
      "epoch": 15.015723270440251,
      "grad_norm": 0.9045798182487488,
      "learning_rate": 8.879563170919566e-06,
      "loss": 0.0273,
      "num_input_tokens_seen": 18726432,
      "step": 28650
    },
    {
      "epoch": 15.018343815513626,
      "grad_norm": 2.697599172592163,
      "learning_rate": 8.870825214950016e-06,
      "loss": 0.0156,
      "num_input_tokens_seen": 18729440,
      "step": 28655
    },
    {
      "epoch": 15.020964360587001,
      "grad_norm": 0.5712617635726929,
      "learning_rate": 8.862090633019818e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 18732256,
      "step": 28660
    },
    {
      "epoch": 15.023584905660377,
      "grad_norm": 2.361072301864624,
      "learning_rate": 8.853359426956146e-06,
      "loss": 0.0387,
      "num_input_tokens_seen": 18736480,
      "step": 28665
    },
    {
      "epoch": 15.026205450733753,
      "grad_norm": 0.6217346787452698,
      "learning_rate": 8.844631598585482e-06,
      "loss": 0.0148,
      "num_input_tokens_seen": 18739200,
      "step": 28670
    },
    {
      "epoch": 15.028825995807129,
      "grad_norm": 1.8595514297485352,
      "learning_rate": 8.835907149733569e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 18741536,
      "step": 28675
    },
    {
      "epoch": 15.031446540880504,
      "grad_norm": 0.3962751626968384,
      "learning_rate": 8.827186082225477e-06,
      "loss": 0.0223,
      "num_input_tokens_seen": 18744672,
      "step": 28680
    },
    {
      "epoch": 15.034067085953879,
      "grad_norm": 1.289770245552063,
      "learning_rate": 8.81846839788554e-06,
      "loss": 0.0291,
      "num_input_tokens_seen": 18747904,
      "step": 28685
    },
    {
      "epoch": 15.036687631027254,
      "grad_norm": 1.2345205545425415,
      "learning_rate": 8.809754098537407e-06,
      "loss": 0.0139,
      "num_input_tokens_seen": 18752000,
      "step": 28690
    },
    {
      "epoch": 15.03930817610063,
      "grad_norm": 0.7258554100990295,
      "learning_rate": 8.801043186004016e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 18755616,
      "step": 28695
    },
    {
      "epoch": 15.041928721174004,
      "grad_norm": 0.399769127368927,
      "learning_rate": 8.792335662107575e-06,
      "loss": 0.013,
      "num_input_tokens_seen": 18758240,
      "step": 28700
    },
    {
      "epoch": 15.04454926624738,
      "grad_norm": 0.7068812251091003,
      "learning_rate": 8.783631528669605e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 18762368,
      "step": 28705
    },
    {
      "epoch": 15.047169811320755,
      "grad_norm": 0.3478952646255493,
      "learning_rate": 8.774930787510924e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 18765152,
      "step": 28710
    },
    {
      "epoch": 15.04979035639413,
      "grad_norm": 1.3808623552322388,
      "learning_rate": 8.766233440451608e-06,
      "loss": 0.0382,
      "num_input_tokens_seen": 18768928,
      "step": 28715
    },
    {
      "epoch": 15.052410901467505,
      "grad_norm": 0.24884432554244995,
      "learning_rate": 8.75753948931105e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 18771488,
      "step": 28720
    },
    {
      "epoch": 15.05503144654088,
      "grad_norm": 1.010399341583252,
      "learning_rate": 8.748848935907936e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 18774272,
      "step": 28725
    },
    {
      "epoch": 15.057651991614255,
      "grad_norm": 0.47566500306129456,
      "learning_rate": 8.740161782060214e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 18776704,
      "step": 28730
    },
    {
      "epoch": 15.06027253668763,
      "grad_norm": 0.5498323440551758,
      "learning_rate": 8.731478029585144e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 18779520,
      "step": 28735
    },
    {
      "epoch": 15.062893081761006,
      "grad_norm": 0.10696715116500854,
      "learning_rate": 8.722797680299278e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 18782336,
      "step": 28740
    },
    {
      "epoch": 15.065513626834381,
      "grad_norm": 0.9343944191932678,
      "learning_rate": 8.71412073601843e-06,
      "loss": 0.0182,
      "num_input_tokens_seen": 18785504,
      "step": 28745
    },
    {
      "epoch": 15.068134171907756,
      "grad_norm": 3.3354077339172363,
      "learning_rate": 8.705447198557731e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 18788480,
      "step": 28750
    },
    {
      "epoch": 15.070754716981131,
      "grad_norm": 1.462590217590332,
      "learning_rate": 8.696777069731574e-06,
      "loss": 0.0527,
      "num_input_tokens_seen": 18794144,
      "step": 28755
    },
    {
      "epoch": 15.073375262054507,
      "grad_norm": 0.4575968384742737,
      "learning_rate": 8.688110351353654e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 18797376,
      "step": 28760
    },
    {
      "epoch": 15.075995807127883,
      "grad_norm": 0.8578583598136902,
      "learning_rate": 8.679447045236962e-06,
      "loss": 0.0258,
      "num_input_tokens_seen": 18800672,
      "step": 28765
    },
    {
      "epoch": 15.078616352201259,
      "grad_norm": 0.3559117019176483,
      "learning_rate": 8.670787153193746e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 18803648,
      "step": 28770
    },
    {
      "epoch": 15.081236897274634,
      "grad_norm": 2.5334887504577637,
      "learning_rate": 8.662130677035574e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 18806144,
      "step": 28775
    },
    {
      "epoch": 15.083857442348009,
      "grad_norm": 0.8742836117744446,
      "learning_rate": 8.65347761857326e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 18809152,
      "step": 28780
    },
    {
      "epoch": 15.086477987421384,
      "grad_norm": 1.0878355503082275,
      "learning_rate": 8.64482797961694e-06,
      "loss": 0.0106,
      "num_input_tokens_seen": 18812192,
      "step": 28785
    },
    {
      "epoch": 15.08909853249476,
      "grad_norm": 5.63104772567749,
      "learning_rate": 8.636181761976016e-06,
      "loss": 0.025,
      "num_input_tokens_seen": 18815168,
      "step": 28790
    },
    {
      "epoch": 15.091719077568134,
      "grad_norm": 0.9097110629081726,
      "learning_rate": 8.62753896745919e-06,
      "loss": 0.0222,
      "num_input_tokens_seen": 18819072,
      "step": 28795
    },
    {
      "epoch": 15.09433962264151,
      "grad_norm": 0.6623551845550537,
      "learning_rate": 8.618899597874413e-06,
      "loss": 0.0133,
      "num_input_tokens_seen": 18821728,
      "step": 28800
    },
    {
      "epoch": 15.096960167714885,
      "grad_norm": 1.7514939308166504,
      "learning_rate": 8.610263655028964e-06,
      "loss": 0.018,
      "num_input_tokens_seen": 18824544,
      "step": 28805
    },
    {
      "epoch": 15.09958071278826,
      "grad_norm": 2.362037181854248,
      "learning_rate": 8.601631140729366e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 18828032,
      "step": 28810
    },
    {
      "epoch": 15.102201257861635,
      "grad_norm": 0.8407651782035828,
      "learning_rate": 8.593002056781451e-06,
      "loss": 0.016,
      "num_input_tokens_seen": 18830688,
      "step": 28815
    },
    {
      "epoch": 15.10482180293501,
      "grad_norm": 3.2607438564300537,
      "learning_rate": 8.584376404990326e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 18833664,
      "step": 28820
    },
    {
      "epoch": 15.107442348008385,
      "grad_norm": 0.778249204158783,
      "learning_rate": 8.575754187160384e-06,
      "loss": 0.0107,
      "num_input_tokens_seen": 18837536,
      "step": 28825
    },
    {
      "epoch": 15.11006289308176,
      "grad_norm": 1.6024012565612793,
      "learning_rate": 8.567135405095291e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 18840000,
      "step": 28830
    },
    {
      "epoch": 15.112683438155136,
      "grad_norm": 1.189666509628296,
      "learning_rate": 8.558520060597985e-06,
      "loss": 0.023,
      "num_input_tokens_seen": 18844224,
      "step": 28835
    },
    {
      "epoch": 15.115303983228511,
      "grad_norm": 1.4897849559783936,
      "learning_rate": 8.54990815547071e-06,
      "loss": 0.0129,
      "num_input_tokens_seen": 18847104,
      "step": 28840
    },
    {
      "epoch": 15.117924528301886,
      "grad_norm": 4.571020126342773,
      "learning_rate": 8.541299691514974e-06,
      "loss": 0.0564,
      "num_input_tokens_seen": 18849536,
      "step": 28845
    },
    {
      "epoch": 15.120545073375261,
      "grad_norm": 0.24637222290039062,
      "learning_rate": 8.53269467053158e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 18854528,
      "step": 28850
    },
    {
      "epoch": 15.123165618448636,
      "grad_norm": 0.3806881904602051,
      "learning_rate": 8.524093094320593e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 18857344,
      "step": 28855
    },
    {
      "epoch": 15.125786163522013,
      "grad_norm": 2.778533458709717,
      "learning_rate": 8.515494964681357e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 18859872,
      "step": 28860
    },
    {
      "epoch": 15.128406708595389,
      "grad_norm": 0.1796269565820694,
      "learning_rate": 8.506900283412506e-06,
      "loss": 0.0037,
      "num_input_tokens_seen": 18862240,
      "step": 28865
    },
    {
      "epoch": 15.131027253668764,
      "grad_norm": 0.3499109148979187,
      "learning_rate": 8.498309052311953e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 18865472,
      "step": 28870
    },
    {
      "epoch": 15.133647798742139,
      "grad_norm": 0.5905930995941162,
      "learning_rate": 8.489721273176887e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 18868960,
      "step": 28875
    },
    {
      "epoch": 15.136268343815514,
      "grad_norm": 0.5020491480827332,
      "learning_rate": 8.481136947803777e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 18871616,
      "step": 28880
    },
    {
      "epoch": 15.13888888888889,
      "grad_norm": 4.190843105316162,
      "learning_rate": 8.472556077988362e-06,
      "loss": 0.0254,
      "num_input_tokens_seen": 18875040,
      "step": 28885
    },
    {
      "epoch": 15.141509433962264,
      "grad_norm": 5.123723030090332,
      "learning_rate": 8.46397866552565e-06,
      "loss": 0.021,
      "num_input_tokens_seen": 18877632,
      "step": 28890
    },
    {
      "epoch": 15.14412997903564,
      "grad_norm": 18.618492126464844,
      "learning_rate": 8.455404712209952e-06,
      "loss": 0.0358,
      "num_input_tokens_seen": 18881056,
      "step": 28895
    },
    {
      "epoch": 15.146750524109015,
      "grad_norm": 0.7548385262489319,
      "learning_rate": 8.446834219834836e-06,
      "loss": 0.0191,
      "num_input_tokens_seen": 18883616,
      "step": 28900
    },
    {
      "epoch": 15.14937106918239,
      "grad_norm": 0.37178581953048706,
      "learning_rate": 8.43826719019316e-06,
      "loss": 0.0037,
      "num_input_tokens_seen": 18886592,
      "step": 28905
    },
    {
      "epoch": 15.151991614255765,
      "grad_norm": 0.8978719711303711,
      "learning_rate": 8.429703625077043e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 18891104,
      "step": 28910
    },
    {
      "epoch": 15.15461215932914,
      "grad_norm": 0.10958582162857056,
      "learning_rate": 8.421143526277875e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 18894528,
      "step": 28915
    },
    {
      "epoch": 15.157232704402515,
      "grad_norm": 2.044553756713867,
      "learning_rate": 8.41258689558634e-06,
      "loss": 0.021,
      "num_input_tokens_seen": 18897760,
      "step": 28920
    },
    {
      "epoch": 15.15985324947589,
      "grad_norm": 0.2278500199317932,
      "learning_rate": 8.404033734792386e-06,
      "loss": 0.0149,
      "num_input_tokens_seen": 18900800,
      "step": 28925
    },
    {
      "epoch": 15.162473794549266,
      "grad_norm": 0.49623334407806396,
      "learning_rate": 8.395484045685248e-06,
      "loss": 0.023,
      "num_input_tokens_seen": 18903680,
      "step": 28930
    },
    {
      "epoch": 15.165094339622641,
      "grad_norm": 0.5353211164474487,
      "learning_rate": 8.386937830053412e-06,
      "loss": 0.0091,
      "num_input_tokens_seen": 18906336,
      "step": 28935
    },
    {
      "epoch": 15.167714884696016,
      "grad_norm": 0.4550411105155945,
      "learning_rate": 8.37839508968464e-06,
      "loss": 0.0184,
      "num_input_tokens_seen": 18910112,
      "step": 28940
    },
    {
      "epoch": 15.170335429769391,
      "grad_norm": 2.8071370124816895,
      "learning_rate": 8.369855826365988e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 18912928,
      "step": 28945
    },
    {
      "epoch": 15.172955974842766,
      "grad_norm": 0.9452383518218994,
      "learning_rate": 8.361320041883772e-06,
      "loss": 0.0129,
      "num_input_tokens_seen": 18916416,
      "step": 28950
    },
    {
      "epoch": 15.175576519916143,
      "grad_norm": 1.1033235788345337,
      "learning_rate": 8.352787738023576e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 18919968,
      "step": 28955
    },
    {
      "epoch": 15.178197064989519,
      "grad_norm": 0.8040234446525574,
      "learning_rate": 8.344258916570274e-06,
      "loss": 0.0158,
      "num_input_tokens_seen": 18923392,
      "step": 28960
    },
    {
      "epoch": 15.180817610062894,
      "grad_norm": 0.7627437710762024,
      "learning_rate": 8.335733579307988e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 18926528,
      "step": 28965
    },
    {
      "epoch": 15.183438155136269,
      "grad_norm": 2.349487543106079,
      "learning_rate": 8.327211728020113e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 18929696,
      "step": 28970
    },
    {
      "epoch": 15.186058700209644,
      "grad_norm": 6.797382831573486,
      "learning_rate": 8.31869336448933e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 18932384,
      "step": 28975
    },
    {
      "epoch": 15.18867924528302,
      "grad_norm": 1.3287830352783203,
      "learning_rate": 8.310178490497586e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 18934880,
      "step": 28980
    },
    {
      "epoch": 15.191299790356394,
      "grad_norm": 0.3613591492176056,
      "learning_rate": 8.301667107826103e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 18938816,
      "step": 28985
    },
    {
      "epoch": 15.19392033542977,
      "grad_norm": 3.2478575706481934,
      "learning_rate": 8.293159218255345e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 18941824,
      "step": 28990
    },
    {
      "epoch": 15.196540880503145,
      "grad_norm": 0.1429366022348404,
      "learning_rate": 8.284654823565088e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 18945056,
      "step": 28995
    },
    {
      "epoch": 15.19916142557652,
      "grad_norm": 0.6122066974639893,
      "learning_rate": 8.276153925534333e-06,
      "loss": 0.0165,
      "num_input_tokens_seen": 18948256,
      "step": 29000
    },
    {
      "epoch": 15.201781970649895,
      "grad_norm": 1.7251880168914795,
      "learning_rate": 8.267656525941383e-06,
      "loss": 0.0201,
      "num_input_tokens_seen": 18951008,
      "step": 29005
    },
    {
      "epoch": 15.20440251572327,
      "grad_norm": 1.6567147970199585,
      "learning_rate": 8.259162626563801e-06,
      "loss": 0.0132,
      "num_input_tokens_seen": 18954464,
      "step": 29010
    },
    {
      "epoch": 15.207023060796645,
      "grad_norm": 0.6104302406311035,
      "learning_rate": 8.250672229178402e-06,
      "loss": 0.0148,
      "num_input_tokens_seen": 18958464,
      "step": 29015
    },
    {
      "epoch": 15.20964360587002,
      "grad_norm": 0.28400924801826477,
      "learning_rate": 8.242185335561295e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 18961152,
      "step": 29020
    },
    {
      "epoch": 15.212264150943396,
      "grad_norm": 0.9653488993644714,
      "learning_rate": 8.23370194748782e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 18964640,
      "step": 29025
    },
    {
      "epoch": 15.214884696016771,
      "grad_norm": 1.9867579936981201,
      "learning_rate": 8.22522206673262e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 18967616,
      "step": 29030
    },
    {
      "epoch": 15.217505241090146,
      "grad_norm": 2.0049972534179688,
      "learning_rate": 8.216745695069589e-06,
      "loss": 0.0375,
      "num_input_tokens_seen": 18971136,
      "step": 29035
    },
    {
      "epoch": 15.220125786163521,
      "grad_norm": 1.1295008659362793,
      "learning_rate": 8.208272834271894e-06,
      "loss": 0.0319,
      "num_input_tokens_seen": 18974208,
      "step": 29040
    },
    {
      "epoch": 15.222746331236896,
      "grad_norm": 0.7590684294700623,
      "learning_rate": 8.19980348611194e-06,
      "loss": 0.0107,
      "num_input_tokens_seen": 18977408,
      "step": 29045
    },
    {
      "epoch": 15.225366876310272,
      "grad_norm": 0.6644713878631592,
      "learning_rate": 8.191337652361439e-06,
      "loss": 0.0179,
      "num_input_tokens_seen": 18980288,
      "step": 29050
    },
    {
      "epoch": 15.227987421383649,
      "grad_norm": 3.914356231689453,
      "learning_rate": 8.182875334791332e-06,
      "loss": 0.0093,
      "num_input_tokens_seen": 18984096,
      "step": 29055
    },
    {
      "epoch": 15.230607966457024,
      "grad_norm": 0.055271416902542114,
      "learning_rate": 8.174416535171841e-06,
      "loss": 0.0346,
      "num_input_tokens_seen": 18987648,
      "step": 29060
    },
    {
      "epoch": 15.233228511530399,
      "grad_norm": 1.4557019472122192,
      "learning_rate": 8.165961255272467e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 18990208,
      "step": 29065
    },
    {
      "epoch": 15.235849056603774,
      "grad_norm": 0.4282313883304596,
      "learning_rate": 8.157509496861935e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 18993440,
      "step": 29070
    },
    {
      "epoch": 15.23846960167715,
      "grad_norm": 0.7987397909164429,
      "learning_rate": 8.149061261708266e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 18997024,
      "step": 29075
    },
    {
      "epoch": 15.241090146750524,
      "grad_norm": 0.5654147267341614,
      "learning_rate": 8.140616551578745e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 19000864,
      "step": 29080
    },
    {
      "epoch": 15.2437106918239,
      "grad_norm": 3.322719097137451,
      "learning_rate": 8.13217536823989e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 19004576,
      "step": 29085
    },
    {
      "epoch": 15.246331236897275,
      "grad_norm": 0.6031779646873474,
      "learning_rate": 8.12373771345752e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 19008128,
      "step": 29090
    },
    {
      "epoch": 15.24895178197065,
      "grad_norm": 0.2173810452222824,
      "learning_rate": 8.115303588996676e-06,
      "loss": 0.0323,
      "num_input_tokens_seen": 19013952,
      "step": 29095
    },
    {
      "epoch": 15.251572327044025,
      "grad_norm": 1.122568964958191,
      "learning_rate": 8.10687299662169e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 19017216,
      "step": 29100
    },
    {
      "epoch": 15.2541928721174,
      "grad_norm": 0.5390076041221619,
      "learning_rate": 8.098445938096147e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 19020096,
      "step": 29105
    },
    {
      "epoch": 15.256813417190775,
      "grad_norm": 15.671405792236328,
      "learning_rate": 8.090022415182898e-06,
      "loss": 0.0132,
      "num_input_tokens_seen": 19023456,
      "step": 29110
    },
    {
      "epoch": 15.25943396226415,
      "grad_norm": 0.8195925951004028,
      "learning_rate": 8.081602429644033e-06,
      "loss": 0.017,
      "num_input_tokens_seen": 19027424,
      "step": 29115
    },
    {
      "epoch": 15.262054507337526,
      "grad_norm": 2.124767541885376,
      "learning_rate": 8.073185983240932e-06,
      "loss": 0.0218,
      "num_input_tokens_seen": 19030976,
      "step": 29120
    },
    {
      "epoch": 15.264675052410901,
      "grad_norm": 1.9043309688568115,
      "learning_rate": 8.064773077734206e-06,
      "loss": 0.0127,
      "num_input_tokens_seen": 19034976,
      "step": 29125
    },
    {
      "epoch": 15.267295597484276,
      "grad_norm": 0.9025167226791382,
      "learning_rate": 8.05636371488374e-06,
      "loss": 0.0149,
      "num_input_tokens_seen": 19038752,
      "step": 29130
    },
    {
      "epoch": 15.269916142557651,
      "grad_norm": 0.7982821464538574,
      "learning_rate": 8.047957896448696e-06,
      "loss": 0.0206,
      "num_input_tokens_seen": 19041536,
      "step": 29135
    },
    {
      "epoch": 15.272536687631026,
      "grad_norm": 0.6989299654960632,
      "learning_rate": 8.039555624187451e-06,
      "loss": 0.0128,
      "num_input_tokens_seen": 19044448,
      "step": 29140
    },
    {
      "epoch": 15.275157232704402,
      "grad_norm": 0.4421520233154297,
      "learning_rate": 8.031156899857681e-06,
      "loss": 0.0254,
      "num_input_tokens_seen": 19047904,
      "step": 29145
    },
    {
      "epoch": 15.277777777777779,
      "grad_norm": 1.130603551864624,
      "learning_rate": 8.022761725216288e-06,
      "loss": 0.0245,
      "num_input_tokens_seen": 19052864,
      "step": 29150
    },
    {
      "epoch": 15.280398322851154,
      "grad_norm": 1.3360081911087036,
      "learning_rate": 8.014370102019456e-06,
      "loss": 0.0747,
      "num_input_tokens_seen": 19055456,
      "step": 29155
    },
    {
      "epoch": 15.283018867924529,
      "grad_norm": 0.47602906823158264,
      "learning_rate": 8.005982032022616e-06,
      "loss": 0.0127,
      "num_input_tokens_seen": 19058240,
      "step": 29160
    },
    {
      "epoch": 15.285639412997904,
      "grad_norm": 1.1271929740905762,
      "learning_rate": 7.997597516980467e-06,
      "loss": 0.0127,
      "num_input_tokens_seen": 19062336,
      "step": 29165
    },
    {
      "epoch": 15.28825995807128,
      "grad_norm": 0.4183765947818756,
      "learning_rate": 7.989216558646942e-06,
      "loss": 0.0088,
      "num_input_tokens_seen": 19066592,
      "step": 29170
    },
    {
      "epoch": 15.290880503144654,
      "grad_norm": 1.3040310144424438,
      "learning_rate": 7.980839158775235e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 19069888,
      "step": 29175
    },
    {
      "epoch": 15.29350104821803,
      "grad_norm": 3.391590118408203,
      "learning_rate": 7.972465319117814e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 19073152,
      "step": 29180
    },
    {
      "epoch": 15.296121593291405,
      "grad_norm": 0.20732048153877258,
      "learning_rate": 7.964095041426386e-06,
      "loss": 0.0166,
      "num_input_tokens_seen": 19075904,
      "step": 29185
    },
    {
      "epoch": 15.29874213836478,
      "grad_norm": 0.3589511811733246,
      "learning_rate": 7.95572832745192e-06,
      "loss": 0.0657,
      "num_input_tokens_seen": 19079584,
      "step": 29190
    },
    {
      "epoch": 15.301362683438155,
      "grad_norm": 1.3459362983703613,
      "learning_rate": 7.947365178944643e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 19082656,
      "step": 29195
    },
    {
      "epoch": 15.30398322851153,
      "grad_norm": 1.1572710275650024,
      "learning_rate": 7.939005597654025e-06,
      "loss": 0.0143,
      "num_input_tokens_seen": 19086208,
      "step": 29200
    },
    {
      "epoch": 15.306603773584905,
      "grad_norm": 0.4790988862514496,
      "learning_rate": 7.930649585328787e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 19089184,
      "step": 29205
    },
    {
      "epoch": 15.30922431865828,
      "grad_norm": 0.4643336832523346,
      "learning_rate": 7.922297143716918e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 19093664,
      "step": 29210
    },
    {
      "epoch": 15.311844863731656,
      "grad_norm": 0.5314231514930725,
      "learning_rate": 7.913948274565652e-06,
      "loss": 0.0123,
      "num_input_tokens_seen": 19098048,
      "step": 29215
    },
    {
      "epoch": 15.314465408805031,
      "grad_norm": 1.1409401893615723,
      "learning_rate": 7.905602979621491e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 19101760,
      "step": 29220
    },
    {
      "epoch": 15.317085953878406,
      "grad_norm": 1.6438169479370117,
      "learning_rate": 7.897261260630159e-06,
      "loss": 0.0186,
      "num_input_tokens_seen": 19106912,
      "step": 29225
    },
    {
      "epoch": 15.319706498951781,
      "grad_norm": 1.120215892791748,
      "learning_rate": 7.888923119336647e-06,
      "loss": 0.0106,
      "num_input_tokens_seen": 19110272,
      "step": 29230
    },
    {
      "epoch": 15.322327044025156,
      "grad_norm": 0.7313971519470215,
      "learning_rate": 7.880588557485203e-06,
      "loss": 0.0259,
      "num_input_tokens_seen": 19113408,
      "step": 29235
    },
    {
      "epoch": 15.324947589098532,
      "grad_norm": 1.7698688507080078,
      "learning_rate": 7.872257576819325e-06,
      "loss": 0.0199,
      "num_input_tokens_seen": 19116864,
      "step": 29240
    },
    {
      "epoch": 15.327568134171909,
      "grad_norm": 1.4751654863357544,
      "learning_rate": 7.863930179081769e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 19120576,
      "step": 29245
    },
    {
      "epoch": 15.330188679245284,
      "grad_norm": 11.352205276489258,
      "learning_rate": 7.855606366014517e-06,
      "loss": 0.022,
      "num_input_tokens_seen": 19123872,
      "step": 29250
    },
    {
      "epoch": 15.332809224318659,
      "grad_norm": 0.16140131652355194,
      "learning_rate": 7.847286139358814e-06,
      "loss": 0.0112,
      "num_input_tokens_seen": 19126688,
      "step": 29255
    },
    {
      "epoch": 15.335429769392034,
      "grad_norm": 1.336893916130066,
      "learning_rate": 7.838969500855162e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 19129504,
      "step": 29260
    },
    {
      "epoch": 15.33805031446541,
      "grad_norm": 0.4130761921405792,
      "learning_rate": 7.830656452243307e-06,
      "loss": 0.0145,
      "num_input_tokens_seen": 19132064,
      "step": 29265
    },
    {
      "epoch": 15.340670859538784,
      "grad_norm": 0.27276721596717834,
      "learning_rate": 7.822346995262241e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 19134848,
      "step": 29270
    },
    {
      "epoch": 15.34329140461216,
      "grad_norm": 1.9402790069580078,
      "learning_rate": 7.814041131650223e-06,
      "loss": 0.0194,
      "num_input_tokens_seen": 19137632,
      "step": 29275
    },
    {
      "epoch": 15.345911949685535,
      "grad_norm": 0.21174445748329163,
      "learning_rate": 7.805738863144731e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 19140672,
      "step": 29280
    },
    {
      "epoch": 15.34853249475891,
      "grad_norm": 0.8632181882858276,
      "learning_rate": 7.797440191482496e-06,
      "loss": 0.0144,
      "num_input_tokens_seen": 19144192,
      "step": 29285
    },
    {
      "epoch": 15.351153039832285,
      "grad_norm": 1.2089459896087646,
      "learning_rate": 7.789145118399518e-06,
      "loss": 0.0175,
      "num_input_tokens_seen": 19146880,
      "step": 29290
    },
    {
      "epoch": 15.35377358490566,
      "grad_norm": 2.265148878097534,
      "learning_rate": 7.780853645631032e-06,
      "loss": 0.0151,
      "num_input_tokens_seen": 19150400,
      "step": 29295
    },
    {
      "epoch": 15.356394129979035,
      "grad_norm": 0.6977131366729736,
      "learning_rate": 7.772565774911522e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 19153504,
      "step": 29300
    },
    {
      "epoch": 15.35901467505241,
      "grad_norm": 2.3506274223327637,
      "learning_rate": 7.76428150797471e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 19156960,
      "step": 29305
    },
    {
      "epoch": 15.361635220125786,
      "grad_norm": 2.420147180557251,
      "learning_rate": 7.756000846553562e-06,
      "loss": 0.0856,
      "num_input_tokens_seen": 19159744,
      "step": 29310
    },
    {
      "epoch": 15.364255765199161,
      "grad_norm": 0.5921143293380737,
      "learning_rate": 7.74772379238031e-06,
      "loss": 0.0094,
      "num_input_tokens_seen": 19163808,
      "step": 29315
    },
    {
      "epoch": 15.366876310272536,
      "grad_norm": 2.1096255779266357,
      "learning_rate": 7.739450347186417e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 19166304,
      "step": 29320
    },
    {
      "epoch": 15.369496855345911,
      "grad_norm": 0.8670904040336609,
      "learning_rate": 7.731180512702599e-06,
      "loss": 0.029,
      "num_input_tokens_seen": 19169792,
      "step": 29325
    },
    {
      "epoch": 15.372117400419286,
      "grad_norm": 0.8416690826416016,
      "learning_rate": 7.722914290658795e-06,
      "loss": 0.0154,
      "num_input_tokens_seen": 19172448,
      "step": 29330
    },
    {
      "epoch": 15.374737945492662,
      "grad_norm": 3.304089307785034,
      "learning_rate": 7.714651682784224e-06,
      "loss": 0.0112,
      "num_input_tokens_seen": 19176352,
      "step": 29335
    },
    {
      "epoch": 15.377358490566039,
      "grad_norm": 0.42580342292785645,
      "learning_rate": 7.706392690807313e-06,
      "loss": 0.022,
      "num_input_tokens_seen": 19179808,
      "step": 29340
    },
    {
      "epoch": 15.379979035639414,
      "grad_norm": 0.596844494342804,
      "learning_rate": 7.69813731645575e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 19183872,
      "step": 29345
    },
    {
      "epoch": 15.382599580712789,
      "grad_norm": 0.9072333574295044,
      "learning_rate": 7.689885561456475e-06,
      "loss": 0.0203,
      "num_input_tokens_seen": 19187584,
      "step": 29350
    },
    {
      "epoch": 15.385220125786164,
      "grad_norm": 1.215686559677124,
      "learning_rate": 7.681637427535663e-06,
      "loss": 0.0114,
      "num_input_tokens_seen": 19192320,
      "step": 29355
    },
    {
      "epoch": 15.38784067085954,
      "grad_norm": 1.7767114639282227,
      "learning_rate": 7.673392916418715e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 19195712,
      "step": 29360
    },
    {
      "epoch": 15.390461215932914,
      "grad_norm": 0.6122862696647644,
      "learning_rate": 7.665152029830303e-06,
      "loss": 0.016,
      "num_input_tokens_seen": 19199328,
      "step": 29365
    },
    {
      "epoch": 15.39308176100629,
      "grad_norm": 0.7902641296386719,
      "learning_rate": 7.656914769494314e-06,
      "loss": 0.0276,
      "num_input_tokens_seen": 19202368,
      "step": 29370
    },
    {
      "epoch": 15.395702306079665,
      "grad_norm": 1.2209373712539673,
      "learning_rate": 7.648681137133892e-06,
      "loss": 0.0236,
      "num_input_tokens_seen": 19205728,
      "step": 29375
    },
    {
      "epoch": 15.39832285115304,
      "grad_norm": 0.4847785234451294,
      "learning_rate": 7.640451134471432e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 19208768,
      "step": 29380
    },
    {
      "epoch": 15.400943396226415,
      "grad_norm": 1.4399775266647339,
      "learning_rate": 7.632224763228538e-06,
      "loss": 0.0196,
      "num_input_tokens_seen": 19213152,
      "step": 29385
    },
    {
      "epoch": 15.40356394129979,
      "grad_norm": 0.12238199263811111,
      "learning_rate": 7.6240020251260896e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 19215392,
      "step": 29390
    },
    {
      "epoch": 15.406184486373165,
      "grad_norm": 1.114962100982666,
      "learning_rate": 7.615782921884174e-06,
      "loss": 0.0176,
      "num_input_tokens_seen": 19218944,
      "step": 29395
    },
    {
      "epoch": 15.40880503144654,
      "grad_norm": 0.44910040497779846,
      "learning_rate": 7.607567455222142e-06,
      "loss": 0.0678,
      "num_input_tokens_seen": 19221984,
      "step": 29400
    },
    {
      "epoch": 15.411425576519916,
      "grad_norm": 1.0342458486557007,
      "learning_rate": 7.599355626858582e-06,
      "loss": 0.1259,
      "num_input_tokens_seen": 19225376,
      "step": 29405
    },
    {
      "epoch": 15.414046121593291,
      "grad_norm": 0.41025179624557495,
      "learning_rate": 7.591147438511298e-06,
      "loss": 0.0101,
      "num_input_tokens_seen": 19228448,
      "step": 29410
    },
    {
      "epoch": 15.416666666666666,
      "grad_norm": 0.6417186856269836,
      "learning_rate": 7.5829428918973625e-06,
      "loss": 0.2129,
      "num_input_tokens_seen": 19231712,
      "step": 29415
    },
    {
      "epoch": 15.419287211740041,
      "grad_norm": 1.5497245788574219,
      "learning_rate": 7.574741988733075e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 19234144,
      "step": 29420
    },
    {
      "epoch": 15.421907756813416,
      "grad_norm": 0.8378811478614807,
      "learning_rate": 7.56654473073396e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 19237056,
      "step": 29425
    },
    {
      "epoch": 15.424528301886792,
      "grad_norm": 1.2392268180847168,
      "learning_rate": 7.558351119614796e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 19239904,
      "step": 29430
    },
    {
      "epoch": 15.427148846960169,
      "grad_norm": 0.4021016061306,
      "learning_rate": 7.5501611570896e-06,
      "loss": 0.0135,
      "num_input_tokens_seen": 19243552,
      "step": 29435
    },
    {
      "epoch": 15.429769392033544,
      "grad_norm": 0.11255314946174622,
      "learning_rate": 7.541974844871602e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 19246560,
      "step": 29440
    },
    {
      "epoch": 15.432389937106919,
      "grad_norm": 2.429370164871216,
      "learning_rate": 7.5337921846732965e-06,
      "loss": 0.0256,
      "num_input_tokens_seen": 19249632,
      "step": 29445
    },
    {
      "epoch": 15.435010482180294,
      "grad_norm": 0.6332119703292847,
      "learning_rate": 7.525613178206409e-06,
      "loss": 0.0134,
      "num_input_tokens_seen": 19252736,
      "step": 29450
    },
    {
      "epoch": 15.43763102725367,
      "grad_norm": 0.707423210144043,
      "learning_rate": 7.517437827181878e-06,
      "loss": 0.0113,
      "num_input_tokens_seen": 19255968,
      "step": 29455
    },
    {
      "epoch": 15.440251572327044,
      "grad_norm": 0.14119547605514526,
      "learning_rate": 7.509266133309908e-06,
      "loss": 0.0304,
      "num_input_tokens_seen": 19259680,
      "step": 29460
    },
    {
      "epoch": 15.44287211740042,
      "grad_norm": 0.22319240868091583,
      "learning_rate": 7.501098098299914e-06,
      "loss": 0.0096,
      "num_input_tokens_seen": 19262848,
      "step": 29465
    },
    {
      "epoch": 15.445492662473795,
      "grad_norm": 1.4391851425170898,
      "learning_rate": 7.492933723860557e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 19265536,
      "step": 29470
    },
    {
      "epoch": 15.44811320754717,
      "grad_norm": 0.6789968013763428,
      "learning_rate": 7.4847730116997334e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 19268384,
      "step": 29475
    },
    {
      "epoch": 15.450733752620545,
      "grad_norm": 1.5259854793548584,
      "learning_rate": 7.476615963524583e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 19272000,
      "step": 29480
    },
    {
      "epoch": 15.45335429769392,
      "grad_norm": 1.4065091609954834,
      "learning_rate": 7.468462581041452e-06,
      "loss": 0.0248,
      "num_input_tokens_seen": 19275360,
      "step": 29485
    },
    {
      "epoch": 15.455974842767295,
      "grad_norm": 2.147346019744873,
      "learning_rate": 7.460312865955932e-06,
      "loss": 0.0237,
      "num_input_tokens_seen": 19279488,
      "step": 29490
    },
    {
      "epoch": 15.45859538784067,
      "grad_norm": 0.13899973034858704,
      "learning_rate": 7.4521668199728584e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 19282176,
      "step": 29495
    },
    {
      "epoch": 15.461215932914046,
      "grad_norm": 0.5493974685668945,
      "learning_rate": 7.4440244447962884e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 19285056,
      "step": 29500
    },
    {
      "epoch": 15.463836477987421,
      "grad_norm": 0.7452818155288696,
      "learning_rate": 7.435885742129523e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 19288704,
      "step": 29505
    },
    {
      "epoch": 15.466457023060796,
      "grad_norm": 1.5242352485656738,
      "learning_rate": 7.427750713675071e-06,
      "loss": 0.0402,
      "num_input_tokens_seen": 19291584,
      "step": 29510
    },
    {
      "epoch": 15.469077568134171,
      "grad_norm": 0.1784522384405136,
      "learning_rate": 7.419619361134702e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 19295072,
      "step": 29515
    },
    {
      "epoch": 15.471698113207546,
      "grad_norm": 1.1042600870132446,
      "learning_rate": 7.411491686209387e-06,
      "loss": 0.0101,
      "num_input_tokens_seen": 19298144,
      "step": 29520
    },
    {
      "epoch": 15.474318658280922,
      "grad_norm": 1.3220598697662354,
      "learning_rate": 7.4033676905993525e-06,
      "loss": 0.0221,
      "num_input_tokens_seen": 19301472,
      "step": 29525
    },
    {
      "epoch": 15.476939203354299,
      "grad_norm": 0.11880683153867722,
      "learning_rate": 7.3952473760040444e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 19303936,
      "step": 29530
    },
    {
      "epoch": 15.479559748427674,
      "grad_norm": 1.104264736175537,
      "learning_rate": 7.387130744122148e-06,
      "loss": 0.0183,
      "num_input_tokens_seen": 19308416,
      "step": 29535
    },
    {
      "epoch": 15.482180293501049,
      "grad_norm": 0.188937246799469,
      "learning_rate": 7.379017796651558e-06,
      "loss": 0.0376,
      "num_input_tokens_seen": 19312992,
      "step": 29540
    },
    {
      "epoch": 15.484800838574424,
      "grad_norm": 0.2540210783481598,
      "learning_rate": 7.3709085352894085e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 19317792,
      "step": 29545
    },
    {
      "epoch": 15.4874213836478,
      "grad_norm": 9.613018035888672,
      "learning_rate": 7.362802961732071e-06,
      "loss": 0.0275,
      "num_input_tokens_seen": 19320928,
      "step": 29550
    },
    {
      "epoch": 15.490041928721174,
      "grad_norm": 0.422463595867157,
      "learning_rate": 7.354701077675136e-06,
      "loss": 0.0426,
      "num_input_tokens_seen": 19323584,
      "step": 29555
    },
    {
      "epoch": 15.49266247379455,
      "grad_norm": 0.27630168199539185,
      "learning_rate": 7.346602884813439e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 19326112,
      "step": 29560
    },
    {
      "epoch": 15.495283018867925,
      "grad_norm": 1.527013897895813,
      "learning_rate": 7.338508384841014e-06,
      "loss": 0.0254,
      "num_input_tokens_seen": 19329824,
      "step": 29565
    },
    {
      "epoch": 15.4979035639413,
      "grad_norm": 1.3976826667785645,
      "learning_rate": 7.3304175794511345e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 19332544,
      "step": 29570
    },
    {
      "epoch": 15.500524109014675,
      "grad_norm": 0.8334566354751587,
      "learning_rate": 7.3223304703363135e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 19335744,
      "step": 29575
    },
    {
      "epoch": 15.50314465408805,
      "grad_norm": 2.1758570671081543,
      "learning_rate": 7.314247059188281e-06,
      "loss": 0.0436,
      "num_input_tokens_seen": 19339456,
      "step": 29580
    },
    {
      "epoch": 15.505765199161425,
      "grad_norm": 1.8412138223648071,
      "learning_rate": 7.306167347697992e-06,
      "loss": 0.018,
      "num_input_tokens_seen": 19342752,
      "step": 29585
    },
    {
      "epoch": 15.5083857442348,
      "grad_norm": 0.5653024911880493,
      "learning_rate": 7.29809133755564e-06,
      "loss": 0.0276,
      "num_input_tokens_seen": 19345536,
      "step": 29590
    },
    {
      "epoch": 15.511006289308176,
      "grad_norm": 1.0876835584640503,
      "learning_rate": 7.290019030450629e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 19348768,
      "step": 29595
    },
    {
      "epoch": 15.51362683438155,
      "grad_norm": 1.0956834554672241,
      "learning_rate": 7.2819504280715794e-06,
      "loss": 0.0065,
      "num_input_tokens_seen": 19352128,
      "step": 29600
    },
    {
      "epoch": 15.516247379454926,
      "grad_norm": 2.223879098892212,
      "learning_rate": 7.273885532106364e-06,
      "loss": 0.0206,
      "num_input_tokens_seen": 19354784,
      "step": 29605
    },
    {
      "epoch": 15.518867924528301,
      "grad_norm": 1.5215378999710083,
      "learning_rate": 7.265824344242064e-06,
      "loss": 0.0145,
      "num_input_tokens_seen": 19358528,
      "step": 29610
    },
    {
      "epoch": 15.521488469601676,
      "grad_norm": 0.5341818332672119,
      "learning_rate": 7.257766866165e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 19361760,
      "step": 29615
    },
    {
      "epoch": 15.524109014675052,
      "grad_norm": 1.5786293745040894,
      "learning_rate": 7.24971309956069e-06,
      "loss": 0.0142,
      "num_input_tokens_seen": 19365504,
      "step": 29620
    },
    {
      "epoch": 15.526729559748428,
      "grad_norm": 0.5541608333587646,
      "learning_rate": 7.241663046113887e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 19369344,
      "step": 29625
    },
    {
      "epoch": 15.529350104821804,
      "grad_norm": 0.4313143789768219,
      "learning_rate": 7.233616707508576e-06,
      "loss": 0.0135,
      "num_input_tokens_seen": 19372224,
      "step": 29630
    },
    {
      "epoch": 15.531970649895179,
      "grad_norm": 0.5682860016822815,
      "learning_rate": 7.225574085427961e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 19374976,
      "step": 29635
    },
    {
      "epoch": 15.534591194968554,
      "grad_norm": 0.25019779801368713,
      "learning_rate": 7.217535181554474e-06,
      "loss": 0.0214,
      "num_input_tokens_seen": 19377440,
      "step": 29640
    },
    {
      "epoch": 15.53721174004193,
      "grad_norm": 0.8291465044021606,
      "learning_rate": 7.209499997569747e-06,
      "loss": 0.0088,
      "num_input_tokens_seen": 19380032,
      "step": 29645
    },
    {
      "epoch": 15.539832285115304,
      "grad_norm": 4.711690902709961,
      "learning_rate": 7.201468535154663e-06,
      "loss": 0.0265,
      "num_input_tokens_seen": 19383488,
      "step": 29650
    },
    {
      "epoch": 15.54245283018868,
      "grad_norm": 0.6535441875457764,
      "learning_rate": 7.193440795989295e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 19386560,
      "step": 29655
    },
    {
      "epoch": 15.545073375262055,
      "grad_norm": 1.859628438949585,
      "learning_rate": 7.185416781752965e-06,
      "loss": 0.03,
      "num_input_tokens_seen": 19389216,
      "step": 29660
    },
    {
      "epoch": 15.54769392033543,
      "grad_norm": 1.8863979578018188,
      "learning_rate": 7.177396494124206e-06,
      "loss": 0.0324,
      "num_input_tokens_seen": 19393056,
      "step": 29665
    },
    {
      "epoch": 15.550314465408805,
      "grad_norm": 1.7544548511505127,
      "learning_rate": 7.169379934780779e-06,
      "loss": 0.0232,
      "num_input_tokens_seen": 19396672,
      "step": 29670
    },
    {
      "epoch": 15.55293501048218,
      "grad_norm": 4.586415767669678,
      "learning_rate": 7.161367105399644e-06,
      "loss": 0.0128,
      "num_input_tokens_seen": 19399168,
      "step": 29675
    },
    {
      "epoch": 15.555555555555555,
      "grad_norm": 0.14371941983699799,
      "learning_rate": 7.153358007656991e-06,
      "loss": 0.0176,
      "num_input_tokens_seen": 19405248,
      "step": 29680
    },
    {
      "epoch": 15.55817610062893,
      "grad_norm": 0.23105759918689728,
      "learning_rate": 7.145352643228237e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 19407712,
      "step": 29685
    },
    {
      "epoch": 15.560796645702306,
      "grad_norm": 1.058577537536621,
      "learning_rate": 7.137351013788013e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 19410816,
      "step": 29690
    },
    {
      "epoch": 15.56341719077568,
      "grad_norm": 0.33966803550720215,
      "learning_rate": 7.129353121010177e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 19413888,
      "step": 29695
    },
    {
      "epoch": 15.566037735849056,
      "grad_norm": 0.4612198770046234,
      "learning_rate": 7.121358966567779e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 19417344,
      "step": 29700
    },
    {
      "epoch": 15.568658280922431,
      "grad_norm": 10.995233535766602,
      "learning_rate": 7.113368552133126e-06,
      "loss": 0.017,
      "num_input_tokens_seen": 19420512,
      "step": 29705
    },
    {
      "epoch": 15.571278825995806,
      "grad_norm": 0.5257472991943359,
      "learning_rate": 7.105381879377701e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 19423328,
      "step": 29710
    },
    {
      "epoch": 15.573899371069182,
      "grad_norm": 0.3558722138404846,
      "learning_rate": 7.097398949972234e-06,
      "loss": 0.0187,
      "num_input_tokens_seen": 19427136,
      "step": 29715
    },
    {
      "epoch": 15.576519916142558,
      "grad_norm": 3.719125747680664,
      "learning_rate": 7.089419765586672e-06,
      "loss": 0.0179,
      "num_input_tokens_seen": 19430496,
      "step": 29720
    },
    {
      "epoch": 15.579140461215934,
      "grad_norm": 1.2572959661483765,
      "learning_rate": 7.081444327890152e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 19433792,
      "step": 29725
    },
    {
      "epoch": 15.581761006289309,
      "grad_norm": 0.8053699731826782,
      "learning_rate": 7.073472638551054e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 19436384,
      "step": 29730
    },
    {
      "epoch": 15.584381551362684,
      "grad_norm": 0.4892231523990631,
      "learning_rate": 7.065504699236969e-06,
      "loss": 0.0103,
      "num_input_tokens_seen": 19439200,
      "step": 29735
    },
    {
      "epoch": 15.58700209643606,
      "grad_norm": 2.057917833328247,
      "learning_rate": 7.057540511614688e-06,
      "loss": 0.0065,
      "num_input_tokens_seen": 19442400,
      "step": 29740
    },
    {
      "epoch": 15.589622641509434,
      "grad_norm": 4.196730136871338,
      "learning_rate": 7.049580077350232e-06,
      "loss": 0.0193,
      "num_input_tokens_seen": 19446464,
      "step": 29745
    },
    {
      "epoch": 15.59224318658281,
      "grad_norm": 4.644940376281738,
      "learning_rate": 7.041623398108843e-06,
      "loss": 0.0182,
      "num_input_tokens_seen": 19449600,
      "step": 29750
    },
    {
      "epoch": 15.594863731656185,
      "grad_norm": 0.18860259652137756,
      "learning_rate": 7.033670475554949e-06,
      "loss": 0.0174,
      "num_input_tokens_seen": 19452608,
      "step": 29755
    },
    {
      "epoch": 15.59748427672956,
      "grad_norm": 0.725473940372467,
      "learning_rate": 7.025721311352224e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 19456832,
      "step": 29760
    },
    {
      "epoch": 15.600104821802935,
      "grad_norm": 4.804191589355469,
      "learning_rate": 7.017775907163546e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 19459136,
      "step": 29765
    },
    {
      "epoch": 15.60272536687631,
      "grad_norm": 0.8393948674201965,
      "learning_rate": 7.009834264650989e-06,
      "loss": 0.0138,
      "num_input_tokens_seen": 19463360,
      "step": 29770
    },
    {
      "epoch": 15.605345911949685,
      "grad_norm": 0.4853430390357971,
      "learning_rate": 7.001896385475867e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 19466432,
      "step": 29775
    },
    {
      "epoch": 15.60796645702306,
      "grad_norm": 0.5538056492805481,
      "learning_rate": 6.993962271298682e-06,
      "loss": 0.0233,
      "num_input_tokens_seen": 19470592,
      "step": 29780
    },
    {
      "epoch": 15.610587002096436,
      "grad_norm": 2.8615293502807617,
      "learning_rate": 6.986031923779166e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 19474496,
      "step": 29785
    },
    {
      "epoch": 15.61320754716981,
      "grad_norm": 1.7037458419799805,
      "learning_rate": 6.978105344576264e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 19477792,
      "step": 29790
    },
    {
      "epoch": 15.615828092243186,
      "grad_norm": 0.8656153082847595,
      "learning_rate": 6.970182535348111e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 19480736,
      "step": 29795
    },
    {
      "epoch": 15.618448637316561,
      "grad_norm": 1.2584407329559326,
      "learning_rate": 6.962263497752086e-06,
      "loss": 0.0273,
      "num_input_tokens_seen": 19485312,
      "step": 29800
    },
    {
      "epoch": 15.621069182389936,
      "grad_norm": 0.7860878705978394,
      "learning_rate": 6.9543482334447436e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 19488544,
      "step": 29805
    },
    {
      "epoch": 15.623689727463312,
      "grad_norm": 4.00771427154541,
      "learning_rate": 6.946436744081875e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 19491744,
      "step": 29810
    },
    {
      "epoch": 15.626310272536688,
      "grad_norm": 0.8921488523483276,
      "learning_rate": 6.938529031318472e-06,
      "loss": 0.0175,
      "num_input_tokens_seen": 19495136,
      "step": 29815
    },
    {
      "epoch": 15.628930817610064,
      "grad_norm": 0.2809044122695923,
      "learning_rate": 6.9306250968087485e-06,
      "loss": 0.0105,
      "num_input_tokens_seen": 19498272,
      "step": 29820
    },
    {
      "epoch": 15.631551362683439,
      "grad_norm": 0.715377151966095,
      "learning_rate": 6.922724942206102e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 19501792,
      "step": 29825
    },
    {
      "epoch": 15.634171907756814,
      "grad_norm": 1.0937221050262451,
      "learning_rate": 6.914828569163167e-06,
      "loss": 0.0601,
      "num_input_tokens_seen": 19504640,
      "step": 29830
    },
    {
      "epoch": 15.63679245283019,
      "grad_norm": 0.9174371957778931,
      "learning_rate": 6.906935979331763e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 19508032,
      "step": 29835
    },
    {
      "epoch": 15.639412997903564,
      "grad_norm": 2.5542454719543457,
      "learning_rate": 6.8990471743629356e-06,
      "loss": 0.0116,
      "num_input_tokens_seen": 19511008,
      "step": 29840
    },
    {
      "epoch": 15.64203354297694,
      "grad_norm": 0.5173631906509399,
      "learning_rate": 6.89116215590693e-06,
      "loss": 0.0197,
      "num_input_tokens_seen": 19514240,
      "step": 29845
    },
    {
      "epoch": 15.644654088050315,
      "grad_norm": 0.25730282068252563,
      "learning_rate": 6.8832809256132146e-06,
      "loss": 0.0155,
      "num_input_tokens_seen": 19517056,
      "step": 29850
    },
    {
      "epoch": 15.64727463312369,
      "grad_norm": 1.847777247428894,
      "learning_rate": 6.875403485130444e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 19520256,
      "step": 29855
    },
    {
      "epoch": 15.649895178197065,
      "grad_norm": 0.8812087774276733,
      "learning_rate": 6.867529836106479e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 19522976,
      "step": 29860
    },
    {
      "epoch": 15.65251572327044,
      "grad_norm": 3.6227173805236816,
      "learning_rate": 6.8596599801884045e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 19526784,
      "step": 29865
    },
    {
      "epoch": 15.655136268343815,
      "grad_norm": 1.195278286933899,
      "learning_rate": 6.851793919022509e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 19529920,
      "step": 29870
    },
    {
      "epoch": 15.65775681341719,
      "grad_norm": 0.5889379978179932,
      "learning_rate": 6.843931654254285e-06,
      "loss": 0.0134,
      "num_input_tokens_seen": 19532608,
      "step": 29875
    },
    {
      "epoch": 15.660377358490566,
      "grad_norm": 0.1461632251739502,
      "learning_rate": 6.836073187528425e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 19535680,
      "step": 29880
    },
    {
      "epoch": 15.66299790356394,
      "grad_norm": 0.5146178603172302,
      "learning_rate": 6.828218520488821e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 19538144,
      "step": 29885
    },
    {
      "epoch": 15.665618448637316,
      "grad_norm": 1.2774895429611206,
      "learning_rate": 6.820367654778589e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 19541344,
      "step": 29890
    },
    {
      "epoch": 15.668238993710691,
      "grad_norm": 2.5893638134002686,
      "learning_rate": 6.812520592040039e-06,
      "loss": 0.0427,
      "num_input_tokens_seen": 19544640,
      "step": 29895
    },
    {
      "epoch": 15.670859538784066,
      "grad_norm": 0.1431138515472412,
      "learning_rate": 6.804677333914689e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 19547872,
      "step": 29900
    },
    {
      "epoch": 15.673480083857442,
      "grad_norm": 1.425045371055603,
      "learning_rate": 6.796837882043261e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 19550880,
      "step": 29905
    },
    {
      "epoch": 15.676100628930818,
      "grad_norm": 0.6211567521095276,
      "learning_rate": 6.7890022380656785e-06,
      "loss": 0.0363,
      "num_input_tokens_seen": 19554112,
      "step": 29910
    },
    {
      "epoch": 15.678721174004194,
      "grad_norm": 1.0437066555023193,
      "learning_rate": 6.781170403621056e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 19557760,
      "step": 29915
    },
    {
      "epoch": 15.681341719077569,
      "grad_norm": 0.3152290880680084,
      "learning_rate": 6.773342380347736e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 19561312,
      "step": 29920
    },
    {
      "epoch": 15.683962264150944,
      "grad_norm": 2.6022019386291504,
      "learning_rate": 6.765518169883248e-06,
      "loss": 0.0706,
      "num_input_tokens_seen": 19565248,
      "step": 29925
    },
    {
      "epoch": 15.68658280922432,
      "grad_norm": 8.850946426391602,
      "learning_rate": 6.757697773864338e-06,
      "loss": 0.0266,
      "num_input_tokens_seen": 19567808,
      "step": 29930
    },
    {
      "epoch": 15.689203354297694,
      "grad_norm": 0.2136101871728897,
      "learning_rate": 6.749881193926932e-06,
      "loss": 0.0211,
      "num_input_tokens_seen": 19570880,
      "step": 29935
    },
    {
      "epoch": 15.69182389937107,
      "grad_norm": 1.7496336698532104,
      "learning_rate": 6.742068431706167e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 19573536,
      "step": 29940
    },
    {
      "epoch": 15.694444444444445,
      "grad_norm": 0.6967637538909912,
      "learning_rate": 6.734259488836386e-06,
      "loss": 0.0266,
      "num_input_tokens_seen": 19577408,
      "step": 29945
    },
    {
      "epoch": 15.69706498951782,
      "grad_norm": 0.14967305958271027,
      "learning_rate": 6.726454366951135e-06,
      "loss": 0.0206,
      "num_input_tokens_seen": 19580960,
      "step": 29950
    },
    {
      "epoch": 15.699685534591195,
      "grad_norm": 0.5478802919387817,
      "learning_rate": 6.7186530676831616e-06,
      "loss": 0.1167,
      "num_input_tokens_seen": 19584448,
      "step": 29955
    },
    {
      "epoch": 15.70230607966457,
      "grad_norm": 0.19273564219474792,
      "learning_rate": 6.710855592664403e-06,
      "loss": 0.0136,
      "num_input_tokens_seen": 19587392,
      "step": 29960
    },
    {
      "epoch": 15.704926624737945,
      "grad_norm": 2.403973340988159,
      "learning_rate": 6.703061943525993e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 19590240,
      "step": 29965
    },
    {
      "epoch": 15.70754716981132,
      "grad_norm": 0.8675986528396606,
      "learning_rate": 6.69527212189828e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 19593056,
      "step": 29970
    },
    {
      "epoch": 15.710167714884696,
      "grad_norm": 0.5192034244537354,
      "learning_rate": 6.687486129410811e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 19596160,
      "step": 29975
    },
    {
      "epoch": 15.71278825995807,
      "grad_norm": 0.2782927453517914,
      "learning_rate": 6.679703967692322e-06,
      "loss": 0.0143,
      "num_input_tokens_seen": 19601184,
      "step": 29980
    },
    {
      "epoch": 15.715408805031446,
      "grad_norm": 0.5462021231651306,
      "learning_rate": 6.67192563837076e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 19604160,
      "step": 29985
    },
    {
      "epoch": 15.718029350104821,
      "grad_norm": 0.24352151155471802,
      "learning_rate": 6.664151143073258e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 19607872,
      "step": 29990
    },
    {
      "epoch": 15.720649895178196,
      "grad_norm": 0.8978573679924011,
      "learning_rate": 6.656380483426141e-06,
      "loss": 0.0129,
      "num_input_tokens_seen": 19612096,
      "step": 29995
    },
    {
      "epoch": 15.723270440251572,
      "grad_norm": 0.10927080363035202,
      "learning_rate": 6.648613661054956e-06,
      "loss": 0.0107,
      "num_input_tokens_seen": 19614976,
      "step": 30000
    },
    {
      "epoch": 15.725890985324948,
      "grad_norm": 1.0184147357940674,
      "learning_rate": 6.6408506775844256e-06,
      "loss": 0.0181,
      "num_input_tokens_seen": 19618624,
      "step": 30005
    },
    {
      "epoch": 15.728511530398324,
      "grad_norm": 1.9532933235168457,
      "learning_rate": 6.633091534638492e-06,
      "loss": 0.0352,
      "num_input_tokens_seen": 19621920,
      "step": 30010
    },
    {
      "epoch": 15.731132075471699,
      "grad_norm": 0.6826417446136475,
      "learning_rate": 6.625336233840257e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 19624352,
      "step": 30015
    },
    {
      "epoch": 15.733752620545074,
      "grad_norm": 2.777509927749634,
      "learning_rate": 6.617584776812064e-06,
      "loss": 0.0439,
      "num_input_tokens_seen": 19631744,
      "step": 30020
    },
    {
      "epoch": 15.73637316561845,
      "grad_norm": 2.7183313369750977,
      "learning_rate": 6.6098371651754085e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 19634592,
      "step": 30025
    },
    {
      "epoch": 15.738993710691824,
      "grad_norm": 1.1589951515197754,
      "learning_rate": 6.602093400551012e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 19637472,
      "step": 30030
    },
    {
      "epoch": 15.7416142557652,
      "grad_norm": 2.200120210647583,
      "learning_rate": 6.59435348455879e-06,
      "loss": 0.0501,
      "num_input_tokens_seen": 19641344,
      "step": 30035
    },
    {
      "epoch": 15.744234800838575,
      "grad_norm": 1.334020733833313,
      "learning_rate": 6.586617418817828e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 19644320,
      "step": 30040
    },
    {
      "epoch": 15.74685534591195,
      "grad_norm": 1.1586339473724365,
      "learning_rate": 6.578885204946439e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 19648512,
      "step": 30045
    },
    {
      "epoch": 15.749475890985325,
      "grad_norm": 8.879515647888184,
      "learning_rate": 6.571156844562098e-06,
      "loss": 0.0393,
      "num_input_tokens_seen": 19651296,
      "step": 30050
    },
    {
      "epoch": 15.7520964360587,
      "grad_norm": 0.6536783576011658,
      "learning_rate": 6.563432339281497e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 19654080,
      "step": 30055
    },
    {
      "epoch": 15.754716981132075,
      "grad_norm": 0.15846361219882965,
      "learning_rate": 6.555711690720517e-06,
      "loss": 0.0298,
      "num_input_tokens_seen": 19658240,
      "step": 30060
    },
    {
      "epoch": 15.75733752620545,
      "grad_norm": 2.576385021209717,
      "learning_rate": 6.547994900494234e-06,
      "loss": 0.0381,
      "num_input_tokens_seen": 19661792,
      "step": 30065
    },
    {
      "epoch": 15.759958071278826,
      "grad_norm": 0.47514137625694275,
      "learning_rate": 6.540281970216899e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 19664288,
      "step": 30070
    },
    {
      "epoch": 15.7625786163522,
      "grad_norm": 0.07994967699050903,
      "learning_rate": 6.5325729015019845e-06,
      "loss": 0.1062,
      "num_input_tokens_seen": 19670656,
      "step": 30075
    },
    {
      "epoch": 15.765199161425576,
      "grad_norm": 0.8987004160881042,
      "learning_rate": 6.524867695962122e-06,
      "loss": 0.0338,
      "num_input_tokens_seen": 19674048,
      "step": 30080
    },
    {
      "epoch": 15.767819706498951,
      "grad_norm": 1.6020675897598267,
      "learning_rate": 6.517166355209165e-06,
      "loss": 0.0085,
      "num_input_tokens_seen": 19676896,
      "step": 30085
    },
    {
      "epoch": 15.770440251572326,
      "grad_norm": 2.18476939201355,
      "learning_rate": 6.50946888085415e-06,
      "loss": 0.0138,
      "num_input_tokens_seen": 19679392,
      "step": 30090
    },
    {
      "epoch": 15.773060796645701,
      "grad_norm": 1.317510962486267,
      "learning_rate": 6.5017752745072855e-06,
      "loss": 0.026,
      "num_input_tokens_seen": 19682560,
      "step": 30095
    },
    {
      "epoch": 15.775681341719078,
      "grad_norm": 3.5398142337799072,
      "learning_rate": 6.494085537777994e-06,
      "loss": 0.0096,
      "num_input_tokens_seen": 19685312,
      "step": 30100
    },
    {
      "epoch": 15.778301886792454,
      "grad_norm": 2.6896865367889404,
      "learning_rate": 6.48639967227489e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 19688064,
      "step": 30105
    },
    {
      "epoch": 15.780922431865829,
      "grad_norm": 2.7160844802856445,
      "learning_rate": 6.47871767960575e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 19690976,
      "step": 30110
    },
    {
      "epoch": 15.783542976939204,
      "grad_norm": 2.874415874481201,
      "learning_rate": 6.471039561377581e-06,
      "loss": 0.024,
      "num_input_tokens_seen": 19693728,
      "step": 30115
    },
    {
      "epoch": 15.786163522012579,
      "grad_norm": 0.5268253087997437,
      "learning_rate": 6.463365319196538e-06,
      "loss": 0.0322,
      "num_input_tokens_seen": 19700864,
      "step": 30120
    },
    {
      "epoch": 15.788784067085954,
      "grad_norm": 0.1713394671678543,
      "learning_rate": 6.4556949546679905e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 19703488,
      "step": 30125
    },
    {
      "epoch": 15.79140461215933,
      "grad_norm": 0.6258844137191772,
      "learning_rate": 6.448028469396497e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 19706944,
      "step": 30130
    },
    {
      "epoch": 15.794025157232705,
      "grad_norm": 0.3723124861717224,
      "learning_rate": 6.440365864985801e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 19710400,
      "step": 30135
    },
    {
      "epoch": 15.79664570230608,
      "grad_norm": 0.47838127613067627,
      "learning_rate": 6.432707143038818e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 19713056,
      "step": 30140
    },
    {
      "epoch": 15.799266247379455,
      "grad_norm": 1.0312154293060303,
      "learning_rate": 6.4250523051576834e-06,
      "loss": 0.0183,
      "num_input_tokens_seen": 19717248,
      "step": 30145
    },
    {
      "epoch": 15.80188679245283,
      "grad_norm": 0.841962993144989,
      "learning_rate": 6.417401352943686e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 19719552,
      "step": 30150
    },
    {
      "epoch": 15.804507337526205,
      "grad_norm": 3.467177391052246,
      "learning_rate": 6.409754287997322e-06,
      "loss": 0.0158,
      "num_input_tokens_seen": 19722304,
      "step": 30155
    },
    {
      "epoch": 15.80712788259958,
      "grad_norm": 2.322770118713379,
      "learning_rate": 6.40211111191828e-06,
      "loss": 0.0213,
      "num_input_tokens_seen": 19725440,
      "step": 30160
    },
    {
      "epoch": 15.809748427672956,
      "grad_norm": 0.4019503891468048,
      "learning_rate": 6.394471826305409e-06,
      "loss": 0.0127,
      "num_input_tokens_seen": 19728864,
      "step": 30165
    },
    {
      "epoch": 15.81236897274633,
      "grad_norm": 2.303856372833252,
      "learning_rate": 6.3868364327567795e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 19732192,
      "step": 30170
    },
    {
      "epoch": 15.814989517819706,
      "grad_norm": 0.32328158617019653,
      "learning_rate": 6.379204932869606e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 19735040,
      "step": 30175
    },
    {
      "epoch": 15.817610062893081,
      "grad_norm": 0.5707583427429199,
      "learning_rate": 6.371577328240327e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 19738368,
      "step": 30180
    },
    {
      "epoch": 15.820230607966456,
      "grad_norm": 2.224302291870117,
      "learning_rate": 6.363953620464547e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 19742176,
      "step": 30185
    },
    {
      "epoch": 15.822851153039831,
      "grad_norm": 1.457457423210144,
      "learning_rate": 6.356333811137064e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 19745760,
      "step": 30190
    },
    {
      "epoch": 15.825471698113208,
      "grad_norm": 0.7146062254905701,
      "learning_rate": 6.34871790185185e-06,
      "loss": 0.0185,
      "num_input_tokens_seen": 19748832,
      "step": 30195
    },
    {
      "epoch": 15.828092243186584,
      "grad_norm": 0.36425548791885376,
      "learning_rate": 6.341105894202057e-06,
      "loss": 0.0237,
      "num_input_tokens_seen": 19752000,
      "step": 30200
    },
    {
      "epoch": 15.830712788259959,
      "grad_norm": 0.5591587424278259,
      "learning_rate": 6.333497789780041e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 19755744,
      "step": 30205
    },
    {
      "epoch": 15.833333333333334,
      "grad_norm": 0.14390096068382263,
      "learning_rate": 6.325893590177329e-06,
      "loss": 0.0186,
      "num_input_tokens_seen": 19759072,
      "step": 30210
    },
    {
      "epoch": 15.835953878406709,
      "grad_norm": 1.2373727560043335,
      "learning_rate": 6.318293296984631e-06,
      "loss": 0.0188,
      "num_input_tokens_seen": 19762432,
      "step": 30215
    },
    {
      "epoch": 15.838574423480084,
      "grad_norm": 0.9532637596130371,
      "learning_rate": 6.3106969117918495e-06,
      "loss": 0.0186,
      "num_input_tokens_seen": 19765248,
      "step": 30220
    },
    {
      "epoch": 15.84119496855346,
      "grad_norm": 1.33212411403656,
      "learning_rate": 6.303104436188057e-06,
      "loss": 0.0263,
      "num_input_tokens_seen": 19768224,
      "step": 30225
    },
    {
      "epoch": 15.843815513626835,
      "grad_norm": 1.4015990495681763,
      "learning_rate": 6.2955158717615036e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 19771232,
      "step": 30230
    },
    {
      "epoch": 15.84643605870021,
      "grad_norm": 0.953679621219635,
      "learning_rate": 6.287931220099638e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 19774016,
      "step": 30235
    },
    {
      "epoch": 15.849056603773585,
      "grad_norm": 2.1729846000671387,
      "learning_rate": 6.280350482789082e-06,
      "loss": 0.021,
      "num_input_tokens_seen": 19777568,
      "step": 30240
    },
    {
      "epoch": 15.85167714884696,
      "grad_norm": 0.2134305238723755,
      "learning_rate": 6.272773661415645e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 19780448,
      "step": 30245
    },
    {
      "epoch": 15.854297693920335,
      "grad_norm": 0.572161078453064,
      "learning_rate": 6.26520075756431e-06,
      "loss": 0.0102,
      "num_input_tokens_seen": 19783808,
      "step": 30250
    },
    {
      "epoch": 15.85691823899371,
      "grad_norm": 3.1374473571777344,
      "learning_rate": 6.2576317728192304e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 19787136,
      "step": 30255
    },
    {
      "epoch": 15.859538784067086,
      "grad_norm": 3.6670520305633545,
      "learning_rate": 6.250066708763761e-06,
      "loss": 0.0282,
      "num_input_tokens_seen": 19789920,
      "step": 30260
    },
    {
      "epoch": 15.86215932914046,
      "grad_norm": 0.043853532522916794,
      "learning_rate": 6.242505566980422e-06,
      "loss": 0.013,
      "num_input_tokens_seen": 19793024,
      "step": 30265
    },
    {
      "epoch": 15.864779874213836,
      "grad_norm": 0.6551027894020081,
      "learning_rate": 6.234948349050931e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 19796672,
      "step": 30270
    },
    {
      "epoch": 15.867400419287211,
      "grad_norm": 2.14483380317688,
      "learning_rate": 6.227395056556162e-06,
      "loss": 0.0088,
      "num_input_tokens_seen": 19798944,
      "step": 30275
    },
    {
      "epoch": 15.870020964360586,
      "grad_norm": 0.09315168112516403,
      "learning_rate": 6.219845691076173e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 19801664,
      "step": 30280
    },
    {
      "epoch": 15.872641509433961,
      "grad_norm": 0.8636082410812378,
      "learning_rate": 6.212300254190206e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 19804384,
      "step": 30285
    },
    {
      "epoch": 15.875262054507338,
      "grad_norm": 2.345627546310425,
      "learning_rate": 6.204758747476688e-06,
      "loss": 0.0132,
      "num_input_tokens_seen": 19807744,
      "step": 30290
    },
    {
      "epoch": 15.877882599580714,
      "grad_norm": 0.611552894115448,
      "learning_rate": 6.1972211725132095e-06,
      "loss": 0.0164,
      "num_input_tokens_seen": 19810336,
      "step": 30295
    },
    {
      "epoch": 15.880503144654089,
      "grad_norm": 1.2272166013717651,
      "learning_rate": 6.189687530876559e-06,
      "loss": 0.0281,
      "num_input_tokens_seen": 19813440,
      "step": 30300
    },
    {
      "epoch": 15.883123689727464,
      "grad_norm": 1.3880008459091187,
      "learning_rate": 6.182157824142676e-06,
      "loss": 0.056,
      "num_input_tokens_seen": 19816128,
      "step": 30305
    },
    {
      "epoch": 15.885744234800839,
      "grad_norm": 1.7044278383255005,
      "learning_rate": 6.174632053886681e-06,
      "loss": 0.0102,
      "num_input_tokens_seen": 19818656,
      "step": 30310
    },
    {
      "epoch": 15.888364779874214,
      "grad_norm": 0.9273452162742615,
      "learning_rate": 6.167110221682893e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 19821856,
      "step": 30315
    },
    {
      "epoch": 15.89098532494759,
      "grad_norm": 0.8401075601577759,
      "learning_rate": 6.159592329104788e-06,
      "loss": 0.0125,
      "num_input_tokens_seen": 19824864,
      "step": 30320
    },
    {
      "epoch": 15.893605870020965,
      "grad_norm": 1.6461795568466187,
      "learning_rate": 6.152078377725032e-06,
      "loss": 0.0191,
      "num_input_tokens_seen": 19828352,
      "step": 30325
    },
    {
      "epoch": 15.89622641509434,
      "grad_norm": 0.8521744608879089,
      "learning_rate": 6.144568369115453e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 19830848,
      "step": 30330
    },
    {
      "epoch": 15.898846960167715,
      "grad_norm": 0.27188384532928467,
      "learning_rate": 6.137062304847046e-06,
      "loss": 0.0282,
      "num_input_tokens_seen": 19836448,
      "step": 30335
    },
    {
      "epoch": 15.90146750524109,
      "grad_norm": 0.4325353503227234,
      "learning_rate": 6.129560186490008e-06,
      "loss": 0.0093,
      "num_input_tokens_seen": 19839296,
      "step": 30340
    },
    {
      "epoch": 15.904088050314465,
      "grad_norm": 0.710800290107727,
      "learning_rate": 6.122062015613694e-06,
      "loss": 0.006,
      "num_input_tokens_seen": 19842176,
      "step": 30345
    },
    {
      "epoch": 15.90670859538784,
      "grad_norm": 1.1258018016815186,
      "learning_rate": 6.114567793786641e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 19845216,
      "step": 30350
    },
    {
      "epoch": 15.909329140461216,
      "grad_norm": 1.252652645111084,
      "learning_rate": 6.107077522576543e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 19848064,
      "step": 30355
    },
    {
      "epoch": 15.91194968553459,
      "grad_norm": 0.6185701489448547,
      "learning_rate": 6.099591203550292e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 19850912,
      "step": 30360
    },
    {
      "epoch": 15.914570230607966,
      "grad_norm": 2.7839248180389404,
      "learning_rate": 6.092108838273927e-06,
      "loss": 0.0349,
      "num_input_tokens_seen": 19854688,
      "step": 30365
    },
    {
      "epoch": 15.917190775681341,
      "grad_norm": 0.7696391344070435,
      "learning_rate": 6.08463042831268e-06,
      "loss": 0.0151,
      "num_input_tokens_seen": 19857792,
      "step": 30370
    },
    {
      "epoch": 15.919811320754716,
      "grad_norm": 1.1036772727966309,
      "learning_rate": 6.0771559752309496e-06,
      "loss": 0.0103,
      "num_input_tokens_seen": 19860672,
      "step": 30375
    },
    {
      "epoch": 15.922431865828091,
      "grad_norm": 1.0218634605407715,
      "learning_rate": 6.069685480592313e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 19863584,
      "step": 30380
    },
    {
      "epoch": 15.925052410901468,
      "grad_norm": 0.4475801885128021,
      "learning_rate": 6.062218945959497e-06,
      "loss": 0.011,
      "num_input_tokens_seen": 19866912,
      "step": 30385
    },
    {
      "epoch": 15.927672955974844,
      "grad_norm": 0.4858594536781311,
      "learning_rate": 6.054756372894435e-06,
      "loss": 0.0296,
      "num_input_tokens_seen": 19869920,
      "step": 30390
    },
    {
      "epoch": 15.930293501048219,
      "grad_norm": 0.7077614665031433,
      "learning_rate": 6.04729776295819e-06,
      "loss": 0.0596,
      "num_input_tokens_seen": 19872736,
      "step": 30395
    },
    {
      "epoch": 15.932914046121594,
      "grad_norm": 0.2779400944709778,
      "learning_rate": 6.0398431177110306e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 19875584,
      "step": 30400
    },
    {
      "epoch": 15.935534591194969,
      "grad_norm": 0.10591208189725876,
      "learning_rate": 6.032392438712389e-06,
      "loss": 0.0134,
      "num_input_tokens_seen": 19879168,
      "step": 30405
    },
    {
      "epoch": 15.938155136268344,
      "grad_norm": 0.7035409212112427,
      "learning_rate": 6.024945727520847e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 19882528,
      "step": 30410
    },
    {
      "epoch": 15.94077568134172,
      "grad_norm": 1.8117568492889404,
      "learning_rate": 6.01750298569419e-06,
      "loss": 0.0145,
      "num_input_tokens_seen": 19885440,
      "step": 30415
    },
    {
      "epoch": 15.943396226415095,
      "grad_norm": 0.38012123107910156,
      "learning_rate": 6.010064214789335e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 19888320,
      "step": 30420
    },
    {
      "epoch": 15.94601677148847,
      "grad_norm": 1.2766863107681274,
      "learning_rate": 6.002629416362399e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 19891520,
      "step": 30425
    },
    {
      "epoch": 15.948637316561845,
      "grad_norm": 0.7567974328994751,
      "learning_rate": 5.995198591968662e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 19894688,
      "step": 30430
    },
    {
      "epoch": 15.95125786163522,
      "grad_norm": 0.6676896214485168,
      "learning_rate": 5.987771743162554e-06,
      "loss": 0.0176,
      "num_input_tokens_seen": 19897408,
      "step": 30435
    },
    {
      "epoch": 15.953878406708595,
      "grad_norm": 1.5729318857192993,
      "learning_rate": 5.9803488714976955e-06,
      "loss": 0.0175,
      "num_input_tokens_seen": 19901056,
      "step": 30440
    },
    {
      "epoch": 15.95649895178197,
      "grad_norm": 1.1785153150558472,
      "learning_rate": 5.972929978526872e-06,
      "loss": 0.0145,
      "num_input_tokens_seen": 19905056,
      "step": 30445
    },
    {
      "epoch": 15.959119496855346,
      "grad_norm": 1.176201581954956,
      "learning_rate": 5.965515065802019e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 19908160,
      "step": 30450
    },
    {
      "epoch": 15.96174004192872,
      "grad_norm": 0.6218587160110474,
      "learning_rate": 5.958104134874254e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 19911040,
      "step": 30455
    },
    {
      "epoch": 15.964360587002096,
      "grad_norm": 2.0798499584198,
      "learning_rate": 5.950697187293872e-06,
      "loss": 0.0206,
      "num_input_tokens_seen": 19914720,
      "step": 30460
    },
    {
      "epoch": 15.966981132075471,
      "grad_norm": 0.19406676292419434,
      "learning_rate": 5.943294224610305e-06,
      "loss": 0.0158,
      "num_input_tokens_seen": 19917312,
      "step": 30465
    },
    {
      "epoch": 15.969601677148846,
      "grad_norm": 0.8631705641746521,
      "learning_rate": 5.935895248372175e-06,
      "loss": 0.0207,
      "num_input_tokens_seen": 19920480,
      "step": 30470
    },
    {
      "epoch": 15.972222222222221,
      "grad_norm": 0.22279535233974457,
      "learning_rate": 5.928500260127273e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 19923104,
      "step": 30475
    },
    {
      "epoch": 15.974842767295598,
      "grad_norm": 1.6483153104782104,
      "learning_rate": 5.921109261422531e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 19926688,
      "step": 30480
    },
    {
      "epoch": 15.977463312368974,
      "grad_norm": 0.31509125232696533,
      "learning_rate": 5.913722253804071e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 19930624,
      "step": 30485
    },
    {
      "epoch": 15.980083857442349,
      "grad_norm": 0.35778507590293884,
      "learning_rate": 5.906339238817163e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 19933984,
      "step": 30490
    },
    {
      "epoch": 15.982704402515724,
      "grad_norm": 0.8355855345726013,
      "learning_rate": 5.8989602180062554e-06,
      "loss": 0.006,
      "num_input_tokens_seen": 19936288,
      "step": 30495
    },
    {
      "epoch": 15.985324947589099,
      "grad_norm": 5.016417503356934,
      "learning_rate": 5.891585192914953e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 19939648,
      "step": 30500
    },
    {
      "epoch": 15.987945492662474,
      "grad_norm": 0.6926714777946472,
      "learning_rate": 5.884214165086036e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 19942848,
      "step": 30505
    },
    {
      "epoch": 15.99056603773585,
      "grad_norm": 4.294330596923828,
      "learning_rate": 5.876847136061428e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 19946176,
      "step": 30510
    },
    {
      "epoch": 15.993186582809225,
      "grad_norm": 2.478041648864746,
      "learning_rate": 5.869484107382228e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 19948800,
      "step": 30515
    },
    {
      "epoch": 15.9958071278826,
      "grad_norm": 0.44312962889671326,
      "learning_rate": 5.862125080588696e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 19952288,
      "step": 30520
    },
    {
      "epoch": 15.998427672955975,
      "grad_norm": 0.1818695217370987,
      "learning_rate": 5.854770057220263e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 19955424,
      "step": 30525
    },
    {
      "epoch": 16.0,
      "eval_loss": 1.4609568119049072,
      "eval_runtime": 11.7734,
      "eval_samples_per_second": 72.027,
      "eval_steps_per_second": 18.007,
      "num_input_tokens_seen": 19956544,
      "step": 30528
    },
    {
      "epoch": 16.00104821802935,
      "grad_norm": 0.272484689950943,
      "learning_rate": 5.8474190388155216e-06,
      "loss": 0.0114,
      "num_input_tokens_seen": 19957376,
      "step": 30530
    },
    {
      "epoch": 16.003668763102727,
      "grad_norm": 0.21193663775920868,
      "learning_rate": 5.840072026912205e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 19960768,
      "step": 30535
    },
    {
      "epoch": 16.0062893081761,
      "grad_norm": 0.34705427289009094,
      "learning_rate": 5.83272902304724e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 19964416,
      "step": 30540
    },
    {
      "epoch": 16.008909853249477,
      "grad_norm": 0.1611567735671997,
      "learning_rate": 5.825390028756683e-06,
      "loss": 0.0155,
      "num_input_tokens_seen": 19967904,
      "step": 30545
    },
    {
      "epoch": 16.01153039832285,
      "grad_norm": 1.0591236352920532,
      "learning_rate": 5.818055045575776e-06,
      "loss": 0.059,
      "num_input_tokens_seen": 19971872,
      "step": 30550
    },
    {
      "epoch": 16.014150943396228,
      "grad_norm": 0.9734894037246704,
      "learning_rate": 5.810724075038917e-06,
      "loss": 0.0123,
      "num_input_tokens_seen": 19975072,
      "step": 30555
    },
    {
      "epoch": 16.0167714884696,
      "grad_norm": 0.20582954585552216,
      "learning_rate": 5.803397118679665e-06,
      "loss": 0.0182,
      "num_input_tokens_seen": 19978144,
      "step": 30560
    },
    {
      "epoch": 16.019392033542978,
      "grad_norm": 0.36887314915657043,
      "learning_rate": 5.796074178030727e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 19981216,
      "step": 30565
    },
    {
      "epoch": 16.02201257861635,
      "grad_norm": 0.5145923495292664,
      "learning_rate": 5.788755254623976e-06,
      "loss": 0.0194,
      "num_input_tokens_seen": 19984896,
      "step": 30570
    },
    {
      "epoch": 16.02463312368973,
      "grad_norm": 0.2391364425420761,
      "learning_rate": 5.7814403499904525e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 19988256,
      "step": 30575
    },
    {
      "epoch": 16.0272536687631,
      "grad_norm": 0.13101787865161896,
      "learning_rate": 5.774129465660352e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 19991744,
      "step": 30580
    },
    {
      "epoch": 16.02987421383648,
      "grad_norm": 0.35245227813720703,
      "learning_rate": 5.7668226031630246e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 19995104,
      "step": 30585
    },
    {
      "epoch": 16.032494758909852,
      "grad_norm": 0.2985983192920685,
      "learning_rate": 5.759519764026999e-06,
      "loss": 0.006,
      "num_input_tokens_seen": 19997472,
      "step": 30590
    },
    {
      "epoch": 16.03511530398323,
      "grad_norm": 5.784962177276611,
      "learning_rate": 5.7522209497799185e-06,
      "loss": 0.016,
      "num_input_tokens_seen": 19999776,
      "step": 30595
    },
    {
      "epoch": 16.037735849056602,
      "grad_norm": 1.5643320083618164,
      "learning_rate": 5.7449261619486246e-06,
      "loss": 0.0169,
      "num_input_tokens_seen": 20002848,
      "step": 30600
    },
    {
      "epoch": 16.04035639412998,
      "grad_norm": 1.8220641613006592,
      "learning_rate": 5.737635402059105e-06,
      "loss": 0.0065,
      "num_input_tokens_seen": 20006176,
      "step": 30605
    },
    {
      "epoch": 16.042976939203353,
      "grad_norm": 3.2202937602996826,
      "learning_rate": 5.730348671636501e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 20009088,
      "step": 30610
    },
    {
      "epoch": 16.04559748427673,
      "grad_norm": 0.5083738565444946,
      "learning_rate": 5.723065972205124e-06,
      "loss": 0.0102,
      "num_input_tokens_seen": 20012000,
      "step": 30615
    },
    {
      "epoch": 16.048218029350103,
      "grad_norm": 1.659100890159607,
      "learning_rate": 5.715787305288423e-06,
      "loss": 0.0609,
      "num_input_tokens_seen": 20016416,
      "step": 30620
    },
    {
      "epoch": 16.05083857442348,
      "grad_norm": 0.2051880657672882,
      "learning_rate": 5.7085126724090024e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 20020608,
      "step": 30625
    },
    {
      "epoch": 16.053459119496857,
      "grad_norm": 3.5998730659484863,
      "learning_rate": 5.7012420750886445e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 20023360,
      "step": 30630
    },
    {
      "epoch": 16.05607966457023,
      "grad_norm": 0.8133007884025574,
      "learning_rate": 5.693975514848271e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 20026240,
      "step": 30635
    },
    {
      "epoch": 16.058700209643607,
      "grad_norm": 1.8562686443328857,
      "learning_rate": 5.686712993207971e-06,
      "loss": 0.0199,
      "num_input_tokens_seen": 20029568,
      "step": 30640
    },
    {
      "epoch": 16.06132075471698,
      "grad_norm": 0.3267999291419983,
      "learning_rate": 5.679454511686977e-06,
      "loss": 0.0275,
      "num_input_tokens_seen": 20032288,
      "step": 30645
    },
    {
      "epoch": 16.063941299790358,
      "grad_norm": 0.6000897288322449,
      "learning_rate": 5.672200071803671e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 20035584,
      "step": 30650
    },
    {
      "epoch": 16.06656184486373,
      "grad_norm": 0.6434508562088013,
      "learning_rate": 5.664949675075607e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 20038912,
      "step": 30655
    },
    {
      "epoch": 16.069182389937108,
      "grad_norm": 0.8900846838951111,
      "learning_rate": 5.657703323019486e-06,
      "loss": 0.0167,
      "num_input_tokens_seen": 20042656,
      "step": 30660
    },
    {
      "epoch": 16.07180293501048,
      "grad_norm": 0.9849187135696411,
      "learning_rate": 5.65046101715116e-06,
      "loss": 0.0163,
      "num_input_tokens_seen": 20046528,
      "step": 30665
    },
    {
      "epoch": 16.07442348008386,
      "grad_norm": 0.1447007954120636,
      "learning_rate": 5.643222758985644e-06,
      "loss": 0.0107,
      "num_input_tokens_seen": 20049600,
      "step": 30670
    },
    {
      "epoch": 16.07704402515723,
      "grad_norm": 1.884300947189331,
      "learning_rate": 5.635988550037094e-06,
      "loss": 0.0171,
      "num_input_tokens_seen": 20052480,
      "step": 30675
    },
    {
      "epoch": 16.07966457023061,
      "grad_norm": 0.9560884833335876,
      "learning_rate": 5.628758391818817e-06,
      "loss": 0.0258,
      "num_input_tokens_seen": 20056128,
      "step": 30680
    },
    {
      "epoch": 16.082285115303982,
      "grad_norm": 0.6271968483924866,
      "learning_rate": 5.621532285843287e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 20059264,
      "step": 30685
    },
    {
      "epoch": 16.08490566037736,
      "grad_norm": 0.2592707574367523,
      "learning_rate": 5.614310233622119e-06,
      "loss": 0.0199,
      "num_input_tokens_seen": 20062112,
      "step": 30690
    },
    {
      "epoch": 16.087526205450732,
      "grad_norm": 0.53103107213974,
      "learning_rate": 5.607092236666093e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 20064896,
      "step": 30695
    },
    {
      "epoch": 16.09014675052411,
      "grad_norm": 1.1407427787780762,
      "learning_rate": 5.599878296485125e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 20067776,
      "step": 30700
    },
    {
      "epoch": 16.092767295597483,
      "grad_norm": 0.37056997418403625,
      "learning_rate": 5.592668414588284e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 20071072,
      "step": 30705
    },
    {
      "epoch": 16.09538784067086,
      "grad_norm": 0.39089834690093994,
      "learning_rate": 5.585462592483801e-06,
      "loss": 0.013,
      "num_input_tokens_seen": 20073920,
      "step": 30710
    },
    {
      "epoch": 16.098008385744233,
      "grad_norm": 1.3339329957962036,
      "learning_rate": 5.578260831679047e-06,
      "loss": 0.0129,
      "num_input_tokens_seen": 20077952,
      "step": 30715
    },
    {
      "epoch": 16.10062893081761,
      "grad_norm": 0.7620701193809509,
      "learning_rate": 5.571063133680562e-06,
      "loss": 0.023,
      "num_input_tokens_seen": 20081472,
      "step": 30720
    },
    {
      "epoch": 16.103249475890987,
      "grad_norm": 1.9903795719146729,
      "learning_rate": 5.563869499994004e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 20084128,
      "step": 30725
    },
    {
      "epoch": 16.10587002096436,
      "grad_norm": 1.0982877016067505,
      "learning_rate": 5.556679932124215e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 20087488,
      "step": 30730
    },
    {
      "epoch": 16.108490566037737,
      "grad_norm": 0.7021395564079285,
      "learning_rate": 5.549494431575159e-06,
      "loss": 0.0187,
      "num_input_tokens_seen": 20090656,
      "step": 30735
    },
    {
      "epoch": 16.11111111111111,
      "grad_norm": 0.8562108874320984,
      "learning_rate": 5.542312999849961e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 20093632,
      "step": 30740
    },
    {
      "epoch": 16.113731656184488,
      "grad_norm": 0.3986206650733948,
      "learning_rate": 5.5351356384509e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 20097312,
      "step": 30745
    },
    {
      "epoch": 16.11635220125786,
      "grad_norm": 0.5863860845565796,
      "learning_rate": 5.527962348879401e-06,
      "loss": 0.0213,
      "num_input_tokens_seen": 20100832,
      "step": 30750
    },
    {
      "epoch": 16.118972746331238,
      "grad_norm": 0.7749519944190979,
      "learning_rate": 5.520793132636027e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 20104800,
      "step": 30755
    },
    {
      "epoch": 16.12159329140461,
      "grad_norm": 1.0694981813430786,
      "learning_rate": 5.513627991220502e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 20108544,
      "step": 30760
    },
    {
      "epoch": 16.12421383647799,
      "grad_norm": 1.1885743141174316,
      "learning_rate": 5.506466926131684e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 20111680,
      "step": 30765
    },
    {
      "epoch": 16.12683438155136,
      "grad_norm": 0.332259863615036,
      "learning_rate": 5.4993099388675916e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 20115328,
      "step": 30770
    },
    {
      "epoch": 16.12945492662474,
      "grad_norm": 0.6958926320075989,
      "learning_rate": 5.49215703092539e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 20118752,
      "step": 30775
    },
    {
      "epoch": 16.132075471698112,
      "grad_norm": 0.8332066535949707,
      "learning_rate": 5.4850082038013755e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 20121504,
      "step": 30780
    },
    {
      "epoch": 16.13469601677149,
      "grad_norm": 0.515683650970459,
      "learning_rate": 5.4778634589910115e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 20125440,
      "step": 30785
    },
    {
      "epoch": 16.137316561844862,
      "grad_norm": 0.24882178008556366,
      "learning_rate": 5.4707227979888834e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 20128288,
      "step": 30790
    },
    {
      "epoch": 16.13993710691824,
      "grad_norm": 0.3668062090873718,
      "learning_rate": 5.463586222288747e-06,
      "loss": 0.0144,
      "num_input_tokens_seen": 20131328,
      "step": 30795
    },
    {
      "epoch": 16.142557651991613,
      "grad_norm": 0.2481415569782257,
      "learning_rate": 5.456453733383493e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 20135616,
      "step": 30800
    },
    {
      "epoch": 16.14517819706499,
      "grad_norm": 0.21640987694263458,
      "learning_rate": 5.449325332765151e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 20139040,
      "step": 30805
    },
    {
      "epoch": 16.147798742138363,
      "grad_norm": 0.3422851860523224,
      "learning_rate": 5.4422010219249e-06,
      "loss": 0.0151,
      "num_input_tokens_seen": 20141984,
      "step": 30810
    },
    {
      "epoch": 16.15041928721174,
      "grad_norm": 4.726858139038086,
      "learning_rate": 5.435080802353077e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 20144448,
      "step": 30815
    },
    {
      "epoch": 16.153039832285117,
      "grad_norm": 1.5375436544418335,
      "learning_rate": 5.4279646755391325e-06,
      "loss": 0.0285,
      "num_input_tokens_seen": 20150048,
      "step": 30820
    },
    {
      "epoch": 16.15566037735849,
      "grad_norm": 0.360013484954834,
      "learning_rate": 5.420852642971691e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 20153600,
      "step": 30825
    },
    {
      "epoch": 16.158280922431867,
      "grad_norm": 0.7715324759483337,
      "learning_rate": 5.41374470613851e-06,
      "loss": 0.0153,
      "num_input_tokens_seen": 20156448,
      "step": 30830
    },
    {
      "epoch": 16.16090146750524,
      "grad_norm": 0.7405257225036621,
      "learning_rate": 5.40664086652648e-06,
      "loss": 0.0287,
      "num_input_tokens_seen": 20160256,
      "step": 30835
    },
    {
      "epoch": 16.163522012578618,
      "grad_norm": 1.804426908493042,
      "learning_rate": 5.399541125621646e-06,
      "loss": 0.006,
      "num_input_tokens_seen": 20163168,
      "step": 30840
    },
    {
      "epoch": 16.16614255765199,
      "grad_norm": 0.5526701807975769,
      "learning_rate": 5.392445484909203e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 20166048,
      "step": 30845
    },
    {
      "epoch": 16.168763102725368,
      "grad_norm": 1.124669075012207,
      "learning_rate": 5.385353945873464e-06,
      "loss": 0.0229,
      "num_input_tokens_seen": 20169632,
      "step": 30850
    },
    {
      "epoch": 16.17138364779874,
      "grad_norm": 2.5595555305480957,
      "learning_rate": 5.378266509997909e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 20173920,
      "step": 30855
    },
    {
      "epoch": 16.17400419287212,
      "grad_norm": 0.29788610339164734,
      "learning_rate": 5.371183178765135e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 20176448,
      "step": 30860
    },
    {
      "epoch": 16.17662473794549,
      "grad_norm": 1.1216540336608887,
      "learning_rate": 5.364103953656904e-06,
      "loss": 0.0093,
      "num_input_tokens_seen": 20180448,
      "step": 30865
    },
    {
      "epoch": 16.17924528301887,
      "grad_norm": 1.029462456703186,
      "learning_rate": 5.35702883615411e-06,
      "loss": 0.0107,
      "num_input_tokens_seen": 20183232,
      "step": 30870
    },
    {
      "epoch": 16.181865828092242,
      "grad_norm": 2.706387758255005,
      "learning_rate": 5.34995782773679e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 20186240,
      "step": 30875
    },
    {
      "epoch": 16.18448637316562,
      "grad_norm": 0.860977828502655,
      "learning_rate": 5.3428909298841115e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 20189888,
      "step": 30880
    },
    {
      "epoch": 16.187106918238992,
      "grad_norm": 0.30756857991218567,
      "learning_rate": 5.335828144074384e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 20193440,
      "step": 30885
    },
    {
      "epoch": 16.18972746331237,
      "grad_norm": 1.5668531656265259,
      "learning_rate": 5.3287694717850695e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 20197088,
      "step": 30890
    },
    {
      "epoch": 16.192348008385743,
      "grad_norm": 0.6987273097038269,
      "learning_rate": 5.321714914492759e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 20200736,
      "step": 30895
    },
    {
      "epoch": 16.19496855345912,
      "grad_norm": 1.076208472251892,
      "learning_rate": 5.314664473673195e-06,
      "loss": 0.011,
      "num_input_tokens_seen": 20203712,
      "step": 30900
    },
    {
      "epoch": 16.197589098532493,
      "grad_norm": 0.7580291628837585,
      "learning_rate": 5.307618150801239e-06,
      "loss": 0.0136,
      "num_input_tokens_seen": 20206688,
      "step": 30905
    },
    {
      "epoch": 16.20020964360587,
      "grad_norm": 2.273745059967041,
      "learning_rate": 5.300575947350897e-06,
      "loss": 0.0373,
      "num_input_tokens_seen": 20210112,
      "step": 30910
    },
    {
      "epoch": 16.202830188679247,
      "grad_norm": 1.3619190454483032,
      "learning_rate": 5.293537864795325e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 20212768,
      "step": 30915
    },
    {
      "epoch": 16.20545073375262,
      "grad_norm": 0.8372661471366882,
      "learning_rate": 5.2865039046068075e-06,
      "loss": 0.0239,
      "num_input_tokens_seen": 20216928,
      "step": 30920
    },
    {
      "epoch": 16.208071278825997,
      "grad_norm": 0.08019724488258362,
      "learning_rate": 5.279474068256767e-06,
      "loss": 0.0016,
      "num_input_tokens_seen": 20219488,
      "step": 30925
    },
    {
      "epoch": 16.21069182389937,
      "grad_norm": 0.17974048852920532,
      "learning_rate": 5.272448357215779e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 20222592,
      "step": 30930
    },
    {
      "epoch": 16.213312368972748,
      "grad_norm": 0.40787258744239807,
      "learning_rate": 5.26542677295353e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 20227616,
      "step": 30935
    },
    {
      "epoch": 16.21593291404612,
      "grad_norm": 0.325705885887146,
      "learning_rate": 5.258409316938848e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 20230240,
      "step": 30940
    },
    {
      "epoch": 16.218553459119498,
      "grad_norm": 0.21761849522590637,
      "learning_rate": 5.251395990639713e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 20233120,
      "step": 30945
    },
    {
      "epoch": 16.22117400419287,
      "grad_norm": 0.3999236524105072,
      "learning_rate": 5.244386795523235e-06,
      "loss": 0.0191,
      "num_input_tokens_seen": 20236512,
      "step": 30950
    },
    {
      "epoch": 16.22379454926625,
      "grad_norm": 0.6222914457321167,
      "learning_rate": 5.23738173305566e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 20241024,
      "step": 30955
    },
    {
      "epoch": 16.22641509433962,
      "grad_norm": 0.3590126633644104,
      "learning_rate": 5.230380804702362e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 20244800,
      "step": 30960
    },
    {
      "epoch": 16.229035639413,
      "grad_norm": 1.0343515872955322,
      "learning_rate": 5.223384011927851e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 20249152,
      "step": 30965
    },
    {
      "epoch": 16.231656184486372,
      "grad_norm": 0.19126547873020172,
      "learning_rate": 5.216391356195779e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 20252512,
      "step": 30970
    },
    {
      "epoch": 16.23427672955975,
      "grad_norm": 0.3399553596973419,
      "learning_rate": 5.209402838968935e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 20256224,
      "step": 30975
    },
    {
      "epoch": 16.236897274633122,
      "grad_norm": 0.38276562094688416,
      "learning_rate": 5.202418461709233e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 20258656,
      "step": 30980
    },
    {
      "epoch": 16.2395178197065,
      "grad_norm": 1.452282428741455,
      "learning_rate": 5.195438225877738e-06,
      "loss": 0.0097,
      "num_input_tokens_seen": 20262528,
      "step": 30985
    },
    {
      "epoch": 16.242138364779873,
      "grad_norm": 1.3494656085968018,
      "learning_rate": 5.18846213293461e-06,
      "loss": 0.0132,
      "num_input_tokens_seen": 20266176,
      "step": 30990
    },
    {
      "epoch": 16.24475890985325,
      "grad_norm": 1.258037805557251,
      "learning_rate": 5.181490184339185e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 20269088,
      "step": 30995
    },
    {
      "epoch": 16.247379454926623,
      "grad_norm": 0.42290765047073364,
      "learning_rate": 5.17452238154991e-06,
      "loss": 0.0044,
      "num_input_tokens_seen": 20272384,
      "step": 31000
    },
    {
      "epoch": 16.25,
      "grad_norm": 3.0660979747772217,
      "learning_rate": 5.1675587260243715e-06,
      "loss": 0.038,
      "num_input_tokens_seen": 20276416,
      "step": 31005
    },
    {
      "epoch": 16.252620545073377,
      "grad_norm": 1.3805598020553589,
      "learning_rate": 5.160599219219295e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 20279904,
      "step": 31010
    },
    {
      "epoch": 16.25524109014675,
      "grad_norm": 1.1369562149047852,
      "learning_rate": 5.1536438625905185e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 20283200,
      "step": 31015
    },
    {
      "epoch": 16.257861635220127,
      "grad_norm": 0.08520563691854477,
      "learning_rate": 5.146692657593025e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 20286336,
      "step": 31020
    },
    {
      "epoch": 16.2604821802935,
      "grad_norm": 0.20125266909599304,
      "learning_rate": 5.139745605680926e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 20289024,
      "step": 31025
    },
    {
      "epoch": 16.263102725366878,
      "grad_norm": 0.2324071079492569,
      "learning_rate": 5.132802708307469e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 20291968,
      "step": 31030
    },
    {
      "epoch": 16.26572327044025,
      "grad_norm": 0.43681275844573975,
      "learning_rate": 5.125863966925035e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 20295072,
      "step": 31035
    },
    {
      "epoch": 16.268343815513628,
      "grad_norm": 1.5713019371032715,
      "learning_rate": 5.118929382985116e-06,
      "loss": 0.0084,
      "num_input_tokens_seen": 20298112,
      "step": 31040
    },
    {
      "epoch": 16.270964360587,
      "grad_norm": 1.8639346361160278,
      "learning_rate": 5.111998957938363e-06,
      "loss": 0.0113,
      "num_input_tokens_seen": 20301664,
      "step": 31045
    },
    {
      "epoch": 16.27358490566038,
      "grad_norm": 0.967265784740448,
      "learning_rate": 5.105072693234525e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 20304608,
      "step": 31050
    },
    {
      "epoch": 16.27620545073375,
      "grad_norm": 1.93880295753479,
      "learning_rate": 5.098150590322506e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 20307840,
      "step": 31055
    },
    {
      "epoch": 16.27882599580713,
      "grad_norm": 0.22018463909626007,
      "learning_rate": 5.091232650650335e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 20310784,
      "step": 31060
    },
    {
      "epoch": 16.281446540880502,
      "grad_norm": 0.07925049215555191,
      "learning_rate": 5.0843188756651645e-06,
      "loss": 0.0085,
      "num_input_tokens_seen": 20313344,
      "step": 31065
    },
    {
      "epoch": 16.28406708595388,
      "grad_norm": 0.3626765012741089,
      "learning_rate": 5.077409266813276e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 20316256,
      "step": 31070
    },
    {
      "epoch": 16.286687631027252,
      "grad_norm": 1.5821666717529297,
      "learning_rate": 5.0705038255400765e-06,
      "loss": 0.0163,
      "num_input_tokens_seen": 20320672,
      "step": 31075
    },
    {
      "epoch": 16.28930817610063,
      "grad_norm": 1.211371898651123,
      "learning_rate": 5.063602553290106e-06,
      "loss": 0.0139,
      "num_input_tokens_seen": 20324128,
      "step": 31080
    },
    {
      "epoch": 16.291928721174003,
      "grad_norm": 0.15771964192390442,
      "learning_rate": 5.056705451507035e-06,
      "loss": 0.0357,
      "num_input_tokens_seen": 20331520,
      "step": 31085
    },
    {
      "epoch": 16.29454926624738,
      "grad_norm": 0.11423321068286896,
      "learning_rate": 5.049812521633668e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 20334560,
      "step": 31090
    },
    {
      "epoch": 16.297169811320753,
      "grad_norm": 0.5316886305809021,
      "learning_rate": 5.042923765111907e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 20337856,
      "step": 31095
    },
    {
      "epoch": 16.29979035639413,
      "grad_norm": 0.823988676071167,
      "learning_rate": 5.036039183382821e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 20340224,
      "step": 31100
    },
    {
      "epoch": 16.302410901467507,
      "grad_norm": 0.9609619379043579,
      "learning_rate": 5.02915877788657e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 20343968,
      "step": 31105
    },
    {
      "epoch": 16.30503144654088,
      "grad_norm": 1.9894063472747803,
      "learning_rate": 5.0222825500624625e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 20348096,
      "step": 31110
    },
    {
      "epoch": 16.307651991614257,
      "grad_norm": 0.4512666165828705,
      "learning_rate": 5.015410501348935e-06,
      "loss": 0.0156,
      "num_input_tokens_seen": 20351072,
      "step": 31115
    },
    {
      "epoch": 16.31027253668763,
      "grad_norm": 0.2100561261177063,
      "learning_rate": 5.008542633183527e-06,
      "loss": 0.0091,
      "num_input_tokens_seen": 20354528,
      "step": 31120
    },
    {
      "epoch": 16.312893081761008,
      "grad_norm": 0.420125275850296,
      "learning_rate": 5.001678947002927e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 20357664,
      "step": 31125
    },
    {
      "epoch": 16.31551362683438,
      "grad_norm": 1.223799467086792,
      "learning_rate": 4.994819444242943e-06,
      "loss": 0.0125,
      "num_input_tokens_seen": 20361056,
      "step": 31130
    },
    {
      "epoch": 16.318134171907758,
      "grad_norm": 0.6334090828895569,
      "learning_rate": 4.9879641263384956e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 20364160,
      "step": 31135
    },
    {
      "epoch": 16.32075471698113,
      "grad_norm": 2.309889793395996,
      "learning_rate": 4.981112994723644e-06,
      "loss": 0.0161,
      "num_input_tokens_seen": 20367072,
      "step": 31140
    },
    {
      "epoch": 16.32337526205451,
      "grad_norm": 0.3439898192882538,
      "learning_rate": 4.974266050831572e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 20369600,
      "step": 31145
    },
    {
      "epoch": 16.32599580712788,
      "grad_norm": 0.4534197151660919,
      "learning_rate": 4.967423296094567e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 20372768,
      "step": 31150
    },
    {
      "epoch": 16.32861635220126,
      "grad_norm": 1.0355679988861084,
      "learning_rate": 4.960584731944076e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 20375680,
      "step": 31155
    },
    {
      "epoch": 16.331236897274632,
      "grad_norm": 0.1601647436618805,
      "learning_rate": 4.953750359810627e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 20378272,
      "step": 31160
    },
    {
      "epoch": 16.33385744234801,
      "grad_norm": 0.6903893351554871,
      "learning_rate": 4.946920181123904e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 20381888,
      "step": 31165
    },
    {
      "epoch": 16.336477987421382,
      "grad_norm": 0.44098398089408875,
      "learning_rate": 4.940094197312709e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 20385856,
      "step": 31170
    },
    {
      "epoch": 16.33909853249476,
      "grad_norm": 0.9457877278327942,
      "learning_rate": 4.933272409804946e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 20389120,
      "step": 31175
    },
    {
      "epoch": 16.341719077568133,
      "grad_norm": 0.41322559118270874,
      "learning_rate": 4.926454820027663e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 20392672,
      "step": 31180
    },
    {
      "epoch": 16.34433962264151,
      "grad_norm": 1.2281084060668945,
      "learning_rate": 4.919641429407024e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 20395232,
      "step": 31185
    },
    {
      "epoch": 16.346960167714883,
      "grad_norm": 0.24689291417598724,
      "learning_rate": 4.912832239368304e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 20398368,
      "step": 31190
    },
    {
      "epoch": 16.34958071278826,
      "grad_norm": 1.0662562847137451,
      "learning_rate": 4.906027251335918e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 20400768,
      "step": 31195
    },
    {
      "epoch": 16.352201257861637,
      "grad_norm": 0.564654529094696,
      "learning_rate": 4.899226466733381e-06,
      "loss": 0.0101,
      "num_input_tokens_seen": 20404448,
      "step": 31200
    },
    {
      "epoch": 16.35482180293501,
      "grad_norm": 0.2002851665019989,
      "learning_rate": 4.892429886983346e-06,
      "loss": 0.022,
      "num_input_tokens_seen": 20407264,
      "step": 31205
    },
    {
      "epoch": 16.357442348008387,
      "grad_norm": 1.0000271797180176,
      "learning_rate": 4.88563751350758e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 20410272,
      "step": 31210
    },
    {
      "epoch": 16.36006289308176,
      "grad_norm": 0.5678783655166626,
      "learning_rate": 4.878849347726977e-06,
      "loss": 0.0088,
      "num_input_tokens_seen": 20413280,
      "step": 31215
    },
    {
      "epoch": 16.362683438155138,
      "grad_norm": 0.5098105669021606,
      "learning_rate": 4.872065391061526e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 20415776,
      "step": 31220
    },
    {
      "epoch": 16.36530398322851,
      "grad_norm": 0.16833876073360443,
      "learning_rate": 4.865285644930373e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 20418944,
      "step": 31225
    },
    {
      "epoch": 16.367924528301888,
      "grad_norm": 0.6425389647483826,
      "learning_rate": 4.858510110751746e-06,
      "loss": 0.0288,
      "num_input_tokens_seen": 20422112,
      "step": 31230
    },
    {
      "epoch": 16.37054507337526,
      "grad_norm": 0.3308309316635132,
      "learning_rate": 4.851738789943019e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 20424736,
      "step": 31235
    },
    {
      "epoch": 16.37316561844864,
      "grad_norm": 0.18365992605686188,
      "learning_rate": 4.844971683920674e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 20428288,
      "step": 31240
    },
    {
      "epoch": 16.37578616352201,
      "grad_norm": 1.805504322052002,
      "learning_rate": 4.838208794100316e-06,
      "loss": 0.0297,
      "num_input_tokens_seen": 20432320,
      "step": 31245
    },
    {
      "epoch": 16.37840670859539,
      "grad_norm": 0.148810476064682,
      "learning_rate": 4.831450121896661e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 20434976,
      "step": 31250
    },
    {
      "epoch": 16.381027253668762,
      "grad_norm": 0.5830580592155457,
      "learning_rate": 4.824695668723542e-06,
      "loss": 0.0093,
      "num_input_tokens_seen": 20437664,
      "step": 31255
    },
    {
      "epoch": 16.38364779874214,
      "grad_norm": 1.2841904163360596,
      "learning_rate": 4.817945435993915e-06,
      "loss": 0.0096,
      "num_input_tokens_seen": 20440896,
      "step": 31260
    },
    {
      "epoch": 16.386268343815512,
      "grad_norm": 0.49149131774902344,
      "learning_rate": 4.811199425119853e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 20444064,
      "step": 31265
    },
    {
      "epoch": 16.38888888888889,
      "grad_norm": 0.7279044985771179,
      "learning_rate": 4.804457637512552e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 20446944,
      "step": 31270
    },
    {
      "epoch": 16.391509433962263,
      "grad_norm": 0.5190425515174866,
      "learning_rate": 4.79772007458231e-06,
      "loss": 0.0174,
      "num_input_tokens_seen": 20449888,
      "step": 31275
    },
    {
      "epoch": 16.39412997903564,
      "grad_norm": 0.7223660945892334,
      "learning_rate": 4.7909867377385424e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 20453824,
      "step": 31280
    },
    {
      "epoch": 16.396750524109013,
      "grad_norm": 1.0030215978622437,
      "learning_rate": 4.784257628389794e-06,
      "loss": 0.0686,
      "num_input_tokens_seen": 20457184,
      "step": 31285
    },
    {
      "epoch": 16.39937106918239,
      "grad_norm": 5.695817947387695,
      "learning_rate": 4.7775327479437166e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 20460864,
      "step": 31290
    },
    {
      "epoch": 16.401991614255767,
      "grad_norm": 0.3605267405509949,
      "learning_rate": 4.770812097807076e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 20463392,
      "step": 31295
    },
    {
      "epoch": 16.40461215932914,
      "grad_norm": 2.46623158454895,
      "learning_rate": 4.764095679385766e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 20466432,
      "step": 31300
    },
    {
      "epoch": 16.407232704402517,
      "grad_norm": 2.0716536045074463,
      "learning_rate": 4.7573834940847745e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 20469312,
      "step": 31305
    },
    {
      "epoch": 16.40985324947589,
      "grad_norm": 0.750223696231842,
      "learning_rate": 4.750675543308212e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 20471936,
      "step": 31310
    },
    {
      "epoch": 16.412473794549268,
      "grad_norm": 0.4148857295513153,
      "learning_rate": 4.743971828459307e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 20474720,
      "step": 31315
    },
    {
      "epoch": 16.41509433962264,
      "grad_norm": 0.6468204855918884,
      "learning_rate": 4.7372723509404025e-06,
      "loss": 0.0273,
      "num_input_tokens_seen": 20478784,
      "step": 31320
    },
    {
      "epoch": 16.417714884696018,
      "grad_norm": 0.6576045751571655,
      "learning_rate": 4.7305771121529615e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 20481824,
      "step": 31325
    },
    {
      "epoch": 16.42033542976939,
      "grad_norm": 0.17122432589530945,
      "learning_rate": 4.723886113497542e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 20484704,
      "step": 31330
    },
    {
      "epoch": 16.42295597484277,
      "grad_norm": 1.0803879499435425,
      "learning_rate": 4.7171993563738195e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 20487936,
      "step": 31335
    },
    {
      "epoch": 16.42557651991614,
      "grad_norm": 0.45841288566589355,
      "learning_rate": 4.7105168421805924e-06,
      "loss": 0.0065,
      "num_input_tokens_seen": 20491264,
      "step": 31340
    },
    {
      "epoch": 16.42819706498952,
      "grad_norm": 2.476886034011841,
      "learning_rate": 4.703838572315767e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 20494112,
      "step": 31345
    },
    {
      "epoch": 16.430817610062892,
      "grad_norm": 0.28139829635620117,
      "learning_rate": 4.697164548176367e-06,
      "loss": 0.0137,
      "num_input_tokens_seen": 20497184,
      "step": 31350
    },
    {
      "epoch": 16.43343815513627,
      "grad_norm": 0.27275800704956055,
      "learning_rate": 4.690494771158519e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 20500864,
      "step": 31355
    },
    {
      "epoch": 16.436058700209642,
      "grad_norm": 0.15698879957199097,
      "learning_rate": 4.683829242657453e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 20504928,
      "step": 31360
    },
    {
      "epoch": 16.43867924528302,
      "grad_norm": 0.32152190804481506,
      "learning_rate": 4.677167964067533e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 20507776,
      "step": 31365
    },
    {
      "epoch": 16.441299790356393,
      "grad_norm": 0.3477771282196045,
      "learning_rate": 4.670510936782216e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 20510944,
      "step": 31370
    },
    {
      "epoch": 16.44392033542977,
      "grad_norm": 0.3025587499141693,
      "learning_rate": 4.6638581621940815e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 20514784,
      "step": 31375
    },
    {
      "epoch": 16.446540880503143,
      "grad_norm": 1.6814371347427368,
      "learning_rate": 4.657209641694821e-06,
      "loss": 0.0161,
      "num_input_tokens_seen": 20517248,
      "step": 31380
    },
    {
      "epoch": 16.44916142557652,
      "grad_norm": 2.557332992553711,
      "learning_rate": 4.65056537667522e-06,
      "loss": 0.0146,
      "num_input_tokens_seen": 20520512,
      "step": 31385
    },
    {
      "epoch": 16.451781970649897,
      "grad_norm": 0.17471663653850555,
      "learning_rate": 4.643925368525176e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 20523328,
      "step": 31390
    },
    {
      "epoch": 16.45440251572327,
      "grad_norm": 1.883101463317871,
      "learning_rate": 4.637289618633711e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 20527296,
      "step": 31395
    },
    {
      "epoch": 16.457023060796647,
      "grad_norm": 0.21906305849552155,
      "learning_rate": 4.630658128388948e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 20529952,
      "step": 31400
    },
    {
      "epoch": 16.45964360587002,
      "grad_norm": 0.5150511860847473,
      "learning_rate": 4.624030899178125e-06,
      "loss": 0.0136,
      "num_input_tokens_seen": 20533760,
      "step": 31405
    },
    {
      "epoch": 16.462264150943398,
      "grad_norm": 0.046131398528814316,
      "learning_rate": 4.617407932387569e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 20536384,
      "step": 31410
    },
    {
      "epoch": 16.46488469601677,
      "grad_norm": 0.8980029225349426,
      "learning_rate": 4.610789229402743e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 20539744,
      "step": 31415
    },
    {
      "epoch": 16.467505241090148,
      "grad_norm": 0.48345452547073364,
      "learning_rate": 4.604174791608193e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 20543904,
      "step": 31420
    },
    {
      "epoch": 16.47012578616352,
      "grad_norm": 1.8176653385162354,
      "learning_rate": 4.597564620387587e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 20547552,
      "step": 31425
    },
    {
      "epoch": 16.4727463312369,
      "grad_norm": 4.669471740722656,
      "learning_rate": 4.590958717123708e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 20549984,
      "step": 31430
    },
    {
      "epoch": 16.47536687631027,
      "grad_norm": 0.4575619101524353,
      "learning_rate": 4.5843570831984174e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 20553280,
      "step": 31435
    },
    {
      "epoch": 16.47798742138365,
      "grad_norm": 0.8789728879928589,
      "learning_rate": 4.577759719992719e-06,
      "loss": 0.0163,
      "num_input_tokens_seen": 20557376,
      "step": 31440
    },
    {
      "epoch": 16.480607966457022,
      "grad_norm": 2.836552143096924,
      "learning_rate": 4.571166628886692e-06,
      "loss": 0.0264,
      "num_input_tokens_seen": 20560288,
      "step": 31445
    },
    {
      "epoch": 16.4832285115304,
      "grad_norm": 0.30806756019592285,
      "learning_rate": 4.564577811259541e-06,
      "loss": 0.0175,
      "num_input_tokens_seen": 20563392,
      "step": 31450
    },
    {
      "epoch": 16.485849056603772,
      "grad_norm": 2.933441638946533,
      "learning_rate": 4.557993268489574e-06,
      "loss": 0.0222,
      "num_input_tokens_seen": 20566272,
      "step": 31455
    },
    {
      "epoch": 16.48846960167715,
      "grad_norm": 2.8618385791778564,
      "learning_rate": 4.551413001954205e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 20568672,
      "step": 31460
    },
    {
      "epoch": 16.491090146750523,
      "grad_norm": 0.7212117314338684,
      "learning_rate": 4.54483701302994e-06,
      "loss": 0.0184,
      "num_input_tokens_seen": 20572096,
      "step": 31465
    },
    {
      "epoch": 16.4937106918239,
      "grad_norm": 1.337409496307373,
      "learning_rate": 4.538265303092415e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 20575008,
      "step": 31470
    },
    {
      "epoch": 16.496331236897273,
      "grad_norm": 0.8211382627487183,
      "learning_rate": 4.531697873516338e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 20578304,
      "step": 31475
    },
    {
      "epoch": 16.49895178197065,
      "grad_norm": 0.5636125802993774,
      "learning_rate": 4.525134725675556e-06,
      "loss": 0.0134,
      "num_input_tokens_seen": 20581248,
      "step": 31480
    },
    {
      "epoch": 16.501572327044027,
      "grad_norm": 0.3479228615760803,
      "learning_rate": 4.518575860943003e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 20584032,
      "step": 31485
    },
    {
      "epoch": 16.5041928721174,
      "grad_norm": 0.39453962445259094,
      "learning_rate": 4.512021280690712e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 20587872,
      "step": 31490
    },
    {
      "epoch": 16.506813417190777,
      "grad_norm": 0.732378363609314,
      "learning_rate": 4.505470986289828e-06,
      "loss": 0.0131,
      "num_input_tokens_seen": 20590880,
      "step": 31495
    },
    {
      "epoch": 16.50943396226415,
      "grad_norm": 0.7918108105659485,
      "learning_rate": 4.498924979110605e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 20593792,
      "step": 31500
    },
    {
      "epoch": 16.512054507337528,
      "grad_norm": 1.5910744667053223,
      "learning_rate": 4.492383260522384e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 20596992,
      "step": 31505
    },
    {
      "epoch": 16.5146750524109,
      "grad_norm": 0.49992427229881287,
      "learning_rate": 4.485845831893626e-06,
      "loss": 0.0111,
      "num_input_tokens_seen": 20599808,
      "step": 31510
    },
    {
      "epoch": 16.517295597484278,
      "grad_norm": 0.45317918062210083,
      "learning_rate": 4.479312694591878e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 20603136,
      "step": 31515
    },
    {
      "epoch": 16.51991614255765,
      "grad_norm": 0.16749341785907745,
      "learning_rate": 4.4727838499838e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 20605792,
      "step": 31520
    },
    {
      "epoch": 16.52253668763103,
      "grad_norm": 0.3501284122467041,
      "learning_rate": 4.4662592994351545e-06,
      "loss": 0.0065,
      "num_input_tokens_seen": 20608928,
      "step": 31525
    },
    {
      "epoch": 16.5251572327044,
      "grad_norm": 0.13059499859809875,
      "learning_rate": 4.4597390443108065e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 20611680,
      "step": 31530
    },
    {
      "epoch": 16.52777777777778,
      "grad_norm": 0.5393795967102051,
      "learning_rate": 4.453223085974712e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 20614528,
      "step": 31535
    },
    {
      "epoch": 16.530398322851152,
      "grad_norm": 2.294525146484375,
      "learning_rate": 4.4467114257899425e-06,
      "loss": 0.0088,
      "num_input_tokens_seen": 20617984,
      "step": 31540
    },
    {
      "epoch": 16.53301886792453,
      "grad_norm": 0.4780781865119934,
      "learning_rate": 4.44020406511865e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 20621952,
      "step": 31545
    },
    {
      "epoch": 16.535639412997902,
      "grad_norm": 0.35939738154411316,
      "learning_rate": 4.43370100532211e-06,
      "loss": 0.0169,
      "num_input_tokens_seen": 20624416,
      "step": 31550
    },
    {
      "epoch": 16.53825995807128,
      "grad_norm": 1.0881109237670898,
      "learning_rate": 4.427202247760695e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 20627456,
      "step": 31555
    },
    {
      "epoch": 16.540880503144653,
      "grad_norm": 0.6447301506996155,
      "learning_rate": 4.420707793793854e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 20630688,
      "step": 31560
    },
    {
      "epoch": 16.54350104821803,
      "grad_norm": 1.7358537912368774,
      "learning_rate": 4.414217644780172e-06,
      "loss": 0.0247,
      "num_input_tokens_seen": 20635552,
      "step": 31565
    },
    {
      "epoch": 16.546121593291403,
      "grad_norm": 0.5180662274360657,
      "learning_rate": 4.407731802077294e-06,
      "loss": 0.0102,
      "num_input_tokens_seen": 20638272,
      "step": 31570
    },
    {
      "epoch": 16.54874213836478,
      "grad_norm": 0.23427347838878632,
      "learning_rate": 4.401250267041996e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 20641536,
      "step": 31575
    },
    {
      "epoch": 16.551362683438157,
      "grad_norm": 2.687893867492676,
      "learning_rate": 4.394773041030137e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 20644864,
      "step": 31580
    },
    {
      "epoch": 16.55398322851153,
      "grad_norm": 0.16128042340278625,
      "learning_rate": 4.388300125396685e-06,
      "loss": 0.019,
      "num_input_tokens_seen": 20648224,
      "step": 31585
    },
    {
      "epoch": 16.556603773584907,
      "grad_norm": 1.6002850532531738,
      "learning_rate": 4.381831521495699e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 20652800,
      "step": 31590
    },
    {
      "epoch": 16.55922431865828,
      "grad_norm": 0.19259966909885406,
      "learning_rate": 4.37536723068033e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 20655776,
      "step": 31595
    },
    {
      "epoch": 16.561844863731658,
      "grad_norm": 1.3718762397766113,
      "learning_rate": 4.368907254302837e-06,
      "loss": 0.0101,
      "num_input_tokens_seen": 20659040,
      "step": 31600
    },
    {
      "epoch": 16.56446540880503,
      "grad_norm": 2.1189889907836914,
      "learning_rate": 4.3624515937145725e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 20662560,
      "step": 31605
    },
    {
      "epoch": 16.567085953878408,
      "grad_norm": 0.3923383057117462,
      "learning_rate": 4.3560002502659905e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 20664992,
      "step": 31610
    },
    {
      "epoch": 16.56970649895178,
      "grad_norm": 0.27288758754730225,
      "learning_rate": 4.349553225306644e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 20667488,
      "step": 31615
    },
    {
      "epoch": 16.572327044025158,
      "grad_norm": 5.853396415710449,
      "learning_rate": 4.343110520185167e-06,
      "loss": 0.0096,
      "num_input_tokens_seen": 20670304,
      "step": 31620
    },
    {
      "epoch": 16.57494758909853,
      "grad_norm": 0.7136334180831909,
      "learning_rate": 4.336672136249298e-06,
      "loss": 0.0181,
      "num_input_tokens_seen": 20674080,
      "step": 31625
    },
    {
      "epoch": 16.57756813417191,
      "grad_norm": 0.20901943743228912,
      "learning_rate": 4.33023807484588e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 20676928,
      "step": 31630
    },
    {
      "epoch": 16.580188679245282,
      "grad_norm": 0.6866011619567871,
      "learning_rate": 4.323808337320842e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 20680096,
      "step": 31635
    },
    {
      "epoch": 16.58280922431866,
      "grad_norm": 2.1397805213928223,
      "learning_rate": 4.31738292501922e-06,
      "loss": 0.0219,
      "num_input_tokens_seen": 20683360,
      "step": 31640
    },
    {
      "epoch": 16.585429769392032,
      "grad_norm": 1.397128939628601,
      "learning_rate": 4.31096183928513e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 20685920,
      "step": 31645
    },
    {
      "epoch": 16.58805031446541,
      "grad_norm": 8.295083045959473,
      "learning_rate": 4.304545081461783e-06,
      "loss": 0.0151,
      "num_input_tokens_seen": 20688576,
      "step": 31650
    },
    {
      "epoch": 16.590670859538783,
      "grad_norm": 0.17071348428726196,
      "learning_rate": 4.298132652891501e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 20691424,
      "step": 31655
    },
    {
      "epoch": 16.59329140461216,
      "grad_norm": 6.61964225769043,
      "learning_rate": 4.291724554915688e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 20694432,
      "step": 31660
    },
    {
      "epoch": 16.595911949685533,
      "grad_norm": 0.713564395904541,
      "learning_rate": 4.285320788874853e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 20697856,
      "step": 31665
    },
    {
      "epoch": 16.59853249475891,
      "grad_norm": 0.6833146810531616,
      "learning_rate": 4.278921356108581e-06,
      "loss": 0.0084,
      "num_input_tokens_seen": 20701504,
      "step": 31670
    },
    {
      "epoch": 16.601153039832283,
      "grad_norm": 0.08971639722585678,
      "learning_rate": 4.272526257955556e-06,
      "loss": 0.0106,
      "num_input_tokens_seen": 20705056,
      "step": 31675
    },
    {
      "epoch": 16.60377358490566,
      "grad_norm": 0.26183372735977173,
      "learning_rate": 4.266135495753567e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 20707872,
      "step": 31680
    },
    {
      "epoch": 16.606394129979037,
      "grad_norm": 0.21492214500904083,
      "learning_rate": 4.259749070839491e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 20710560,
      "step": 31685
    },
    {
      "epoch": 16.60901467505241,
      "grad_norm": 1.5811998844146729,
      "learning_rate": 4.253366984549289e-06,
      "loss": 0.0617,
      "num_input_tokens_seen": 20714016,
      "step": 31690
    },
    {
      "epoch": 16.611635220125788,
      "grad_norm": 0.11259902268648148,
      "learning_rate": 4.24698923821803e-06,
      "loss": 0.0093,
      "num_input_tokens_seen": 20717120,
      "step": 31695
    },
    {
      "epoch": 16.61425576519916,
      "grad_norm": 0.16574779152870178,
      "learning_rate": 4.240615833179861e-06,
      "loss": 0.0114,
      "num_input_tokens_seen": 20720160,
      "step": 31700
    },
    {
      "epoch": 16.616876310272538,
      "grad_norm": 0.392354279756546,
      "learning_rate": 4.234246770768016e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 20723328,
      "step": 31705
    },
    {
      "epoch": 16.61949685534591,
      "grad_norm": 0.6008899211883545,
      "learning_rate": 4.227882052314841e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 20725632,
      "step": 31710
    },
    {
      "epoch": 16.622117400419288,
      "grad_norm": 2.2040011882781982,
      "learning_rate": 4.2215216791517575e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 20729600,
      "step": 31715
    },
    {
      "epoch": 16.62473794549266,
      "grad_norm": 1.5205656290054321,
      "learning_rate": 4.215165652609293e-06,
      "loss": 0.024,
      "num_input_tokens_seen": 20732512,
      "step": 31720
    },
    {
      "epoch": 16.62735849056604,
      "grad_norm": 0.18293477594852448,
      "learning_rate": 4.208813974017048e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 20736544,
      "step": 31725
    },
    {
      "epoch": 16.629979035639412,
      "grad_norm": 0.25040921568870544,
      "learning_rate": 4.202466644703712e-06,
      "loss": 0.0461,
      "num_input_tokens_seen": 20739968,
      "step": 31730
    },
    {
      "epoch": 16.63259958071279,
      "grad_norm": 0.3407100439071655,
      "learning_rate": 4.196123665997087e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 20743296,
      "step": 31735
    },
    {
      "epoch": 16.635220125786162,
      "grad_norm": 0.34698575735092163,
      "learning_rate": 4.189785039224045e-06,
      "loss": 0.0103,
      "num_input_tokens_seen": 20745696,
      "step": 31740
    },
    {
      "epoch": 16.63784067085954,
      "grad_norm": 1.4341939687728882,
      "learning_rate": 4.183450765710567e-06,
      "loss": 0.0373,
      "num_input_tokens_seen": 20750272,
      "step": 31745
    },
    {
      "epoch": 16.640461215932913,
      "grad_norm": 0.354428768157959,
      "learning_rate": 4.177120846781693e-06,
      "loss": 0.018,
      "num_input_tokens_seen": 20754112,
      "step": 31750
    },
    {
      "epoch": 16.64308176100629,
      "grad_norm": 1.0536433458328247,
      "learning_rate": 4.170795283761583e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 20756928,
      "step": 31755
    },
    {
      "epoch": 16.645702306079663,
      "grad_norm": 0.41578420996665955,
      "learning_rate": 4.164474077973462e-06,
      "loss": 0.0771,
      "num_input_tokens_seen": 20759968,
      "step": 31760
    },
    {
      "epoch": 16.64832285115304,
      "grad_norm": 0.35220322012901306,
      "learning_rate": 4.158157230739659e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 20762560,
      "step": 31765
    },
    {
      "epoch": 16.650943396226417,
      "grad_norm": 0.1693086475133896,
      "learning_rate": 4.151844743381589e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 20765120,
      "step": 31770
    },
    {
      "epoch": 16.65356394129979,
      "grad_norm": 0.2646382749080658,
      "learning_rate": 4.145536617219756e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 20767840,
      "step": 31775
    },
    {
      "epoch": 16.656184486373167,
      "grad_norm": 0.5829858183860779,
      "learning_rate": 4.139232853573735e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 20771200,
      "step": 31780
    },
    {
      "epoch": 16.65880503144654,
      "grad_norm": 0.16605328023433685,
      "learning_rate": 4.132933453762214e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 20774336,
      "step": 31785
    },
    {
      "epoch": 16.661425576519918,
      "grad_norm": 0.5759866237640381,
      "learning_rate": 4.126638419102946e-06,
      "loss": 0.0181,
      "num_input_tokens_seen": 20777248,
      "step": 31790
    },
    {
      "epoch": 16.66404612159329,
      "grad_norm": 0.7179591059684753,
      "learning_rate": 4.120347750912784e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 20780288,
      "step": 31795
    },
    {
      "epoch": 16.666666666666668,
      "grad_norm": 2.725431442260742,
      "learning_rate": 4.1140614505076724e-06,
      "loss": 0.0327,
      "num_input_tokens_seen": 20782848,
      "step": 31800
    },
    {
      "epoch": 16.66928721174004,
      "grad_norm": 4.358381748199463,
      "learning_rate": 4.10777951920262e-06,
      "loss": 0.0288,
      "num_input_tokens_seen": 20787168,
      "step": 31805
    },
    {
      "epoch": 16.671907756813418,
      "grad_norm": 0.31468909978866577,
      "learning_rate": 4.1015019583117455e-06,
      "loss": 0.0114,
      "num_input_tokens_seen": 20790176,
      "step": 31810
    },
    {
      "epoch": 16.67452830188679,
      "grad_norm": 0.15838439762592316,
      "learning_rate": 4.095228769148232e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 20792864,
      "step": 31815
    },
    {
      "epoch": 16.67714884696017,
      "grad_norm": 0.8289487957954407,
      "learning_rate": 4.088959953024368e-06,
      "loss": 0.0127,
      "num_input_tokens_seen": 20797408,
      "step": 31820
    },
    {
      "epoch": 16.679769392033542,
      "grad_norm": 0.15091072022914886,
      "learning_rate": 4.082695511251522e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 20800128,
      "step": 31825
    },
    {
      "epoch": 16.68238993710692,
      "grad_norm": 0.11671854555606842,
      "learning_rate": 4.076435445140131e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 20803008,
      "step": 31830
    },
    {
      "epoch": 16.685010482180292,
      "grad_norm": 0.961989164352417,
      "learning_rate": 4.070179755999737e-06,
      "loss": 0.0445,
      "num_input_tokens_seen": 20806144,
      "step": 31835
    },
    {
      "epoch": 16.68763102725367,
      "grad_norm": 0.780023992061615,
      "learning_rate": 4.063928445138962e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 20809472,
      "step": 31840
    },
    {
      "epoch": 16.690251572327043,
      "grad_norm": 0.7817131876945496,
      "learning_rate": 4.057681513865499e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 20812288,
      "step": 31845
    },
    {
      "epoch": 16.69287211740042,
      "grad_norm": 1.39159095287323,
      "learning_rate": 4.0514389634861426e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 20815296,
      "step": 31850
    },
    {
      "epoch": 16.695492662473793,
      "grad_norm": 1.5187498331069946,
      "learning_rate": 4.045200795306767e-06,
      "loss": 0.0313,
      "num_input_tokens_seen": 20822656,
      "step": 31855
    },
    {
      "epoch": 16.69811320754717,
      "grad_norm": 0.36913490295410156,
      "learning_rate": 4.038967010632314e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 20826016,
      "step": 31860
    },
    {
      "epoch": 16.700733752620543,
      "grad_norm": 0.6105825901031494,
      "learning_rate": 4.032737610766826e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 20829824,
      "step": 31865
    },
    {
      "epoch": 16.70335429769392,
      "grad_norm": 0.7155320048332214,
      "learning_rate": 4.026512597013429e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 20832512,
      "step": 31870
    },
    {
      "epoch": 16.705974842767297,
      "grad_norm": 0.9372411966323853,
      "learning_rate": 4.020291970674312e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 20836000,
      "step": 31875
    },
    {
      "epoch": 16.70859538784067,
      "grad_norm": 0.20095908641815186,
      "learning_rate": 4.0140757330507705e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 20838976,
      "step": 31880
    },
    {
      "epoch": 16.711215932914047,
      "grad_norm": 2.382290840148926,
      "learning_rate": 4.007863885443161e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 20842208,
      "step": 31885
    },
    {
      "epoch": 16.71383647798742,
      "grad_norm": 0.5364439487457275,
      "learning_rate": 4.001656429150935e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 20846272,
      "step": 31890
    },
    {
      "epoch": 16.716457023060798,
      "grad_norm": 0.3963487446308136,
      "learning_rate": 3.9954533654726235e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 20848896,
      "step": 31895
    },
    {
      "epoch": 16.71907756813417,
      "grad_norm": 0.39778852462768555,
      "learning_rate": 3.989254695705841e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 20852288,
      "step": 31900
    },
    {
      "epoch": 16.721698113207548,
      "grad_norm": 0.38050222396850586,
      "learning_rate": 3.983060421147272e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 20855136,
      "step": 31905
    },
    {
      "epoch": 16.72431865828092,
      "grad_norm": 2.5780956745147705,
      "learning_rate": 3.976870543092684e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 20858432,
      "step": 31910
    },
    {
      "epoch": 16.7269392033543,
      "grad_norm": 0.14039507508277893,
      "learning_rate": 3.970685062836932e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 20861280,
      "step": 31915
    },
    {
      "epoch": 16.729559748427672,
      "grad_norm": 0.2628186047077179,
      "learning_rate": 3.96450398167395e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 20863968,
      "step": 31920
    },
    {
      "epoch": 16.73218029350105,
      "grad_norm": 1.219728946685791,
      "learning_rate": 3.958327300896758e-06,
      "loss": 0.0204,
      "num_input_tokens_seen": 20866112,
      "step": 31925
    },
    {
      "epoch": 16.734800838574422,
      "grad_norm": 0.5874159336090088,
      "learning_rate": 3.952155021797429e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 20869856,
      "step": 31930
    },
    {
      "epoch": 16.7374213836478,
      "grad_norm": 0.21840237081050873,
      "learning_rate": 3.945987145667151e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 20872448,
      "step": 31935
    },
    {
      "epoch": 16.740041928721173,
      "grad_norm": 0.5989457964897156,
      "learning_rate": 3.939823673796162e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 20875008,
      "step": 31940
    },
    {
      "epoch": 16.74266247379455,
      "grad_norm": 2.319122552871704,
      "learning_rate": 3.933664607473789e-06,
      "loss": 0.0207,
      "num_input_tokens_seen": 20878144,
      "step": 31945
    },
    {
      "epoch": 16.745283018867923,
      "grad_norm": 0.6592478156089783,
      "learning_rate": 3.927509947988445e-06,
      "loss": 0.0094,
      "num_input_tokens_seen": 20881216,
      "step": 31950
    },
    {
      "epoch": 16.7479035639413,
      "grad_norm": 1.5888628959655762,
      "learning_rate": 3.9213596966276164e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 20885120,
      "step": 31955
    },
    {
      "epoch": 16.750524109014677,
      "grad_norm": 0.7226315140724182,
      "learning_rate": 3.9152138546778625e-06,
      "loss": 0.028,
      "num_input_tokens_seen": 20887808,
      "step": 31960
    },
    {
      "epoch": 16.75314465408805,
      "grad_norm": 1.522025465965271,
      "learning_rate": 3.9090724234248175e-06,
      "loss": 0.0131,
      "num_input_tokens_seen": 20890720,
      "step": 31965
    },
    {
      "epoch": 16.755765199161427,
      "grad_norm": 0.9218038320541382,
      "learning_rate": 3.902935404153202e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 20893440,
      "step": 31970
    },
    {
      "epoch": 16.7583857442348,
      "grad_norm": 0.39323291182518005,
      "learning_rate": 3.896802798146809e-06,
      "loss": 0.016,
      "num_input_tokens_seen": 20897792,
      "step": 31975
    },
    {
      "epoch": 16.761006289308177,
      "grad_norm": 1.6909393072128296,
      "learning_rate": 3.890674606688524e-06,
      "loss": 0.0141,
      "num_input_tokens_seen": 20899936,
      "step": 31980
    },
    {
      "epoch": 16.76362683438155,
      "grad_norm": 0.6725255250930786,
      "learning_rate": 3.88455083106028e-06,
      "loss": 0.0101,
      "num_input_tokens_seen": 20905728,
      "step": 31985
    },
    {
      "epoch": 16.766247379454928,
      "grad_norm": 2.903658151626587,
      "learning_rate": 3.878431472543098e-06,
      "loss": 0.0458,
      "num_input_tokens_seen": 20909376,
      "step": 31990
    },
    {
      "epoch": 16.7688679245283,
      "grad_norm": 1.318692922592163,
      "learning_rate": 3.872316532417086e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 20913088,
      "step": 31995
    },
    {
      "epoch": 16.771488469601678,
      "grad_norm": 1.6658583879470825,
      "learning_rate": 3.866206011961418e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 20915808,
      "step": 32000
    },
    {
      "epoch": 16.77410901467505,
      "grad_norm": 0.6228649020195007,
      "learning_rate": 3.860099912454346e-06,
      "loss": 0.0214,
      "num_input_tokens_seen": 20920032,
      "step": 32005
    },
    {
      "epoch": 16.77672955974843,
      "grad_norm": 1.098249077796936,
      "learning_rate": 3.853998235173201e-06,
      "loss": 0.0044,
      "num_input_tokens_seen": 20922912,
      "step": 32010
    },
    {
      "epoch": 16.779350104821802,
      "grad_norm": 1.5305598974227905,
      "learning_rate": 3.847900981394379e-06,
      "loss": 0.0146,
      "num_input_tokens_seen": 20926176,
      "step": 32015
    },
    {
      "epoch": 16.78197064989518,
      "grad_norm": 0.6355205178260803,
      "learning_rate": 3.841808152393348e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 20928896,
      "step": 32020
    },
    {
      "epoch": 16.784591194968552,
      "grad_norm": 1.1213688850402832,
      "learning_rate": 3.835719749444666e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 20931968,
      "step": 32025
    },
    {
      "epoch": 16.78721174004193,
      "grad_norm": 1.2502162456512451,
      "learning_rate": 3.8296357738219606e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 20935168,
      "step": 32030
    },
    {
      "epoch": 16.789832285115303,
      "grad_norm": 0.230081245303154,
      "learning_rate": 3.823556226797931e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 20937856,
      "step": 32035
    },
    {
      "epoch": 16.79245283018868,
      "grad_norm": 8.644547462463379,
      "learning_rate": 3.817481109644344e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 20941120,
      "step": 32040
    },
    {
      "epoch": 16.795073375262053,
      "grad_norm": 0.6859988570213318,
      "learning_rate": 3.811410423632042e-06,
      "loss": 0.0094,
      "num_input_tokens_seen": 20944416,
      "step": 32045
    },
    {
      "epoch": 16.79769392033543,
      "grad_norm": 0.28318673372268677,
      "learning_rate": 3.8053441700309473e-06,
      "loss": 0.0159,
      "num_input_tokens_seen": 20948384,
      "step": 32050
    },
    {
      "epoch": 16.800314465408803,
      "grad_norm": 0.085172139108181,
      "learning_rate": 3.7992823501100507e-06,
      "loss": 0.0232,
      "num_input_tokens_seen": 20951328,
      "step": 32055
    },
    {
      "epoch": 16.80293501048218,
      "grad_norm": 0.3057703375816345,
      "learning_rate": 3.7932249651374207e-06,
      "loss": 0.0131,
      "num_input_tokens_seen": 20954304,
      "step": 32060
    },
    {
      "epoch": 16.805555555555557,
      "grad_norm": 0.20598076283931732,
      "learning_rate": 3.787172016380183e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 20957600,
      "step": 32065
    },
    {
      "epoch": 16.80817610062893,
      "grad_norm": 0.8395466804504395,
      "learning_rate": 3.7811235051045558e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 20961248,
      "step": 32070
    },
    {
      "epoch": 16.810796645702307,
      "grad_norm": 0.20154649019241333,
      "learning_rate": 3.775079432575809e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 20964864,
      "step": 32075
    },
    {
      "epoch": 16.81341719077568,
      "grad_norm": 0.343783438205719,
      "learning_rate": 3.7690398000582964e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 20968160,
      "step": 32080
    },
    {
      "epoch": 16.816037735849058,
      "grad_norm": 0.25690823793411255,
      "learning_rate": 3.763004608815443e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 20971072,
      "step": 32085
    },
    {
      "epoch": 16.81865828092243,
      "grad_norm": 4.148865699768066,
      "learning_rate": 3.756973860109747e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 20974016,
      "step": 32090
    },
    {
      "epoch": 16.821278825995808,
      "grad_norm": 1.1595189571380615,
      "learning_rate": 3.7509475552027663e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 20977280,
      "step": 32095
    },
    {
      "epoch": 16.82389937106918,
      "grad_norm": 1.6870908737182617,
      "learning_rate": 3.744925695355131e-06,
      "loss": 0.0142,
      "num_input_tokens_seen": 20980704,
      "step": 32100
    },
    {
      "epoch": 16.82651991614256,
      "grad_norm": 1.1119664907455444,
      "learning_rate": 3.7389082818265476e-06,
      "loss": 0.0771,
      "num_input_tokens_seen": 20983552,
      "step": 32105
    },
    {
      "epoch": 16.829140461215932,
      "grad_norm": 0.3971253037452698,
      "learning_rate": 3.732895315875792e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 20986464,
      "step": 32110
    },
    {
      "epoch": 16.83176100628931,
      "grad_norm": 1.0225659608840942,
      "learning_rate": 3.726886798760715e-06,
      "loss": 0.0193,
      "num_input_tokens_seen": 20990112,
      "step": 32115
    },
    {
      "epoch": 16.834381551362682,
      "grad_norm": 0.16060949862003326,
      "learning_rate": 3.720882731738215e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 20993184,
      "step": 32120
    },
    {
      "epoch": 16.83700209643606,
      "grad_norm": 1.671285629272461,
      "learning_rate": 3.714883116064291e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 20995936,
      "step": 32125
    },
    {
      "epoch": 16.839622641509433,
      "grad_norm": 1.3956999778747559,
      "learning_rate": 3.708887952993978e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 20999008,
      "step": 32130
    },
    {
      "epoch": 16.84224318658281,
      "grad_norm": 0.34192928671836853,
      "learning_rate": 3.7028972437814057e-06,
      "loss": 0.0205,
      "num_input_tokens_seen": 21002304,
      "step": 32135
    },
    {
      "epoch": 16.844863731656183,
      "grad_norm": 0.11870086938142776,
      "learning_rate": 3.696910989679764e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 21005344,
      "step": 32140
    },
    {
      "epoch": 16.84748427672956,
      "grad_norm": 0.22016537189483643,
      "learning_rate": 3.6909291919413003e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 21009312,
      "step": 32145
    },
    {
      "epoch": 16.850104821802937,
      "grad_norm": 0.08141901344060898,
      "learning_rate": 3.684951851817342e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 21012544,
      "step": 32150
    },
    {
      "epoch": 16.85272536687631,
      "grad_norm": 0.9046382904052734,
      "learning_rate": 3.678978970558286e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 21015200,
      "step": 32155
    },
    {
      "epoch": 16.855345911949687,
      "grad_norm": 0.17505382001399994,
      "learning_rate": 3.673010549413583e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 21017696,
      "step": 32160
    },
    {
      "epoch": 16.85796645702306,
      "grad_norm": 0.6377444863319397,
      "learning_rate": 3.6670465896317592e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21020800,
      "step": 32165
    },
    {
      "epoch": 16.860587002096437,
      "grad_norm": 2.5233099460601807,
      "learning_rate": 3.661087092460419e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 21023552,
      "step": 32170
    },
    {
      "epoch": 16.86320754716981,
      "grad_norm": 0.761066734790802,
      "learning_rate": 3.655132059146202e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 21026656,
      "step": 32175
    },
    {
      "epoch": 16.865828092243188,
      "grad_norm": 4.180483341217041,
      "learning_rate": 3.6491814909348526e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 21030464,
      "step": 32180
    },
    {
      "epoch": 16.86844863731656,
      "grad_norm": 0.8972131609916687,
      "learning_rate": 3.6432353890711424e-06,
      "loss": 0.0112,
      "num_input_tokens_seen": 21033472,
      "step": 32185
    },
    {
      "epoch": 16.871069182389938,
      "grad_norm": 15.640043258666992,
      "learning_rate": 3.6372937547989425e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 21036512,
      "step": 32190
    },
    {
      "epoch": 16.87368972746331,
      "grad_norm": 0.8599612712860107,
      "learning_rate": 3.631356589361173e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 21039936,
      "step": 32195
    },
    {
      "epoch": 16.87631027253669,
      "grad_norm": 9.673925399780273,
      "learning_rate": 3.625423893999813e-06,
      "loss": 0.0301,
      "num_input_tokens_seen": 21043360,
      "step": 32200
    },
    {
      "epoch": 16.878930817610062,
      "grad_norm": 1.2249882221221924,
      "learning_rate": 3.6194956699559217e-06,
      "loss": 0.0139,
      "num_input_tokens_seen": 21046880,
      "step": 32205
    },
    {
      "epoch": 16.88155136268344,
      "grad_norm": 0.33742621541023254,
      "learning_rate": 3.6135719184696188e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 21049664,
      "step": 32210
    },
    {
      "epoch": 16.884171907756812,
      "grad_norm": 0.09436578303575516,
      "learning_rate": 3.6076526407800766e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 21052832,
      "step": 32215
    },
    {
      "epoch": 16.88679245283019,
      "grad_norm": 0.4884236454963684,
      "learning_rate": 3.601737838125552e-06,
      "loss": 0.0094,
      "num_input_tokens_seen": 21057120,
      "step": 32220
    },
    {
      "epoch": 16.889412997903563,
      "grad_norm": 3.083065986633301,
      "learning_rate": 3.595827511743341e-06,
      "loss": 0.0277,
      "num_input_tokens_seen": 21060512,
      "step": 32225
    },
    {
      "epoch": 16.89203354297694,
      "grad_norm": 1.804264783859253,
      "learning_rate": 3.5899216628698217e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 21062720,
      "step": 32230
    },
    {
      "epoch": 16.894654088050313,
      "grad_norm": 0.059678614139556885,
      "learning_rate": 3.584020292740431e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 21065504,
      "step": 32235
    },
    {
      "epoch": 16.89727463312369,
      "grad_norm": 0.031885165721178055,
      "learning_rate": 3.578123402589675e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 21069472,
      "step": 32240
    },
    {
      "epoch": 16.899895178197063,
      "grad_norm": 1.3908092975616455,
      "learning_rate": 3.5722309936511044e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 21072192,
      "step": 32245
    },
    {
      "epoch": 16.90251572327044,
      "grad_norm": 1.1280832290649414,
      "learning_rate": 3.5663430671573572e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21075264,
      "step": 32250
    },
    {
      "epoch": 16.905136268343817,
      "grad_norm": 0.8072314262390137,
      "learning_rate": 3.5604596243401038e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 21079008,
      "step": 32255
    },
    {
      "epoch": 16.90775681341719,
      "grad_norm": 0.2567293643951416,
      "learning_rate": 3.554580666430102e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 21081888,
      "step": 32260
    },
    {
      "epoch": 16.910377358490567,
      "grad_norm": 0.14579862356185913,
      "learning_rate": 3.5487061946571614e-06,
      "loss": 0.0133,
      "num_input_tokens_seen": 21084608,
      "step": 32265
    },
    {
      "epoch": 16.91299790356394,
      "grad_norm": 1.3922003507614136,
      "learning_rate": 3.542836210250161e-06,
      "loss": 0.0211,
      "num_input_tokens_seen": 21087680,
      "step": 32270
    },
    {
      "epoch": 16.915618448637318,
      "grad_norm": 8.859701156616211,
      "learning_rate": 3.536970714437032e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 21090944,
      "step": 32275
    },
    {
      "epoch": 16.91823899371069,
      "grad_norm": 0.47699424624443054,
      "learning_rate": 3.5311097084447575e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 21094272,
      "step": 32280
    },
    {
      "epoch": 16.920859538784068,
      "grad_norm": 0.5160732865333557,
      "learning_rate": 3.5252531934994033e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 21097952,
      "step": 32285
    },
    {
      "epoch": 16.92348008385744,
      "grad_norm": 5.3692779541015625,
      "learning_rate": 3.5194011708260866e-06,
      "loss": 0.0146,
      "num_input_tokens_seen": 21101408,
      "step": 32290
    },
    {
      "epoch": 16.92610062893082,
      "grad_norm": 0.37370529770851135,
      "learning_rate": 3.513553641648984e-06,
      "loss": 0.0148,
      "num_input_tokens_seen": 21105888,
      "step": 32295
    },
    {
      "epoch": 16.928721174004192,
      "grad_norm": 12.722748756408691,
      "learning_rate": 3.5077106071913307e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 21108992,
      "step": 32300
    },
    {
      "epoch": 16.93134171907757,
      "grad_norm": 0.3139748275279999,
      "learning_rate": 3.5018720686754133e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 21111648,
      "step": 32305
    },
    {
      "epoch": 16.933962264150942,
      "grad_norm": 0.3184037208557129,
      "learning_rate": 3.4960380273225985e-06,
      "loss": 0.0169,
      "num_input_tokens_seen": 21115872,
      "step": 32310
    },
    {
      "epoch": 16.93658280922432,
      "grad_norm": 0.44631367921829224,
      "learning_rate": 3.490208484353297e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 21118880,
      "step": 32315
    },
    {
      "epoch": 16.939203354297693,
      "grad_norm": 0.6432901620864868,
      "learning_rate": 3.4843834409869824e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 21121792,
      "step": 32320
    },
    {
      "epoch": 16.94182389937107,
      "grad_norm": 0.5341950058937073,
      "learning_rate": 3.4785628984421977e-06,
      "loss": 0.194,
      "num_input_tokens_seen": 21125024,
      "step": 32325
    },
    {
      "epoch": 16.944444444444443,
      "grad_norm": 0.2903744578361511,
      "learning_rate": 3.4727468579365235e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 21128000,
      "step": 32330
    },
    {
      "epoch": 16.94706498951782,
      "grad_norm": 0.7128163576126099,
      "learning_rate": 3.466935320686604e-06,
      "loss": 0.0451,
      "num_input_tokens_seen": 21131808,
      "step": 32335
    },
    {
      "epoch": 16.949685534591197,
      "grad_norm": 0.11905195564031601,
      "learning_rate": 3.4611282879081553e-06,
      "loss": 0.014,
      "num_input_tokens_seen": 21135488,
      "step": 32340
    },
    {
      "epoch": 16.95230607966457,
      "grad_norm": 0.4819779694080353,
      "learning_rate": 3.455325760815939e-06,
      "loss": 0.0166,
      "num_input_tokens_seen": 21138656,
      "step": 32345
    },
    {
      "epoch": 16.954926624737947,
      "grad_norm": 2.7985987663269043,
      "learning_rate": 3.4495277406237843e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 21141184,
      "step": 32350
    },
    {
      "epoch": 16.95754716981132,
      "grad_norm": 0.3034976124763489,
      "learning_rate": 3.443734228544565e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 21144288,
      "step": 32355
    },
    {
      "epoch": 16.960167714884697,
      "grad_norm": 1.7092373371124268,
      "learning_rate": 3.437945225790215e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 21146912,
      "step": 32360
    },
    {
      "epoch": 16.96278825995807,
      "grad_norm": 2.2904953956604004,
      "learning_rate": 3.432160733571729e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 21150304,
      "step": 32365
    },
    {
      "epoch": 16.965408805031448,
      "grad_norm": 0.7232529520988464,
      "learning_rate": 3.4263807530991595e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 21153760,
      "step": 32370
    },
    {
      "epoch": 16.96802935010482,
      "grad_norm": 0.28616467118263245,
      "learning_rate": 3.4206052855816155e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 21156960,
      "step": 32375
    },
    {
      "epoch": 16.970649895178198,
      "grad_norm": 1.8195914030075073,
      "learning_rate": 3.4148343322272567e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 21159744,
      "step": 32380
    },
    {
      "epoch": 16.97327044025157,
      "grad_norm": 0.506141722202301,
      "learning_rate": 3.4090678942432923e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 21166304,
      "step": 32385
    },
    {
      "epoch": 16.97589098532495,
      "grad_norm": 0.23830311000347137,
      "learning_rate": 3.4033059728360018e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 21169440,
      "step": 32390
    },
    {
      "epoch": 16.978511530398322,
      "grad_norm": 0.740675687789917,
      "learning_rate": 3.397548569210715e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 21173696,
      "step": 32395
    },
    {
      "epoch": 16.9811320754717,
      "grad_norm": 1.1174732446670532,
      "learning_rate": 3.3917956845718136e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 21176672,
      "step": 32400
    },
    {
      "epoch": 16.983752620545072,
      "grad_norm": 0.20409978926181793,
      "learning_rate": 3.38604732012274e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21180064,
      "step": 32405
    },
    {
      "epoch": 16.98637316561845,
      "grad_norm": 7.3625264167785645,
      "learning_rate": 3.3803034770659825e-06,
      "loss": 0.0692,
      "num_input_tokens_seen": 21183680,
      "step": 32410
    },
    {
      "epoch": 16.988993710691823,
      "grad_norm": 0.13708725571632385,
      "learning_rate": 3.3745641566030826e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 21187072,
      "step": 32415
    },
    {
      "epoch": 16.9916142557652,
      "grad_norm": 2.708285093307495,
      "learning_rate": 3.3688293599346442e-06,
      "loss": 0.0336,
      "num_input_tokens_seen": 21191776,
      "step": 32420
    },
    {
      "epoch": 16.994234800838573,
      "grad_norm": 2.1982834339141846,
      "learning_rate": 3.3630990882603213e-06,
      "loss": 0.0964,
      "num_input_tokens_seen": 21196288,
      "step": 32425
    },
    {
      "epoch": 16.99685534591195,
      "grad_norm": 0.9836459159851074,
      "learning_rate": 3.35737334277883e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 21200992,
      "step": 32430
    },
    {
      "epoch": 16.999475890985323,
      "grad_norm": 1.6417367458343506,
      "learning_rate": 3.351652124687918e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 21204320,
      "step": 32435
    },
    {
      "epoch": 17.0,
      "eval_loss": 1.514715552330017,
      "eval_runtime": 11.8628,
      "eval_samples_per_second": 71.484,
      "eval_steps_per_second": 17.871,
      "num_input_tokens_seen": 21204416,
      "step": 32436
    },
    {
      "epoch": 17.0020964360587,
      "grad_norm": 0.06999629735946655,
      "learning_rate": 3.345935435184408e-06,
      "loss": 0.0307,
      "num_input_tokens_seen": 21206752,
      "step": 32440
    },
    {
      "epoch": 17.004716981132077,
      "grad_norm": 0.20864485204219818,
      "learning_rate": 3.3402232754641576e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 21210656,
      "step": 32445
    },
    {
      "epoch": 17.00733752620545,
      "grad_norm": 2.7341911792755127,
      "learning_rate": 3.334515646722089e-06,
      "loss": 0.0195,
      "num_input_tokens_seen": 21213216,
      "step": 32450
    },
    {
      "epoch": 17.009958071278827,
      "grad_norm": 0.253219872713089,
      "learning_rate": 3.32881255015218e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 21216416,
      "step": 32455
    },
    {
      "epoch": 17.0125786163522,
      "grad_norm": 0.36813464760780334,
      "learning_rate": 3.3231139869474433e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 21220352,
      "step": 32460
    },
    {
      "epoch": 17.015199161425578,
      "grad_norm": 0.23373527824878693,
      "learning_rate": 3.317419958299961e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 21223296,
      "step": 32465
    },
    {
      "epoch": 17.01781970649895,
      "grad_norm": 0.15973161160945892,
      "learning_rate": 3.31173046540085e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 21225824,
      "step": 32470
    },
    {
      "epoch": 17.020440251572328,
      "grad_norm": 0.44862332940101624,
      "learning_rate": 3.306045509440295e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 21228704,
      "step": 32475
    },
    {
      "epoch": 17.0230607966457,
      "grad_norm": 0.18711671233177185,
      "learning_rate": 3.300365091607521e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 21231648,
      "step": 32480
    },
    {
      "epoch": 17.02568134171908,
      "grad_norm": 2.4603564739227295,
      "learning_rate": 3.2946892130908127e-06,
      "loss": 0.0165,
      "num_input_tokens_seen": 21234048,
      "step": 32485
    },
    {
      "epoch": 17.028301886792452,
      "grad_norm": 1.3365691900253296,
      "learning_rate": 3.289017875077488e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 21237568,
      "step": 32490
    },
    {
      "epoch": 17.03092243186583,
      "grad_norm": 0.32222312688827515,
      "learning_rate": 3.283351078753938e-06,
      "loss": 0.0183,
      "num_input_tokens_seen": 21240672,
      "step": 32495
    },
    {
      "epoch": 17.033542976939202,
      "grad_norm": 1.2915138006210327,
      "learning_rate": 3.2776888253055777e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 21243200,
      "step": 32500
    },
    {
      "epoch": 17.03616352201258,
      "grad_norm": 0.5389934182167053,
      "learning_rate": 3.272031115916896e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 21247936,
      "step": 32505
    },
    {
      "epoch": 17.038784067085953,
      "grad_norm": 0.2898378372192383,
      "learning_rate": 3.2663779517714238e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 21250560,
      "step": 32510
    },
    {
      "epoch": 17.04140461215933,
      "grad_norm": 0.2647784352302551,
      "learning_rate": 3.260729334051729e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 21254432,
      "step": 32515
    },
    {
      "epoch": 17.044025157232703,
      "grad_norm": 0.6323239803314209,
      "learning_rate": 3.2550852639394446e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 21257568,
      "step": 32520
    },
    {
      "epoch": 17.04664570230608,
      "grad_norm": 1.5417723655700684,
      "learning_rate": 3.2494457426152452e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 21260992,
      "step": 32525
    },
    {
      "epoch": 17.049266247379453,
      "grad_norm": 2.877469301223755,
      "learning_rate": 3.243810771258851e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 21263328,
      "step": 32530
    },
    {
      "epoch": 17.05188679245283,
      "grad_norm": 1.9878339767456055,
      "learning_rate": 3.2381803510490387e-06,
      "loss": 0.0087,
      "num_input_tokens_seen": 21266752,
      "step": 32535
    },
    {
      "epoch": 17.054507337526207,
      "grad_norm": 0.3921281397342682,
      "learning_rate": 3.2325544831636217e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 21269728,
      "step": 32540
    },
    {
      "epoch": 17.05712788259958,
      "grad_norm": 0.47041261196136475,
      "learning_rate": 3.22693316877947e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 21272864,
      "step": 32545
    },
    {
      "epoch": 17.059748427672957,
      "grad_norm": 1.044430136680603,
      "learning_rate": 3.2213164090725057e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 21276320,
      "step": 32550
    },
    {
      "epoch": 17.06236897274633,
      "grad_norm": 0.23086708784103394,
      "learning_rate": 3.2157042052176763e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 21279360,
      "step": 32555
    },
    {
      "epoch": 17.064989517819708,
      "grad_norm": 0.10141181945800781,
      "learning_rate": 3.2100965583890003e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 21282112,
      "step": 32560
    },
    {
      "epoch": 17.06761006289308,
      "grad_norm": 1.1966371536254883,
      "learning_rate": 3.204493469759537e-06,
      "loss": 0.0099,
      "num_input_tokens_seen": 21284832,
      "step": 32565
    },
    {
      "epoch": 17.070230607966458,
      "grad_norm": 0.1687471568584442,
      "learning_rate": 3.1988949405013764e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21288256,
      "step": 32570
    },
    {
      "epoch": 17.07285115303983,
      "grad_norm": 0.3307425081729889,
      "learning_rate": 3.193300971785676e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 21291616,
      "step": 32575
    },
    {
      "epoch": 17.07547169811321,
      "grad_norm": 0.3964137136936188,
      "learning_rate": 3.1877115647826335e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 21294720,
      "step": 32580
    },
    {
      "epoch": 17.078092243186582,
      "grad_norm": 1.2241685390472412,
      "learning_rate": 3.1821267206614797e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 21298400,
      "step": 32585
    },
    {
      "epoch": 17.08071278825996,
      "grad_norm": 0.13479715585708618,
      "learning_rate": 3.176546440590508e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 21301376,
      "step": 32590
    },
    {
      "epoch": 17.083333333333332,
      "grad_norm": 0.26145222783088684,
      "learning_rate": 3.170970725737041e-06,
      "loss": 0.0014,
      "num_input_tokens_seen": 21303840,
      "step": 32595
    },
    {
      "epoch": 17.08595387840671,
      "grad_norm": 0.5210472941398621,
      "learning_rate": 3.165399577267461e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 21307136,
      "step": 32600
    },
    {
      "epoch": 17.088574423480082,
      "grad_norm": 0.4151609241962433,
      "learning_rate": 3.159832996347184e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 21309824,
      "step": 32605
    },
    {
      "epoch": 17.09119496855346,
      "grad_norm": 0.34732574224472046,
      "learning_rate": 3.1542709841406865e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21312448,
      "step": 32610
    },
    {
      "epoch": 17.093815513626833,
      "grad_norm": 0.14373676478862762,
      "learning_rate": 3.148713541811468e-06,
      "loss": 0.0037,
      "num_input_tokens_seen": 21315040,
      "step": 32615
    },
    {
      "epoch": 17.09643605870021,
      "grad_norm": 0.2629750072956085,
      "learning_rate": 3.1431606705220803e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21317696,
      "step": 32620
    },
    {
      "epoch": 17.099056603773583,
      "grad_norm": 0.15562395751476288,
      "learning_rate": 3.1376123714341234e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 21320320,
      "step": 32625
    },
    {
      "epoch": 17.10167714884696,
      "grad_norm": 0.1018417477607727,
      "learning_rate": 3.1320686457082387e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 21323136,
      "step": 32630
    },
    {
      "epoch": 17.104297693920337,
      "grad_norm": 0.5384577512741089,
      "learning_rate": 3.126529494504113e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 21325664,
      "step": 32635
    },
    {
      "epoch": 17.10691823899371,
      "grad_norm": 0.6350106596946716,
      "learning_rate": 3.1209949189804744e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 21328352,
      "step": 32640
    },
    {
      "epoch": 17.109538784067087,
      "grad_norm": 0.26164156198501587,
      "learning_rate": 3.115464920295089e-06,
      "loss": 0.0138,
      "num_input_tokens_seen": 21332704,
      "step": 32645
    },
    {
      "epoch": 17.11215932914046,
      "grad_norm": 0.7074905633926392,
      "learning_rate": 3.1099394996047666e-06,
      "loss": 0.0204,
      "num_input_tokens_seen": 21336352,
      "step": 32650
    },
    {
      "epoch": 17.114779874213838,
      "grad_norm": 0.09467960894107819,
      "learning_rate": 3.104418658065367e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 21340288,
      "step": 32655
    },
    {
      "epoch": 17.11740041928721,
      "grad_norm": 0.6256729960441589,
      "learning_rate": 3.098902396831782e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 21344128,
      "step": 32660
    },
    {
      "epoch": 17.120020964360588,
      "grad_norm": 0.5680224299430847,
      "learning_rate": 3.0933907170579646e-06,
      "loss": 0.0186,
      "num_input_tokens_seen": 21347296,
      "step": 32665
    },
    {
      "epoch": 17.12264150943396,
      "grad_norm": 0.5564082264900208,
      "learning_rate": 3.0878836198968835e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 21350240,
      "step": 32670
    },
    {
      "epoch": 17.12526205450734,
      "grad_norm": 0.1377526968717575,
      "learning_rate": 3.0823811065005564e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 21352704,
      "step": 32675
    },
    {
      "epoch": 17.127882599580712,
      "grad_norm": 1.702911615371704,
      "learning_rate": 3.0768831780200542e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 21355904,
      "step": 32680
    },
    {
      "epoch": 17.13050314465409,
      "grad_norm": 0.7292142510414124,
      "learning_rate": 3.071389835605476e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 21359136,
      "step": 32685
    },
    {
      "epoch": 17.133123689727462,
      "grad_norm": 0.3284044861793518,
      "learning_rate": 3.065901080405978e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 21362496,
      "step": 32690
    },
    {
      "epoch": 17.13574423480084,
      "grad_norm": 0.10700962692499161,
      "learning_rate": 3.060416913569733e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 21364928,
      "step": 32695
    },
    {
      "epoch": 17.138364779874212,
      "grad_norm": 0.5438835620880127,
      "learning_rate": 3.054937336243968e-06,
      "loss": 0.0095,
      "num_input_tokens_seen": 21367808,
      "step": 32700
    },
    {
      "epoch": 17.14098532494759,
      "grad_norm": 1.4874030351638794,
      "learning_rate": 3.049462349574947e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 21371040,
      "step": 32705
    },
    {
      "epoch": 17.143605870020963,
      "grad_norm": 0.11814378947019577,
      "learning_rate": 3.043991954707981e-06,
      "loss": 0.0116,
      "num_input_tokens_seen": 21374528,
      "step": 32710
    },
    {
      "epoch": 17.14622641509434,
      "grad_norm": 0.3810107707977295,
      "learning_rate": 3.0385261527874075e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 21376864,
      "step": 32715
    },
    {
      "epoch": 17.148846960167713,
      "grad_norm": 0.3930932879447937,
      "learning_rate": 3.0330649449566205e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21379840,
      "step": 32720
    },
    {
      "epoch": 17.15146750524109,
      "grad_norm": 0.9041348099708557,
      "learning_rate": 3.027608332358034e-06,
      "loss": 0.023,
      "num_input_tokens_seen": 21383072,
      "step": 32725
    },
    {
      "epoch": 17.154088050314467,
      "grad_norm": 0.30385622382164,
      "learning_rate": 3.022156316133104e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 21386080,
      "step": 32730
    },
    {
      "epoch": 17.15670859538784,
      "grad_norm": 1.7432385683059692,
      "learning_rate": 3.016708897422335e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 21388640,
      "step": 32735
    },
    {
      "epoch": 17.159329140461217,
      "grad_norm": 1.12300443649292,
      "learning_rate": 3.0112660773652685e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 21391552,
      "step": 32740
    },
    {
      "epoch": 17.16194968553459,
      "grad_norm": 0.511772871017456,
      "learning_rate": 3.005827857100482e-06,
      "loss": 0.0091,
      "num_input_tokens_seen": 21395040,
      "step": 32745
    },
    {
      "epoch": 17.164570230607968,
      "grad_norm": 0.8964090943336487,
      "learning_rate": 3.000394237765586e-06,
      "loss": 0.0037,
      "num_input_tokens_seen": 21398688,
      "step": 32750
    },
    {
      "epoch": 17.16719077568134,
      "grad_norm": 1.3368043899536133,
      "learning_rate": 2.9949652204972254e-06,
      "loss": 0.0085,
      "num_input_tokens_seen": 21402144,
      "step": 32755
    },
    {
      "epoch": 17.169811320754718,
      "grad_norm": 1.097851276397705,
      "learning_rate": 2.989540806431093e-06,
      "loss": 0.024,
      "num_input_tokens_seen": 21404352,
      "step": 32760
    },
    {
      "epoch": 17.17243186582809,
      "grad_norm": 0.343098521232605,
      "learning_rate": 2.984120996701917e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 21407808,
      "step": 32765
    },
    {
      "epoch": 17.17505241090147,
      "grad_norm": 0.6512918472290039,
      "learning_rate": 2.9787057924434618e-06,
      "loss": 0.0598,
      "num_input_tokens_seen": 21410976,
      "step": 32770
    },
    {
      "epoch": 17.177672955974842,
      "grad_norm": 0.22028425335884094,
      "learning_rate": 2.9732951947885166e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 21414976,
      "step": 32775
    },
    {
      "epoch": 17.18029350104822,
      "grad_norm": 0.5537497997283936,
      "learning_rate": 2.967889204868929e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 21418368,
      "step": 32780
    },
    {
      "epoch": 17.182914046121592,
      "grad_norm": 0.306264191865921,
      "learning_rate": 2.96248782381556e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 21421664,
      "step": 32785
    },
    {
      "epoch": 17.18553459119497,
      "grad_norm": 0.7150886654853821,
      "learning_rate": 2.957091052758321e-06,
      "loss": 0.0308,
      "num_input_tokens_seen": 21428544,
      "step": 32790
    },
    {
      "epoch": 17.188155136268342,
      "grad_norm": 0.3915163278579712,
      "learning_rate": 2.9516988928261517e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 21431520,
      "step": 32795
    },
    {
      "epoch": 17.19077568134172,
      "grad_norm": 0.6644130349159241,
      "learning_rate": 2.9463113451470403e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 21435104,
      "step": 32800
    },
    {
      "epoch": 17.193396226415093,
      "grad_norm": 1.1894549131393433,
      "learning_rate": 2.94092841084799e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 21437888,
      "step": 32805
    },
    {
      "epoch": 17.19601677148847,
      "grad_norm": 0.6477234959602356,
      "learning_rate": 2.9355500910550586e-06,
      "loss": 0.0135,
      "num_input_tokens_seen": 21441280,
      "step": 32810
    },
    {
      "epoch": 17.198637316561843,
      "grad_norm": 0.1716027557849884,
      "learning_rate": 2.9301763868933157e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 21445472,
      "step": 32815
    },
    {
      "epoch": 17.20125786163522,
      "grad_norm": 0.26858264207839966,
      "learning_rate": 2.9248072994868896e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 21447872,
      "step": 32820
    },
    {
      "epoch": 17.203878406708597,
      "grad_norm": 0.45322269201278687,
      "learning_rate": 2.919442829958932e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 21451456,
      "step": 32825
    },
    {
      "epoch": 17.20649895178197,
      "grad_norm": 0.5721315145492554,
      "learning_rate": 2.9140829794316234e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 21454560,
      "step": 32830
    },
    {
      "epoch": 17.209119496855347,
      "grad_norm": 0.24997498095035553,
      "learning_rate": 2.908727749026191e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 21457248,
      "step": 32835
    },
    {
      "epoch": 17.21174004192872,
      "grad_norm": 0.07748988270759583,
      "learning_rate": 2.9033771398628756e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 21461056,
      "step": 32840
    },
    {
      "epoch": 17.214360587002098,
      "grad_norm": 1.0092511177062988,
      "learning_rate": 2.898031153060973e-06,
      "loss": 0.031,
      "num_input_tokens_seen": 21466176,
      "step": 32845
    },
    {
      "epoch": 17.21698113207547,
      "grad_norm": 1.1127777099609375,
      "learning_rate": 2.8926897897388033e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21469824,
      "step": 32850
    },
    {
      "epoch": 17.219601677148848,
      "grad_norm": 0.06870836764574051,
      "learning_rate": 2.8873530510137136e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 21473472,
      "step": 32855
    },
    {
      "epoch": 17.22222222222222,
      "grad_norm": 0.44919443130493164,
      "learning_rate": 2.8820209380020892e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 21476640,
      "step": 32860
    },
    {
      "epoch": 17.2248427672956,
      "grad_norm": 0.20197045803070068,
      "learning_rate": 2.8766934518193566e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 21479712,
      "step": 32865
    },
    {
      "epoch": 17.22746331236897,
      "grad_norm": 0.6292629241943359,
      "learning_rate": 2.87137059357995e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 21483104,
      "step": 32870
    },
    {
      "epoch": 17.23008385744235,
      "grad_norm": 0.2603228986263275,
      "learning_rate": 2.8660523643973618e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 21486080,
      "step": 32875
    },
    {
      "epoch": 17.232704402515722,
      "grad_norm": 0.03999394550919533,
      "learning_rate": 2.860738765384105e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 21489792,
      "step": 32880
    },
    {
      "epoch": 17.2353249475891,
      "grad_norm": 0.39329439401626587,
      "learning_rate": 2.8554297976517154e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 21492704,
      "step": 32885
    },
    {
      "epoch": 17.237945492662472,
      "grad_norm": 0.3251906633377075,
      "learning_rate": 2.8501254623107707e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 21496992,
      "step": 32890
    },
    {
      "epoch": 17.24056603773585,
      "grad_norm": 3.440009832382202,
      "learning_rate": 2.84482576047089e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 21500320,
      "step": 32895
    },
    {
      "epoch": 17.243186582809223,
      "grad_norm": 2.2966859340667725,
      "learning_rate": 2.839530693240694e-06,
      "loss": 0.0134,
      "num_input_tokens_seen": 21503584,
      "step": 32900
    },
    {
      "epoch": 17.2458071278826,
      "grad_norm": 0.5203356146812439,
      "learning_rate": 2.8342402617278606e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 21506240,
      "step": 32905
    },
    {
      "epoch": 17.248427672955973,
      "grad_norm": 0.18865929543972015,
      "learning_rate": 2.82895446703908e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 21510208,
      "step": 32910
    },
    {
      "epoch": 17.25104821802935,
      "grad_norm": 0.19496096670627594,
      "learning_rate": 2.823673310280084e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 21513344,
      "step": 32915
    },
    {
      "epoch": 17.253668763102727,
      "grad_norm": 0.27958813309669495,
      "learning_rate": 2.8183967925556305e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21515968,
      "step": 32920
    },
    {
      "epoch": 17.2562893081761,
      "grad_norm": 0.6019533276557922,
      "learning_rate": 2.8131249149695115e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 21519200,
      "step": 32925
    },
    {
      "epoch": 17.258909853249477,
      "grad_norm": 0.13584983348846436,
      "learning_rate": 2.8078576786245397e-06,
      "loss": 0.0179,
      "num_input_tokens_seen": 21522368,
      "step": 32930
    },
    {
      "epoch": 17.26153039832285,
      "grad_norm": 0.2532923221588135,
      "learning_rate": 2.8025950846225557e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 21525280,
      "step": 32935
    },
    {
      "epoch": 17.264150943396228,
      "grad_norm": 0.18212297558784485,
      "learning_rate": 2.7973371340644373e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 21528736,
      "step": 32940
    },
    {
      "epoch": 17.2667714884696,
      "grad_norm": 0.44070863723754883,
      "learning_rate": 2.7920838280500907e-06,
      "loss": 0.0467,
      "num_input_tokens_seen": 21531904,
      "step": 32945
    },
    {
      "epoch": 17.269392033542978,
      "grad_norm": 1.6947550773620605,
      "learning_rate": 2.7868351676784504e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 21534592,
      "step": 32950
    },
    {
      "epoch": 17.27201257861635,
      "grad_norm": 1.4508458375930786,
      "learning_rate": 2.7815911540474665e-06,
      "loss": 0.0168,
      "num_input_tokens_seen": 21538880,
      "step": 32955
    },
    {
      "epoch": 17.27463312368973,
      "grad_norm": 0.5190322399139404,
      "learning_rate": 2.7763517882541364e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 21542688,
      "step": 32960
    },
    {
      "epoch": 17.2772536687631,
      "grad_norm": 0.4787108302116394,
      "learning_rate": 2.771117071394466e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 21545760,
      "step": 32965
    },
    {
      "epoch": 17.27987421383648,
      "grad_norm": 1.2988903522491455,
      "learning_rate": 2.7658870045635055e-06,
      "loss": 0.0133,
      "num_input_tokens_seen": 21548736,
      "step": 32970
    },
    {
      "epoch": 17.282494758909852,
      "grad_norm": 0.25654077529907227,
      "learning_rate": 2.7606615888553234e-06,
      "loss": 0.0192,
      "num_input_tokens_seen": 21551936,
      "step": 32975
    },
    {
      "epoch": 17.28511530398323,
      "grad_norm": 1.1007981300354004,
      "learning_rate": 2.755440825363023e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 21555296,
      "step": 32980
    },
    {
      "epoch": 17.287735849056602,
      "grad_norm": 0.7350364923477173,
      "learning_rate": 2.75022471517872e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 21558144,
      "step": 32985
    },
    {
      "epoch": 17.29035639412998,
      "grad_norm": 0.25142884254455566,
      "learning_rate": 2.745013259393564e-06,
      "loss": 0.0166,
      "num_input_tokens_seen": 21562048,
      "step": 32990
    },
    {
      "epoch": 17.292976939203353,
      "grad_norm": 1.4842274188995361,
      "learning_rate": 2.739806459097735e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 21565312,
      "step": 32995
    },
    {
      "epoch": 17.29559748427673,
      "grad_norm": 1.4421671628952026,
      "learning_rate": 2.7346043153804403e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 21568960,
      "step": 33000
    },
    {
      "epoch": 17.298218029350103,
      "grad_norm": 0.2942492663860321,
      "learning_rate": 2.729406829329903e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 21572064,
      "step": 33005
    },
    {
      "epoch": 17.30083857442348,
      "grad_norm": 2.0562734603881836,
      "learning_rate": 2.72421400203339e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21575328,
      "step": 33010
    },
    {
      "epoch": 17.303459119496857,
      "grad_norm": 0.9779902100563049,
      "learning_rate": 2.719025834577163e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 21578912,
      "step": 33015
    },
    {
      "epoch": 17.30607966457023,
      "grad_norm": 0.3237854242324829,
      "learning_rate": 2.713842328046537e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 21582560,
      "step": 33020
    },
    {
      "epoch": 17.308700209643607,
      "grad_norm": 0.2490975260734558,
      "learning_rate": 2.7086634835258416e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 21586912,
      "step": 33025
    },
    {
      "epoch": 17.31132075471698,
      "grad_norm": 0.27827104926109314,
      "learning_rate": 2.7034893020984333e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21589920,
      "step": 33030
    },
    {
      "epoch": 17.313941299790358,
      "grad_norm": 0.24374981224536896,
      "learning_rate": 2.6983197848466952e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 21592416,
      "step": 33035
    },
    {
      "epoch": 17.31656184486373,
      "grad_norm": 1.5342626571655273,
      "learning_rate": 2.6931549328520265e-06,
      "loss": 0.0253,
      "num_input_tokens_seen": 21596448,
      "step": 33040
    },
    {
      "epoch": 17.319182389937108,
      "grad_norm": 0.22583477199077606,
      "learning_rate": 2.6879947471948523e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 21599968,
      "step": 33045
    },
    {
      "epoch": 17.32180293501048,
      "grad_norm": 0.23924706876277924,
      "learning_rate": 2.6828392289546285e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 21603168,
      "step": 33050
    },
    {
      "epoch": 17.32442348008386,
      "grad_norm": 0.41076213121414185,
      "learning_rate": 2.677688379209828e-06,
      "loss": 0.0158,
      "num_input_tokens_seen": 21606336,
      "step": 33055
    },
    {
      "epoch": 17.32704402515723,
      "grad_norm": 2.579105854034424,
      "learning_rate": 2.6725421990379586e-06,
      "loss": 0.007,
      "num_input_tokens_seen": 21609728,
      "step": 33060
    },
    {
      "epoch": 17.32966457023061,
      "grad_norm": 0.19788357615470886,
      "learning_rate": 2.667400689515537e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 21612448,
      "step": 33065
    },
    {
      "epoch": 17.332285115303982,
      "grad_norm": 0.6156572103500366,
      "learning_rate": 2.6622638517181e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 21615424,
      "step": 33070
    },
    {
      "epoch": 17.33490566037736,
      "grad_norm": 19.123088836669922,
      "learning_rate": 2.657131686720224e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 21618400,
      "step": 33075
    },
    {
      "epoch": 17.337526205450732,
      "grad_norm": 1.2829508781433105,
      "learning_rate": 2.652004195595498e-06,
      "loss": 0.0096,
      "num_input_tokens_seen": 21622688,
      "step": 33080
    },
    {
      "epoch": 17.34014675052411,
      "grad_norm": 0.4228057265281677,
      "learning_rate": 2.6468813794165356e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21626176,
      "step": 33085
    },
    {
      "epoch": 17.342767295597483,
      "grad_norm": 0.7790247797966003,
      "learning_rate": 2.641763239254977e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 21629440,
      "step": 33090
    },
    {
      "epoch": 17.34538784067086,
      "grad_norm": 0.28918692469596863,
      "learning_rate": 2.6366497761814713e-06,
      "loss": 0.0417,
      "num_input_tokens_seen": 21632000,
      "step": 33095
    },
    {
      "epoch": 17.348008385744233,
      "grad_norm": 0.4651499092578888,
      "learning_rate": 2.631540991265691e-06,
      "loss": 0.0191,
      "num_input_tokens_seen": 21635104,
      "step": 33100
    },
    {
      "epoch": 17.35062893081761,
      "grad_norm": 0.6357519030570984,
      "learning_rate": 2.6264368855763443e-06,
      "loss": 0.0198,
      "num_input_tokens_seen": 21638624,
      "step": 33105
    },
    {
      "epoch": 17.353249475890987,
      "grad_norm": 0.7236025929450989,
      "learning_rate": 2.6213374601811504e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 21641664,
      "step": 33110
    },
    {
      "epoch": 17.35587002096436,
      "grad_norm": 1.2504018545150757,
      "learning_rate": 2.616242716146855e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 21644672,
      "step": 33115
    },
    {
      "epoch": 17.358490566037737,
      "grad_norm": 0.7484932541847229,
      "learning_rate": 2.6111526545392145e-06,
      "loss": 0.0075,
      "num_input_tokens_seen": 21647328,
      "step": 33120
    },
    {
      "epoch": 17.36111111111111,
      "grad_norm": 0.4739953577518463,
      "learning_rate": 2.6060672764230106e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 21650368,
      "step": 33125
    },
    {
      "epoch": 17.363731656184488,
      "grad_norm": 0.19820915162563324,
      "learning_rate": 2.600986582862047e-06,
      "loss": 0.0238,
      "num_input_tokens_seen": 21654400,
      "step": 33130
    },
    {
      "epoch": 17.36635220125786,
      "grad_norm": 0.6684425473213196,
      "learning_rate": 2.5959105749191498e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 21657856,
      "step": 33135
    },
    {
      "epoch": 17.368972746331238,
      "grad_norm": 0.1721796989440918,
      "learning_rate": 2.590839253656166e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 21661088,
      "step": 33140
    },
    {
      "epoch": 17.37159329140461,
      "grad_norm": 0.3299154043197632,
      "learning_rate": 2.5857726201339473e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 21666144,
      "step": 33145
    },
    {
      "epoch": 17.37421383647799,
      "grad_norm": 0.5897979736328125,
      "learning_rate": 2.5807106754123873e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 21669152,
      "step": 33150
    },
    {
      "epoch": 17.37683438155136,
      "grad_norm": 0.9842673540115356,
      "learning_rate": 2.575653420550378e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 21671840,
      "step": 33155
    },
    {
      "epoch": 17.37945492662474,
      "grad_norm": 0.17997734248638153,
      "learning_rate": 2.57060085660584e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 21674400,
      "step": 33160
    },
    {
      "epoch": 17.382075471698112,
      "grad_norm": 0.9844767451286316,
      "learning_rate": 2.565552984635716e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 21677280,
      "step": 33165
    },
    {
      "epoch": 17.38469601677149,
      "grad_norm": 1.1182665824890137,
      "learning_rate": 2.5605098056959652e-06,
      "loss": 0.0093,
      "num_input_tokens_seen": 21680352,
      "step": 33170
    },
    {
      "epoch": 17.387316561844862,
      "grad_norm": 0.16040624678134918,
      "learning_rate": 2.555471320841557e-06,
      "loss": 0.0309,
      "num_input_tokens_seen": 21684768,
      "step": 33175
    },
    {
      "epoch": 17.38993710691824,
      "grad_norm": 0.4432581961154938,
      "learning_rate": 2.55043753112649e-06,
      "loss": 0.0293,
      "num_input_tokens_seen": 21687776,
      "step": 33180
    },
    {
      "epoch": 17.392557651991613,
      "grad_norm": 0.13984932005405426,
      "learning_rate": 2.5454084376037717e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 21691392,
      "step": 33185
    },
    {
      "epoch": 17.39517819706499,
      "grad_norm": 0.3172493875026703,
      "learning_rate": 2.5403840413254304e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 21694016,
      "step": 33190
    },
    {
      "epoch": 17.397798742138363,
      "grad_norm": 0.9076021313667297,
      "learning_rate": 2.5353643433425196e-06,
      "loss": 0.0083,
      "num_input_tokens_seen": 21697152,
      "step": 33195
    },
    {
      "epoch": 17.40041928721174,
      "grad_norm": 1.1331799030303955,
      "learning_rate": 2.5303493447050965e-06,
      "loss": 0.0153,
      "num_input_tokens_seen": 21701504,
      "step": 33200
    },
    {
      "epoch": 17.403039832285117,
      "grad_norm": 0.44509217143058777,
      "learning_rate": 2.525339046462244e-06,
      "loss": 0.0193,
      "num_input_tokens_seen": 21704640,
      "step": 33205
    },
    {
      "epoch": 17.40566037735849,
      "grad_norm": 1.218760371208191,
      "learning_rate": 2.5203334496620544e-06,
      "loss": 0.0127,
      "num_input_tokens_seen": 21708160,
      "step": 33210
    },
    {
      "epoch": 17.408280922431867,
      "grad_norm": 0.18184088170528412,
      "learning_rate": 2.515332555351646e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 21710976,
      "step": 33215
    },
    {
      "epoch": 17.41090146750524,
      "grad_norm": 0.3744472861289978,
      "learning_rate": 2.510336364577154e-06,
      "loss": 0.051,
      "num_input_tokens_seen": 21714368,
      "step": 33220
    },
    {
      "epoch": 17.413522012578618,
      "grad_norm": 1.3828686475753784,
      "learning_rate": 2.5053448783837143e-06,
      "loss": 0.0347,
      "num_input_tokens_seen": 21716896,
      "step": 33225
    },
    {
      "epoch": 17.41614255765199,
      "grad_norm": 0.3262661099433899,
      "learning_rate": 2.5003580978154917e-06,
      "loss": 0.0148,
      "num_input_tokens_seen": 21719488,
      "step": 33230
    },
    {
      "epoch": 17.418763102725368,
      "grad_norm": 0.5695001482963562,
      "learning_rate": 2.4953760239156716e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 21724000,
      "step": 33235
    },
    {
      "epoch": 17.42138364779874,
      "grad_norm": 0.2718450129032135,
      "learning_rate": 2.4903986577264366e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 21727424,
      "step": 33240
    },
    {
      "epoch": 17.42400419287212,
      "grad_norm": 0.0435965433716774,
      "learning_rate": 2.4854260002889986e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 21732032,
      "step": 33245
    },
    {
      "epoch": 17.42662473794549,
      "grad_norm": 1.796641230583191,
      "learning_rate": 2.480458052643586e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 21735488,
      "step": 33250
    },
    {
      "epoch": 17.42924528301887,
      "grad_norm": 0.19029663503170013,
      "learning_rate": 2.4754948158294293e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 21737984,
      "step": 33255
    },
    {
      "epoch": 17.431865828092242,
      "grad_norm": 0.21530145406723022,
      "learning_rate": 2.4705362908847814e-06,
      "loss": 0.0154,
      "num_input_tokens_seen": 21741088,
      "step": 33260
    },
    {
      "epoch": 17.43448637316562,
      "grad_norm": 1.1329681873321533,
      "learning_rate": 2.4655824788469172e-06,
      "loss": 0.011,
      "num_input_tokens_seen": 21744928,
      "step": 33265
    },
    {
      "epoch": 17.437106918238992,
      "grad_norm": 1.2483069896697998,
      "learning_rate": 2.4606333807521087e-06,
      "loss": 0.0101,
      "num_input_tokens_seen": 21748288,
      "step": 33270
    },
    {
      "epoch": 17.43972746331237,
      "grad_norm": 0.4264671802520752,
      "learning_rate": 2.455688997635658e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 21751968,
      "step": 33275
    },
    {
      "epoch": 17.442348008385743,
      "grad_norm": 0.835537314414978,
      "learning_rate": 2.4507493305318652e-06,
      "loss": 0.0097,
      "num_input_tokens_seen": 21755648,
      "step": 33280
    },
    {
      "epoch": 17.44496855345912,
      "grad_norm": 0.600723147392273,
      "learning_rate": 2.445814380474057e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 21759040,
      "step": 33285
    },
    {
      "epoch": 17.447589098532493,
      "grad_norm": 0.5933822393417358,
      "learning_rate": 2.4408841484945684e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 21762272,
      "step": 33290
    },
    {
      "epoch": 17.45020964360587,
      "grad_norm": 0.48890286684036255,
      "learning_rate": 2.435958635624755e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 21765600,
      "step": 33295
    },
    {
      "epoch": 17.452830188679247,
      "grad_norm": 0.32249194383621216,
      "learning_rate": 2.4310378428949678e-06,
      "loss": 0.0094,
      "num_input_tokens_seen": 21768032,
      "step": 33300
    },
    {
      "epoch": 17.45545073375262,
      "grad_norm": 1.0417981147766113,
      "learning_rate": 2.4261217713345804e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21770304,
      "step": 33305
    },
    {
      "epoch": 17.458071278825997,
      "grad_norm": 0.7626853585243225,
      "learning_rate": 2.421210421971981e-06,
      "loss": 0.0044,
      "num_input_tokens_seen": 21774016,
      "step": 33310
    },
    {
      "epoch": 17.46069182389937,
      "grad_norm": 0.6404761672019958,
      "learning_rate": 2.416303795834571e-06,
      "loss": 0.0137,
      "num_input_tokens_seen": 21778048,
      "step": 33315
    },
    {
      "epoch": 17.463312368972748,
      "grad_norm": 0.2885684669017792,
      "learning_rate": 2.4114018939487615e-06,
      "loss": 0.0019,
      "num_input_tokens_seen": 21780736,
      "step": 33320
    },
    {
      "epoch": 17.46593291404612,
      "grad_norm": 0.23274724185466766,
      "learning_rate": 2.4065047173399662e-06,
      "loss": 0.0261,
      "num_input_tokens_seen": 21784704,
      "step": 33325
    },
    {
      "epoch": 17.468553459119498,
      "grad_norm": 0.501167893409729,
      "learning_rate": 2.4016122670326295e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 21788096,
      "step": 33330
    },
    {
      "epoch": 17.47117400419287,
      "grad_norm": 0.6795707941055298,
      "learning_rate": 2.396724544050186e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21791808,
      "step": 33335
    },
    {
      "epoch": 17.47379454926625,
      "grad_norm": 0.5535025000572205,
      "learning_rate": 2.3918415494150944e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 21795072,
      "step": 33340
    },
    {
      "epoch": 17.47641509433962,
      "grad_norm": 0.1482686549425125,
      "learning_rate": 2.3869632841488245e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 21798016,
      "step": 33345
    },
    {
      "epoch": 17.479035639413,
      "grad_norm": 0.2677736282348633,
      "learning_rate": 2.382089749271857e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 21801632,
      "step": 33350
    },
    {
      "epoch": 17.481656184486372,
      "grad_norm": 0.11384277790784836,
      "learning_rate": 2.3772209458036737e-06,
      "loss": 0.001,
      "num_input_tokens_seen": 21804416,
      "step": 33355
    },
    {
      "epoch": 17.48427672955975,
      "grad_norm": 0.3248349726200104,
      "learning_rate": 2.372356874762771e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 21808160,
      "step": 33360
    },
    {
      "epoch": 17.486897274633122,
      "grad_norm": 1.2799934148788452,
      "learning_rate": 2.36749753716666e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 21810976,
      "step": 33365
    },
    {
      "epoch": 17.4895178197065,
      "grad_norm": 0.6829182505607605,
      "learning_rate": 2.3626429340318613e-06,
      "loss": 0.006,
      "num_input_tokens_seen": 21814880,
      "step": 33370
    },
    {
      "epoch": 17.492138364779873,
      "grad_norm": 0.1467026323080063,
      "learning_rate": 2.357793066373901e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 21817792,
      "step": 33375
    },
    {
      "epoch": 17.49475890985325,
      "grad_norm": 0.6991633176803589,
      "learning_rate": 2.352947935207317e-06,
      "loss": 0.0044,
      "num_input_tokens_seen": 21821312,
      "step": 33380
    },
    {
      "epoch": 17.497379454926623,
      "grad_norm": 0.33466118574142456,
      "learning_rate": 2.3481075415456495e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21825664,
      "step": 33385
    },
    {
      "epoch": 17.5,
      "grad_norm": 2.514009714126587,
      "learning_rate": 2.3432718864014576e-06,
      "loss": 0.0099,
      "num_input_tokens_seen": 21829088,
      "step": 33390
    },
    {
      "epoch": 17.502620545073377,
      "grad_norm": 0.29960373044013977,
      "learning_rate": 2.338440970786304e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 21832000,
      "step": 33395
    },
    {
      "epoch": 17.50524109014675,
      "grad_norm": 0.5589448809623718,
      "learning_rate": 2.3336147957107644e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 21834816,
      "step": 33400
    },
    {
      "epoch": 17.507861635220127,
      "grad_norm": 0.15210029482841492,
      "learning_rate": 2.3287933621844227e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 21837184,
      "step": 33405
    },
    {
      "epoch": 17.5104821802935,
      "grad_norm": 0.16332164406776428,
      "learning_rate": 2.3239766712158633e-06,
      "loss": 0.0263,
      "num_input_tokens_seen": 21842080,
      "step": 33410
    },
    {
      "epoch": 17.513102725366878,
      "grad_norm": 1.0049479007720947,
      "learning_rate": 2.319164723812675e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 21844736,
      "step": 33415
    },
    {
      "epoch": 17.51572327044025,
      "grad_norm": 1.0578961372375488,
      "learning_rate": 2.314357520981472e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 21848736,
      "step": 33420
    },
    {
      "epoch": 17.518343815513628,
      "grad_norm": 1.2338340282440186,
      "learning_rate": 2.309555063727864e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 21851808,
      "step": 33425
    },
    {
      "epoch": 17.520964360587,
      "grad_norm": 0.06416268646717072,
      "learning_rate": 2.304757353056475e-06,
      "loss": 0.0013,
      "num_input_tokens_seen": 21854752,
      "step": 33430
    },
    {
      "epoch": 17.52358490566038,
      "grad_norm": 0.21961365640163422,
      "learning_rate": 2.2999643899709267e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 21858016,
      "step": 33435
    },
    {
      "epoch": 17.52620545073375,
      "grad_norm": 1.9934340715408325,
      "learning_rate": 2.2951761754738456e-06,
      "loss": 0.0044,
      "num_input_tokens_seen": 21860928,
      "step": 33440
    },
    {
      "epoch": 17.52882599580713,
      "grad_norm": 0.22452308237552643,
      "learning_rate": 2.29039271056688e-06,
      "loss": 0.0017,
      "num_input_tokens_seen": 21864416,
      "step": 33445
    },
    {
      "epoch": 17.531446540880502,
      "grad_norm": 0.6045808792114258,
      "learning_rate": 2.285613996250674e-06,
      "loss": 0.0063,
      "num_input_tokens_seen": 21867808,
      "step": 33450
    },
    {
      "epoch": 17.53406708595388,
      "grad_norm": 0.9938685297966003,
      "learning_rate": 2.2808400335248865e-06,
      "loss": 0.0107,
      "num_input_tokens_seen": 21871008,
      "step": 33455
    },
    {
      "epoch": 17.536687631027252,
      "grad_norm": 0.14440277218818665,
      "learning_rate": 2.2760708233881654e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 21873984,
      "step": 33460
    },
    {
      "epoch": 17.53930817610063,
      "grad_norm": 0.40084847807884216,
      "learning_rate": 2.271306366838183e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 21877216,
      "step": 33465
    },
    {
      "epoch": 17.541928721174003,
      "grad_norm": 0.08970408141613007,
      "learning_rate": 2.266546664871602e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 21880544,
      "step": 33470
    },
    {
      "epoch": 17.54454926624738,
      "grad_norm": 2.5275566577911377,
      "learning_rate": 2.261791718484102e-06,
      "loss": 0.0272,
      "num_input_tokens_seen": 21883296,
      "step": 33475
    },
    {
      "epoch": 17.547169811320753,
      "grad_norm": 0.8514797687530518,
      "learning_rate": 2.257041528670362e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 21885888,
      "step": 33480
    },
    {
      "epoch": 17.54979035639413,
      "grad_norm": 0.4958018362522125,
      "learning_rate": 2.2522960964240753e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 21888736,
      "step": 33485
    },
    {
      "epoch": 17.552410901467507,
      "grad_norm": 0.12129591405391693,
      "learning_rate": 2.247555422737924e-06,
      "loss": 0.0077,
      "num_input_tokens_seen": 21891264,
      "step": 33490
    },
    {
      "epoch": 17.55503144654088,
      "grad_norm": 1.87421715259552,
      "learning_rate": 2.2428195086036024e-06,
      "loss": 0.0133,
      "num_input_tokens_seen": 21893760,
      "step": 33495
    },
    {
      "epoch": 17.557651991614257,
      "grad_norm": 0.6586657762527466,
      "learning_rate": 2.238088355011811e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 21897280,
      "step": 33500
    },
    {
      "epoch": 17.56027253668763,
      "grad_norm": 0.7095693945884705,
      "learning_rate": 2.2333619629522518e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 21900640,
      "step": 33505
    },
    {
      "epoch": 17.562893081761008,
      "grad_norm": 0.9188531637191772,
      "learning_rate": 2.2286403334136424e-06,
      "loss": 0.0067,
      "num_input_tokens_seen": 21903744,
      "step": 33510
    },
    {
      "epoch": 17.56551362683438,
      "grad_norm": 0.3158656060695648,
      "learning_rate": 2.223923467383679e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 21905984,
      "step": 33515
    },
    {
      "epoch": 17.568134171907758,
      "grad_norm": 0.30537283420562744,
      "learning_rate": 2.2192113658490886e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 21910272,
      "step": 33520
    },
    {
      "epoch": 17.57075471698113,
      "grad_norm": 0.07524927705526352,
      "learning_rate": 2.2145040297955776e-06,
      "loss": 0.0389,
      "num_input_tokens_seen": 21913216,
      "step": 33525
    },
    {
      "epoch": 17.57337526205451,
      "grad_norm": 1.6995259523391724,
      "learning_rate": 2.209801460207875e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 21916128,
      "step": 33530
    },
    {
      "epoch": 17.57599580712788,
      "grad_norm": 0.5596534013748169,
      "learning_rate": 2.2051036580697042e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 21918592,
      "step": 33535
    },
    {
      "epoch": 17.57861635220126,
      "grad_norm": 0.3199828565120697,
      "learning_rate": 2.2004106243637885e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 21921216,
      "step": 33540
    },
    {
      "epoch": 17.581236897274632,
      "grad_norm": 0.4513566792011261,
      "learning_rate": 2.1957223600718587e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 21925056,
      "step": 33545
    },
    {
      "epoch": 17.58385744234801,
      "grad_norm": 0.2573901116847992,
      "learning_rate": 2.1910388661746493e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 21928320,
      "step": 33550
    },
    {
      "epoch": 17.586477987421382,
      "grad_norm": 1.1291688680648804,
      "learning_rate": 2.186360143651886e-06,
      "loss": 0.0204,
      "num_input_tokens_seen": 21931968,
      "step": 33555
    },
    {
      "epoch": 17.58909853249476,
      "grad_norm": 0.22441062331199646,
      "learning_rate": 2.1816861934823103e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21935008,
      "step": 33560
    },
    {
      "epoch": 17.591719077568133,
      "grad_norm": 1.7959887981414795,
      "learning_rate": 2.1770170166436595e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 21937344,
      "step": 33565
    },
    {
      "epoch": 17.59433962264151,
      "grad_norm": 0.22596067190170288,
      "learning_rate": 2.172352614112669e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 21939936,
      "step": 33570
    },
    {
      "epoch": 17.596960167714883,
      "grad_norm": 0.31358015537261963,
      "learning_rate": 2.167692986865083e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 21942400,
      "step": 33575
    },
    {
      "epoch": 17.59958071278826,
      "grad_norm": 0.20082570612430573,
      "learning_rate": 2.1630381358756362e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 21945632,
      "step": 33580
    },
    {
      "epoch": 17.602201257861637,
      "grad_norm": 2.1141269207000732,
      "learning_rate": 2.158388062118072e-06,
      "loss": 0.0167,
      "num_input_tokens_seen": 21949312,
      "step": 33585
    },
    {
      "epoch": 17.60482180293501,
      "grad_norm": 1.9062328338623047,
      "learning_rate": 2.1537427665651393e-06,
      "loss": 0.0231,
      "num_input_tokens_seen": 21955872,
      "step": 33590
    },
    {
      "epoch": 17.607442348008387,
      "grad_norm": 0.580771803855896,
      "learning_rate": 2.1491022501885726e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 21959712,
      "step": 33595
    },
    {
      "epoch": 17.61006289308176,
      "grad_norm": 0.6017234325408936,
      "learning_rate": 2.1444665139591174e-06,
      "loss": 0.0092,
      "num_input_tokens_seen": 21963136,
      "step": 33600
    },
    {
      "epoch": 17.612683438155138,
      "grad_norm": 0.3437265157699585,
      "learning_rate": 2.139835558846523e-06,
      "loss": 0.0518,
      "num_input_tokens_seen": 21966912,
      "step": 33605
    },
    {
      "epoch": 17.61530398322851,
      "grad_norm": 0.22183313965797424,
      "learning_rate": 2.1352093858195225e-06,
      "loss": 0.0025,
      "num_input_tokens_seen": 21969152,
      "step": 33610
    },
    {
      "epoch": 17.617924528301888,
      "grad_norm": 0.23078270256519318,
      "learning_rate": 2.1305879958458676e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 21971744,
      "step": 33615
    },
    {
      "epoch": 17.62054507337526,
      "grad_norm": 0.3736843168735504,
      "learning_rate": 2.1259713898922894e-06,
      "loss": 0.002,
      "num_input_tokens_seen": 21974592,
      "step": 33620
    },
    {
      "epoch": 17.62316561844864,
      "grad_norm": 0.3728029429912567,
      "learning_rate": 2.1213595689245386e-06,
      "loss": 0.0124,
      "num_input_tokens_seen": 21978528,
      "step": 33625
    },
    {
      "epoch": 17.62578616352201,
      "grad_norm": 0.11922096461057663,
      "learning_rate": 2.1167525339073513e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 21981792,
      "step": 33630
    },
    {
      "epoch": 17.62840670859539,
      "grad_norm": 1.6015740633010864,
      "learning_rate": 2.112150285804473e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 21985344,
      "step": 33635
    },
    {
      "epoch": 17.631027253668762,
      "grad_norm": 0.28951430320739746,
      "learning_rate": 2.107552825578632e-06,
      "loss": 0.0154,
      "num_input_tokens_seen": 21988352,
      "step": 33640
    },
    {
      "epoch": 17.63364779874214,
      "grad_norm": 0.2503635883331299,
      "learning_rate": 2.102960154191572e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 21992608,
      "step": 33645
    },
    {
      "epoch": 17.636268343815512,
      "grad_norm": 0.3369728624820709,
      "learning_rate": 2.098372272604021e-06,
      "loss": 0.0162,
      "num_input_tokens_seen": 21996160,
      "step": 33650
    },
    {
      "epoch": 17.63888888888889,
      "grad_norm": 0.3423464596271515,
      "learning_rate": 2.0937891817757155e-06,
      "loss": 0.0037,
      "num_input_tokens_seen": 21999712,
      "step": 33655
    },
    {
      "epoch": 17.641509433962263,
      "grad_norm": 0.1004156619310379,
      "learning_rate": 2.089210882665385e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 22002432,
      "step": 33660
    },
    {
      "epoch": 17.64412997903564,
      "grad_norm": 0.4597180187702179,
      "learning_rate": 2.0846373762307635e-06,
      "loss": 0.0117,
      "num_input_tokens_seen": 22004992,
      "step": 33665
    },
    {
      "epoch": 17.646750524109013,
      "grad_norm": 0.6232246160507202,
      "learning_rate": 2.080068663428569e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 22008480,
      "step": 33670
    },
    {
      "epoch": 17.64937106918239,
      "grad_norm": 1.3019558191299438,
      "learning_rate": 2.075504745214521e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 22011968,
      "step": 33675
    },
    {
      "epoch": 17.651991614255767,
      "grad_norm": 0.8225328326225281,
      "learning_rate": 2.0709456225433437e-06,
      "loss": 0.0144,
      "num_input_tokens_seen": 22015424,
      "step": 33680
    },
    {
      "epoch": 17.65461215932914,
      "grad_norm": 0.4843885004520416,
      "learning_rate": 2.0663912963687554e-06,
      "loss": 0.0089,
      "num_input_tokens_seen": 22019776,
      "step": 33685
    },
    {
      "epoch": 17.657232704402517,
      "grad_norm": 0.10512956231832504,
      "learning_rate": 2.0618417676434713e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 22023072,
      "step": 33690
    },
    {
      "epoch": 17.65985324947589,
      "grad_norm": 0.6168329119682312,
      "learning_rate": 2.0572970373191975e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 22026272,
      "step": 33695
    },
    {
      "epoch": 17.662473794549268,
      "grad_norm": 1.3677767515182495,
      "learning_rate": 2.052757106346634e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 22029216,
      "step": 33700
    },
    {
      "epoch": 17.66509433962264,
      "grad_norm": 0.752845823764801,
      "learning_rate": 2.048221975675488e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 22032416,
      "step": 33705
    },
    {
      "epoch": 17.667714884696018,
      "grad_norm": 0.24444319307804108,
      "learning_rate": 2.043691646254456e-06,
      "loss": 0.0024,
      "num_input_tokens_seen": 22035488,
      "step": 33710
    },
    {
      "epoch": 17.67033542976939,
      "grad_norm": 0.9367683529853821,
      "learning_rate": 2.039166119031233e-06,
      "loss": 0.0152,
      "num_input_tokens_seen": 22039232,
      "step": 33715
    },
    {
      "epoch": 17.67295597484277,
      "grad_norm": 0.13572049140930176,
      "learning_rate": 2.034645394952511e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 22041888,
      "step": 33720
    },
    {
      "epoch": 17.67557651991614,
      "grad_norm": 0.4906250238418579,
      "learning_rate": 2.030129474963971e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 22044416,
      "step": 33725
    },
    {
      "epoch": 17.67819706498952,
      "grad_norm": 0.09134310483932495,
      "learning_rate": 2.025618360010284e-06,
      "loss": 0.0016,
      "num_input_tokens_seen": 22047552,
      "step": 33730
    },
    {
      "epoch": 17.680817610062892,
      "grad_norm": 0.8366460800170898,
      "learning_rate": 2.0211120510351317e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 22049856,
      "step": 33735
    },
    {
      "epoch": 17.68343815513627,
      "grad_norm": 0.32075998187065125,
      "learning_rate": 2.0166105489811822e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 22052672,
      "step": 33740
    },
    {
      "epoch": 17.686058700209642,
      "grad_norm": 0.9117177724838257,
      "learning_rate": 2.0121138547901016e-06,
      "loss": 0.0143,
      "num_input_tokens_seen": 22055392,
      "step": 33745
    },
    {
      "epoch": 17.68867924528302,
      "grad_norm": 0.20893393456935883,
      "learning_rate": 2.0076219694025427e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 22058176,
      "step": 33750
    },
    {
      "epoch": 17.691299790356393,
      "grad_norm": 0.12737874686717987,
      "learning_rate": 2.003134893758152e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 22061152,
      "step": 33755
    },
    {
      "epoch": 17.69392033542977,
      "grad_norm": 0.12968342006206512,
      "learning_rate": 1.9986526287955805e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 22064288,
      "step": 33760
    },
    {
      "epoch": 17.696540880503143,
      "grad_norm": 0.2806413769721985,
      "learning_rate": 1.9941751754524644e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 22067008,
      "step": 33765
    },
    {
      "epoch": 17.69916142557652,
      "grad_norm": 0.3628174960613251,
      "learning_rate": 1.9897025346654412e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 22070400,
      "step": 33770
    },
    {
      "epoch": 17.701781970649897,
      "grad_norm": 2.6921072006225586,
      "learning_rate": 1.985234707370134e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 22073312,
      "step": 33775
    },
    {
      "epoch": 17.70440251572327,
      "grad_norm": 0.6821253299713135,
      "learning_rate": 1.9807716945011544e-06,
      "loss": 0.0086,
      "num_input_tokens_seen": 22076480,
      "step": 33780
    },
    {
      "epoch": 17.707023060796647,
      "grad_norm": 0.16973042488098145,
      "learning_rate": 1.9763134969921215e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22079968,
      "step": 33785
    },
    {
      "epoch": 17.70964360587002,
      "grad_norm": 0.11322102695703506,
      "learning_rate": 1.971860115775634e-06,
      "loss": 0.002,
      "num_input_tokens_seen": 22083136,
      "step": 33790
    },
    {
      "epoch": 17.712264150943398,
      "grad_norm": 0.3681521713733673,
      "learning_rate": 1.967411551783294e-06,
      "loss": 0.0105,
      "num_input_tokens_seen": 22086400,
      "step": 33795
    },
    {
      "epoch": 17.71488469601677,
      "grad_norm": 0.47065165638923645,
      "learning_rate": 1.9629678059456895e-06,
      "loss": 0.0116,
      "num_input_tokens_seen": 22089632,
      "step": 33800
    },
    {
      "epoch": 17.717505241090148,
      "grad_norm": 0.9155102372169495,
      "learning_rate": 1.9585288791924004e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 22092576,
      "step": 33805
    },
    {
      "epoch": 17.72012578616352,
      "grad_norm": 0.5957792401313782,
      "learning_rate": 1.9540947724519947e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 22096768,
      "step": 33810
    },
    {
      "epoch": 17.7227463312369,
      "grad_norm": 1.0657894611358643,
      "learning_rate": 1.9496654866520414e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 22099296,
      "step": 33815
    },
    {
      "epoch": 17.72536687631027,
      "grad_norm": 0.21707946062088013,
      "learning_rate": 1.9452410227190937e-06,
      "loss": 0.0438,
      "num_input_tokens_seen": 22101888,
      "step": 33820
    },
    {
      "epoch": 17.72798742138365,
      "grad_norm": 0.2682397663593292,
      "learning_rate": 1.9408213815787087e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 22105568,
      "step": 33825
    },
    {
      "epoch": 17.730607966457022,
      "grad_norm": 2.0632877349853516,
      "learning_rate": 1.936406564155413e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 22109248,
      "step": 33830
    },
    {
      "epoch": 17.7332285115304,
      "grad_norm": 0.2273605912923813,
      "learning_rate": 1.9319965713727435e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22112352,
      "step": 33835
    },
    {
      "epoch": 17.735849056603772,
      "grad_norm": 0.5220072269439697,
      "learning_rate": 1.9275914041532144e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 22115200,
      "step": 33840
    },
    {
      "epoch": 17.73846960167715,
      "grad_norm": 0.331553190946579,
      "learning_rate": 1.9231910634183394e-06,
      "loss": 0.0202,
      "num_input_tokens_seen": 22118272,
      "step": 33845
    },
    {
      "epoch": 17.741090146750523,
      "grad_norm": 0.2514130473136902,
      "learning_rate": 1.918795550088623e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 22121248,
      "step": 33850
    },
    {
      "epoch": 17.7437106918239,
      "grad_norm": 0.7754105925559998,
      "learning_rate": 1.9144048650835496e-06,
      "loss": 0.0094,
      "num_input_tokens_seen": 22124416,
      "step": 33855
    },
    {
      "epoch": 17.746331236897273,
      "grad_norm": 0.5069817304611206,
      "learning_rate": 1.9100190093216093e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 22127264,
      "step": 33860
    },
    {
      "epoch": 17.74895178197065,
      "grad_norm": 1.1522084474563599,
      "learning_rate": 1.9056379837202627e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 22130816,
      "step": 33865
    },
    {
      "epoch": 17.751572327044027,
      "grad_norm": 2.662755012512207,
      "learning_rate": 1.9012617891959766e-06,
      "loss": 0.0163,
      "num_input_tokens_seen": 22135200,
      "step": 33870
    },
    {
      "epoch": 17.7541928721174,
      "grad_norm": 3.2454028129577637,
      "learning_rate": 1.8968904266641996e-06,
      "loss": 0.0235,
      "num_input_tokens_seen": 22138304,
      "step": 33875
    },
    {
      "epoch": 17.756813417190777,
      "grad_norm": 0.4511446952819824,
      "learning_rate": 1.892523897039372e-06,
      "loss": 0.0019,
      "num_input_tokens_seen": 22141024,
      "step": 33880
    },
    {
      "epoch": 17.75943396226415,
      "grad_norm": 0.4141886532306671,
      "learning_rate": 1.888162201234922e-06,
      "loss": 0.0156,
      "num_input_tokens_seen": 22144800,
      "step": 33885
    },
    {
      "epoch": 17.762054507337528,
      "grad_norm": 4.091605186462402,
      "learning_rate": 1.8838053401632666e-06,
      "loss": 0.0985,
      "num_input_tokens_seen": 22148256,
      "step": 33890
    },
    {
      "epoch": 17.7646750524109,
      "grad_norm": 0.5525266528129578,
      "learning_rate": 1.8794533147358074e-06,
      "loss": 0.0146,
      "num_input_tokens_seen": 22151840,
      "step": 33895
    },
    {
      "epoch": 17.767295597484278,
      "grad_norm": 0.16347342729568481,
      "learning_rate": 1.875106125862941e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 22155616,
      "step": 33900
    },
    {
      "epoch": 17.76991614255765,
      "grad_norm": 0.09072303771972656,
      "learning_rate": 1.870763774454057e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 22158656,
      "step": 33905
    },
    {
      "epoch": 17.77253668763103,
      "grad_norm": 0.3715427815914154,
      "learning_rate": 1.8664262614175115e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22161152,
      "step": 33910
    },
    {
      "epoch": 17.7751572327044,
      "grad_norm": 0.30173712968826294,
      "learning_rate": 1.862093587660671e-06,
      "loss": 0.009,
      "num_input_tokens_seen": 22164544,
      "step": 33915
    },
    {
      "epoch": 17.77777777777778,
      "grad_norm": 0.5426985621452332,
      "learning_rate": 1.8577657540898851e-06,
      "loss": 0.0121,
      "num_input_tokens_seen": 22167712,
      "step": 33920
    },
    {
      "epoch": 17.780398322851152,
      "grad_norm": 0.4980272054672241,
      "learning_rate": 1.8534427616104798e-06,
      "loss": 0.0116,
      "num_input_tokens_seen": 22171648,
      "step": 33925
    },
    {
      "epoch": 17.78301886792453,
      "grad_norm": 0.3646451234817505,
      "learning_rate": 1.8491246111267846e-06,
      "loss": 0.0838,
      "num_input_tokens_seen": 22178912,
      "step": 33930
    },
    {
      "epoch": 17.785639412997902,
      "grad_norm": 0.7631343603134155,
      "learning_rate": 1.8448113035420934e-06,
      "loss": 0.0227,
      "num_input_tokens_seen": 22181568,
      "step": 33935
    },
    {
      "epoch": 17.78825995807128,
      "grad_norm": 0.15974026918411255,
      "learning_rate": 1.840502839758712e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22185408,
      "step": 33940
    },
    {
      "epoch": 17.790880503144653,
      "grad_norm": 0.17393945157527924,
      "learning_rate": 1.8361992206779226e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 22188224,
      "step": 33945
    },
    {
      "epoch": 17.79350104821803,
      "grad_norm": 1.4833624362945557,
      "learning_rate": 1.8319004471999856e-06,
      "loss": 0.0079,
      "num_input_tokens_seen": 22190688,
      "step": 33950
    },
    {
      "epoch": 17.796121593291403,
      "grad_norm": 0.411803662776947,
      "learning_rate": 1.8276065202241592e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 22193184,
      "step": 33955
    },
    {
      "epoch": 17.79874213836478,
      "grad_norm": 0.08635004609823227,
      "learning_rate": 1.8233174406486863e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 22195520,
      "step": 33960
    },
    {
      "epoch": 17.801362683438157,
      "grad_norm": 0.9469172358512878,
      "learning_rate": 1.8190332093707907e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 22198208,
      "step": 33965
    },
    {
      "epoch": 17.80398322851153,
      "grad_norm": 3.051386594772339,
      "learning_rate": 1.8147538272866833e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 22201792,
      "step": 33970
    },
    {
      "epoch": 17.806603773584907,
      "grad_norm": 0.36068642139434814,
      "learning_rate": 1.8104792952915672e-06,
      "loss": 0.0156,
      "num_input_tokens_seen": 22205632,
      "step": 33975
    },
    {
      "epoch": 17.80922431865828,
      "grad_norm": 0.27956464886665344,
      "learning_rate": 1.8062096142796192e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 22208192,
      "step": 33980
    },
    {
      "epoch": 17.811844863731658,
      "grad_norm": 1.44036066532135,
      "learning_rate": 1.8019447851440163e-06,
      "loss": 0.0112,
      "num_input_tokens_seen": 22211584,
      "step": 33985
    },
    {
      "epoch": 17.81446540880503,
      "grad_norm": 1.0856468677520752,
      "learning_rate": 1.7976848087769032e-06,
      "loss": 0.0204,
      "num_input_tokens_seen": 22215360,
      "step": 33990
    },
    {
      "epoch": 17.817085953878408,
      "grad_norm": 0.8975173830986023,
      "learning_rate": 1.79342968606942e-06,
      "loss": 0.0202,
      "num_input_tokens_seen": 22218848,
      "step": 33995
    },
    {
      "epoch": 17.81970649895178,
      "grad_norm": 0.38085681200027466,
      "learning_rate": 1.7891794179116933e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 22222880,
      "step": 34000
    },
    {
      "epoch": 17.822327044025158,
      "grad_norm": 1.6147385835647583,
      "learning_rate": 1.7849340051928316e-06,
      "loss": 0.0273,
      "num_input_tokens_seen": 22229120,
      "step": 34005
    },
    {
      "epoch": 17.82494758909853,
      "grad_norm": 0.21035881340503693,
      "learning_rate": 1.780693448800927e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 22232352,
      "step": 34010
    },
    {
      "epoch": 17.82756813417191,
      "grad_norm": 0.14708919823169708,
      "learning_rate": 1.7764577496230478e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 22237856,
      "step": 34015
    },
    {
      "epoch": 17.830188679245282,
      "grad_norm": 0.21534667909145355,
      "learning_rate": 1.7722269085452602e-06,
      "loss": 0.0131,
      "num_input_tokens_seen": 22240992,
      "step": 34020
    },
    {
      "epoch": 17.83280922431866,
      "grad_norm": 0.07315356284379959,
      "learning_rate": 1.7680009264526038e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 22243776,
      "step": 34025
    },
    {
      "epoch": 17.835429769392032,
      "grad_norm": 0.30290108919143677,
      "learning_rate": 1.7637798042291126e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22246592,
      "step": 34030
    },
    {
      "epoch": 17.83805031446541,
      "grad_norm": 0.3941293954849243,
      "learning_rate": 1.7595635427577944e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 22249664,
      "step": 34035
    },
    {
      "epoch": 17.840670859538783,
      "grad_norm": 0.11419674754142761,
      "learning_rate": 1.7553521429206465e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 22253312,
      "step": 34040
    },
    {
      "epoch": 17.84329140461216,
      "grad_norm": 1.4172642230987549,
      "learning_rate": 1.7511456055986335e-06,
      "loss": 0.0152,
      "num_input_tokens_seen": 22256768,
      "step": 34045
    },
    {
      "epoch": 17.845911949685533,
      "grad_norm": 0.6683011054992676,
      "learning_rate": 1.746943931671724e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 22260224,
      "step": 34050
    },
    {
      "epoch": 17.84853249475891,
      "grad_norm": 0.22409719228744507,
      "learning_rate": 1.742747122018859e-06,
      "loss": 0.0103,
      "num_input_tokens_seen": 22262912,
      "step": 34055
    },
    {
      "epoch": 17.851153039832283,
      "grad_norm": 0.005287005100399256,
      "learning_rate": 1.738555177517967e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 22266240,
      "step": 34060
    },
    {
      "epoch": 17.85377358490566,
      "grad_norm": 0.39103978872299194,
      "learning_rate": 1.734368099045952e-06,
      "loss": 0.0118,
      "num_input_tokens_seen": 22269472,
      "step": 34065
    },
    {
      "epoch": 17.856394129979037,
      "grad_norm": 0.42811858654022217,
      "learning_rate": 1.730185887478697e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 22273184,
      "step": 34070
    },
    {
      "epoch": 17.85901467505241,
      "grad_norm": 1.486480951309204,
      "learning_rate": 1.7260085436910794e-06,
      "loss": 0.0238,
      "num_input_tokens_seen": 22280032,
      "step": 34075
    },
    {
      "epoch": 17.861635220125788,
      "grad_norm": 0.3896450400352478,
      "learning_rate": 1.7218360685569506e-06,
      "loss": 0.0177,
      "num_input_tokens_seen": 22284128,
      "step": 34080
    },
    {
      "epoch": 17.86425576519916,
      "grad_norm": 0.5037846565246582,
      "learning_rate": 1.7176684629491513e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 22286880,
      "step": 34085
    },
    {
      "epoch": 17.866876310272538,
      "grad_norm": 0.16970033943653107,
      "learning_rate": 1.7135057277394868e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 22290016,
      "step": 34090
    },
    {
      "epoch": 17.86949685534591,
      "grad_norm": 0.4178564250469208,
      "learning_rate": 1.7093478637987548e-06,
      "loss": 0.0156,
      "num_input_tokens_seen": 22294688,
      "step": 34095
    },
    {
      "epoch": 17.872117400419288,
      "grad_norm": 0.5863781571388245,
      "learning_rate": 1.7051948719967375e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 22297856,
      "step": 34100
    },
    {
      "epoch": 17.87473794549266,
      "grad_norm": 1.092299222946167,
      "learning_rate": 1.7010467532021901e-06,
      "loss": 0.0154,
      "num_input_tokens_seen": 22301024,
      "step": 34105
    },
    {
      "epoch": 17.87735849056604,
      "grad_norm": 0.7208559513092041,
      "learning_rate": 1.6969035082828544e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 22303456,
      "step": 34110
    },
    {
      "epoch": 17.879979035639412,
      "grad_norm": 0.45313745737075806,
      "learning_rate": 1.692765138105451e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 22306528,
      "step": 34115
    },
    {
      "epoch": 17.88259958071279,
      "grad_norm": 1.0350114107131958,
      "learning_rate": 1.6886316435356792e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 22309824,
      "step": 34120
    },
    {
      "epoch": 17.885220125786162,
      "grad_norm": 0.09965566545724869,
      "learning_rate": 1.6845030254382138e-06,
      "loss": 0.003,
      "num_input_tokens_seen": 22313024,
      "step": 34125
    },
    {
      "epoch": 17.88784067085954,
      "grad_norm": 2.4072837829589844,
      "learning_rate": 1.6803792846767197e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 22315584,
      "step": 34130
    },
    {
      "epoch": 17.890461215932913,
      "grad_norm": 0.16494330763816833,
      "learning_rate": 1.6762604221138316e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 22318144,
      "step": 34135
    },
    {
      "epoch": 17.89308176100629,
      "grad_norm": 0.4443223774433136,
      "learning_rate": 1.6721464386111774e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 22320800,
      "step": 34140
    },
    {
      "epoch": 17.895702306079663,
      "grad_norm": 1.8142105340957642,
      "learning_rate": 1.6680373350293516e-06,
      "loss": 0.119,
      "num_input_tokens_seen": 22324064,
      "step": 34145
    },
    {
      "epoch": 17.89832285115304,
      "grad_norm": 0.08666913211345673,
      "learning_rate": 1.6639331122279223e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 22327104,
      "step": 34150
    },
    {
      "epoch": 17.900943396226417,
      "grad_norm": 0.4499680697917938,
      "learning_rate": 1.6598337710654583e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22330528,
      "step": 34155
    },
    {
      "epoch": 17.90356394129979,
      "grad_norm": 0.27497759461402893,
      "learning_rate": 1.6557393123994874e-06,
      "loss": 0.0055,
      "num_input_tokens_seen": 22333472,
      "step": 34160
    },
    {
      "epoch": 17.906184486373167,
      "grad_norm": 0.1092856302857399,
      "learning_rate": 1.651649737086533e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 22336192,
      "step": 34165
    },
    {
      "epoch": 17.90880503144654,
      "grad_norm": 1.2556232213974,
      "learning_rate": 1.6475650459820774e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 22338720,
      "step": 34170
    },
    {
      "epoch": 17.911425576519918,
      "grad_norm": 0.9706639647483826,
      "learning_rate": 1.643485239940598e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22341152,
      "step": 34175
    },
    {
      "epoch": 17.91404612159329,
      "grad_norm": 0.46907269954681396,
      "learning_rate": 1.6394103198155375e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 22344352,
      "step": 34180
    },
    {
      "epoch": 17.916666666666668,
      "grad_norm": 0.7943724393844604,
      "learning_rate": 1.6353402864593276e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 22346944,
      "step": 34185
    },
    {
      "epoch": 17.91928721174004,
      "grad_norm": 0.5236281156539917,
      "learning_rate": 1.6312751407233706e-06,
      "loss": 0.012,
      "num_input_tokens_seen": 22351040,
      "step": 34190
    },
    {
      "epoch": 17.921907756813418,
      "grad_norm": 0.6553868651390076,
      "learning_rate": 1.6272148834580559e-06,
      "loss": 0.0217,
      "num_input_tokens_seen": 22353728,
      "step": 34195
    },
    {
      "epoch": 17.92452830188679,
      "grad_norm": 3.2589364051818848,
      "learning_rate": 1.6231595155127317e-06,
      "loss": 0.0098,
      "num_input_tokens_seen": 22356928,
      "step": 34200
    },
    {
      "epoch": 17.92714884696017,
      "grad_norm": 0.26174452900886536,
      "learning_rate": 1.6191090377357414e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 22359168,
      "step": 34205
    },
    {
      "epoch": 17.929769392033542,
      "grad_norm": 0.19618235528469086,
      "learning_rate": 1.6150634509743939e-06,
      "loss": 0.0061,
      "num_input_tokens_seen": 22362400,
      "step": 34210
    },
    {
      "epoch": 17.93238993710692,
      "grad_norm": 0.21752193570137024,
      "learning_rate": 1.611022756074984e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22365408,
      "step": 34215
    },
    {
      "epoch": 17.935010482180292,
      "grad_norm": 4.530148029327393,
      "learning_rate": 1.606986953882783e-06,
      "loss": 0.0078,
      "num_input_tokens_seen": 22368768,
      "step": 34220
    },
    {
      "epoch": 17.93763102725367,
      "grad_norm": 1.393035650253296,
      "learning_rate": 1.6029560452420238e-06,
      "loss": 0.0239,
      "num_input_tokens_seen": 22372128,
      "step": 34225
    },
    {
      "epoch": 17.940251572327043,
      "grad_norm": 0.2952802777290344,
      "learning_rate": 1.5989300309959377e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 22375520,
      "step": 34230
    },
    {
      "epoch": 17.94287211740042,
      "grad_norm": 0.3580104112625122,
      "learning_rate": 1.5949089119867116e-06,
      "loss": 0.0214,
      "num_input_tokens_seen": 22378656,
      "step": 34235
    },
    {
      "epoch": 17.945492662473793,
      "grad_norm": 0.40339982509613037,
      "learning_rate": 1.5908926890555204e-06,
      "loss": 0.0069,
      "num_input_tokens_seen": 22381888,
      "step": 34240
    },
    {
      "epoch": 17.94811320754717,
      "grad_norm": 0.37017250061035156,
      "learning_rate": 1.5868813630425167e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 22384928,
      "step": 34245
    },
    {
      "epoch": 17.950733752620543,
      "grad_norm": 0.0960448756814003,
      "learning_rate": 1.5828749347868183e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 22388096,
      "step": 34250
    },
    {
      "epoch": 17.95335429769392,
      "grad_norm": 0.4719812273979187,
      "learning_rate": 1.5788734051265268e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 22393632,
      "step": 34255
    },
    {
      "epoch": 17.955974842767297,
      "grad_norm": 1.2298206090927124,
      "learning_rate": 1.5748767748987231e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 22396768,
      "step": 34260
    },
    {
      "epoch": 17.95859538784067,
      "grad_norm": 0.6267217993736267,
      "learning_rate": 1.5708850449394436e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 22399776,
      "step": 34265
    },
    {
      "epoch": 17.961215932914047,
      "grad_norm": 3.7070975303649902,
      "learning_rate": 1.5668982160837176e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 22403104,
      "step": 34270
    },
    {
      "epoch": 17.96383647798742,
      "grad_norm": 0.08557040989398956,
      "learning_rate": 1.562916289165553e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22406592,
      "step": 34275
    },
    {
      "epoch": 17.966457023060798,
      "grad_norm": 0.6514779925346375,
      "learning_rate": 1.5589392650179114e-06,
      "loss": 0.0122,
      "num_input_tokens_seen": 22411264,
      "step": 34280
    },
    {
      "epoch": 17.96907756813417,
      "grad_norm": 0.26065751910209656,
      "learning_rate": 1.5549671444727437e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 22414976,
      "step": 34285
    },
    {
      "epoch": 17.971698113207548,
      "grad_norm": 1.6193984746932983,
      "learning_rate": 1.55099992836098e-06,
      "loss": 0.0318,
      "num_input_tokens_seen": 22418016,
      "step": 34290
    },
    {
      "epoch": 17.97431865828092,
      "grad_norm": 0.6259325742721558,
      "learning_rate": 1.5470376175125061e-06,
      "loss": 0.0131,
      "num_input_tokens_seen": 22421408,
      "step": 34295
    },
    {
      "epoch": 17.9769392033543,
      "grad_norm": 0.4374333322048187,
      "learning_rate": 1.5430802127562038e-06,
      "loss": 0.0638,
      "num_input_tokens_seen": 22423744,
      "step": 34300
    },
    {
      "epoch": 17.979559748427672,
      "grad_norm": 0.7588679790496826,
      "learning_rate": 1.5391277149199023e-06,
      "loss": 0.0057,
      "num_input_tokens_seen": 22428672,
      "step": 34305
    },
    {
      "epoch": 17.98218029350105,
      "grad_norm": 2.9649012088775635,
      "learning_rate": 1.5351801248304292e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 22431648,
      "step": 34310
    },
    {
      "epoch": 17.984800838574422,
      "grad_norm": 0.18005824089050293,
      "learning_rate": 1.5312374433135767e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 22434496,
      "step": 34315
    },
    {
      "epoch": 17.9874213836478,
      "grad_norm": 0.4819847643375397,
      "learning_rate": 1.5272996711941018e-06,
      "loss": 0.002,
      "num_input_tokens_seen": 22438432,
      "step": 34320
    },
    {
      "epoch": 17.990041928721173,
      "grad_norm": 0.33249545097351074,
      "learning_rate": 1.5233668092957482e-06,
      "loss": 0.0062,
      "num_input_tokens_seen": 22441600,
      "step": 34325
    },
    {
      "epoch": 17.99266247379455,
      "grad_norm": 0.23882487416267395,
      "learning_rate": 1.5194388584412217e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 22444160,
      "step": 34330
    },
    {
      "epoch": 17.995283018867923,
      "grad_norm": 0.6617596745491028,
      "learning_rate": 1.5155158194522013e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 22447328,
      "step": 34335
    },
    {
      "epoch": 17.9979035639413,
      "grad_norm": 0.30303525924682617,
      "learning_rate": 1.5115976931493498e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 22450112,
      "step": 34340
    },
    {
      "epoch": 18.0,
      "eval_loss": 1.5930230617523193,
      "eval_runtime": 11.7552,
      "eval_samples_per_second": 72.138,
      "eval_steps_per_second": 18.035,
      "num_input_tokens_seen": 22451928,
      "step": 34344
    },
    {
      "epoch": 18.000524109014677,
      "grad_norm": 0.2221655547618866,
      "learning_rate": 1.5076844803522922e-06,
      "loss": 0.0024,
      "num_input_tokens_seen": 22452408,
      "step": 34345
    },
    {
      "epoch": 18.00314465408805,
      "grad_norm": 1.5442392826080322,
      "learning_rate": 1.5037761818796236e-06,
      "loss": 0.0125,
      "num_input_tokens_seen": 22456184,
      "step": 34350
    },
    {
      "epoch": 18.005765199161427,
      "grad_norm": 0.07731743901968002,
      "learning_rate": 1.4998727985489235e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 22458968,
      "step": 34355
    },
    {
      "epoch": 18.0083857442348,
      "grad_norm": 0.6369814872741699,
      "learning_rate": 1.4959743311767249e-06,
      "loss": 0.0072,
      "num_input_tokens_seen": 22462392,
      "step": 34360
    },
    {
      "epoch": 18.011006289308177,
      "grad_norm": 0.3522917330265045,
      "learning_rate": 1.4920807805785502e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 22466360,
      "step": 34365
    },
    {
      "epoch": 18.01362683438155,
      "grad_norm": 0.6898306608200073,
      "learning_rate": 1.4881921475688815e-06,
      "loss": 0.0052,
      "num_input_tokens_seen": 22470104,
      "step": 34370
    },
    {
      "epoch": 18.016247379454928,
      "grad_norm": 0.2807955741882324,
      "learning_rate": 1.4843084329611844e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 22473144,
      "step": 34375
    },
    {
      "epoch": 18.0188679245283,
      "grad_norm": 0.046751078218221664,
      "learning_rate": 1.480429637567879e-06,
      "loss": 0.0224,
      "num_input_tokens_seen": 22475640,
      "step": 34380
    },
    {
      "epoch": 18.021488469601678,
      "grad_norm": 1.1552257537841797,
      "learning_rate": 1.4765557622003684e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 22478712,
      "step": 34385
    },
    {
      "epoch": 18.02410901467505,
      "grad_norm": 2.1870663166046143,
      "learning_rate": 1.4726868076690186e-06,
      "loss": 0.0167,
      "num_input_tokens_seen": 22481624,
      "step": 34390
    },
    {
      "epoch": 18.02672955974843,
      "grad_norm": 0.734443724155426,
      "learning_rate": 1.4688227747831767e-06,
      "loss": 0.0032,
      "num_input_tokens_seen": 22484664,
      "step": 34395
    },
    {
      "epoch": 18.029350104821802,
      "grad_norm": 0.4005950391292572,
      "learning_rate": 1.4649636643511573e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 22487128,
      "step": 34400
    },
    {
      "epoch": 18.03197064989518,
      "grad_norm": 0.43320778012275696,
      "learning_rate": 1.4611094771802364e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 22489528,
      "step": 34405
    },
    {
      "epoch": 18.034591194968552,
      "grad_norm": 0.07536036521196365,
      "learning_rate": 1.457260214076664e-06,
      "loss": 0.0495,
      "num_input_tokens_seen": 22493560,
      "step": 34410
    },
    {
      "epoch": 18.03721174004193,
      "grad_norm": 0.10873979330062866,
      "learning_rate": 1.4534158758456623e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22496408,
      "step": 34415
    },
    {
      "epoch": 18.039832285115303,
      "grad_norm": 0.2616318464279175,
      "learning_rate": 1.4495764632914271e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 22500440,
      "step": 34420
    },
    {
      "epoch": 18.04245283018868,
      "grad_norm": 0.35518428683280945,
      "learning_rate": 1.4457419772171215e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22503640,
      "step": 34425
    },
    {
      "epoch": 18.045073375262053,
      "grad_norm": 0.29498985409736633,
      "learning_rate": 1.4419124184248733e-06,
      "loss": 0.0257,
      "num_input_tokens_seen": 22507768,
      "step": 34430
    },
    {
      "epoch": 18.04769392033543,
      "grad_norm": 0.3315601348876953,
      "learning_rate": 1.4380877877157834e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 22511256,
      "step": 34435
    },
    {
      "epoch": 18.050314465408807,
      "grad_norm": 4.5817413330078125,
      "learning_rate": 1.4342680858899143e-06,
      "loss": 0.0148,
      "num_input_tokens_seen": 22514872,
      "step": 34440
    },
    {
      "epoch": 18.05293501048218,
      "grad_norm": 0.1513662338256836,
      "learning_rate": 1.4304533137463134e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22518008,
      "step": 34445
    },
    {
      "epoch": 18.055555555555557,
      "grad_norm": 0.1960364580154419,
      "learning_rate": 1.4266434720829808e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 22520664,
      "step": 34450
    },
    {
      "epoch": 18.05817610062893,
      "grad_norm": 0.8386849164962769,
      "learning_rate": 1.4228385616968987e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22523096,
      "step": 34455
    },
    {
      "epoch": 18.060796645702307,
      "grad_norm": 0.16016629338264465,
      "learning_rate": 1.419038583384008e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 22526616,
      "step": 34460
    },
    {
      "epoch": 18.06341719077568,
      "grad_norm": 0.14541757106781006,
      "learning_rate": 1.4152435379392148e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 22529560,
      "step": 34465
    },
    {
      "epoch": 18.066037735849058,
      "grad_norm": 0.20295235514640808,
      "learning_rate": 1.411453426156406e-06,
      "loss": 0.0126,
      "num_input_tokens_seen": 22533016,
      "step": 34470
    },
    {
      "epoch": 18.06865828092243,
      "grad_norm": 0.4341767430305481,
      "learning_rate": 1.4076682488284282e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 22535416,
      "step": 34475
    },
    {
      "epoch": 18.071278825995808,
      "grad_norm": 0.40287113189697266,
      "learning_rate": 1.4038880067470978e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 22538936,
      "step": 34480
    },
    {
      "epoch": 18.07389937106918,
      "grad_norm": 0.6988785266876221,
      "learning_rate": 1.400112700703196e-06,
      "loss": 0.0056,
      "num_input_tokens_seen": 22541944,
      "step": 34485
    },
    {
      "epoch": 18.07651991614256,
      "grad_norm": 0.40072450041770935,
      "learning_rate": 1.3963423314864805e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 22544568,
      "step": 34490
    },
    {
      "epoch": 18.079140461215932,
      "grad_norm": 0.2751006782054901,
      "learning_rate": 1.3925768998856587e-06,
      "loss": 0.0022,
      "num_input_tokens_seen": 22547768,
      "step": 34495
    },
    {
      "epoch": 18.08176100628931,
      "grad_norm": 0.21519848704338074,
      "learning_rate": 1.388816406688423e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 22550776,
      "step": 34500
    },
    {
      "epoch": 18.084381551362682,
      "grad_norm": 0.18350611627101898,
      "learning_rate": 1.3850608526814246e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 22554392,
      "step": 34505
    },
    {
      "epoch": 18.08700209643606,
      "grad_norm": 0.6907799243927002,
      "learning_rate": 1.3813102386502853e-06,
      "loss": 0.0037,
      "num_input_tokens_seen": 22557304,
      "step": 34510
    },
    {
      "epoch": 18.089622641509433,
      "grad_norm": 0.256392240524292,
      "learning_rate": 1.3775645653795881e-06,
      "loss": 0.0039,
      "num_input_tokens_seen": 22560536,
      "step": 34515
    },
    {
      "epoch": 18.09224318658281,
      "grad_norm": 0.19099774956703186,
      "learning_rate": 1.373823833652882e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 22563832,
      "step": 34520
    },
    {
      "epoch": 18.094863731656183,
      "grad_norm": 0.20678140223026276,
      "learning_rate": 1.3700880442526876e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22566616,
      "step": 34525
    },
    {
      "epoch": 18.09748427672956,
      "grad_norm": 0.12549330294132233,
      "learning_rate": 1.3663571979604884e-06,
      "loss": 0.0203,
      "num_input_tokens_seen": 22569976,
      "step": 34530
    },
    {
      "epoch": 18.100104821802937,
      "grad_norm": 1.1223219633102417,
      "learning_rate": 1.3626312955567406e-06,
      "loss": 0.008,
      "num_input_tokens_seen": 22573112,
      "step": 34535
    },
    {
      "epoch": 18.10272536687631,
      "grad_norm": 0.10635208338499069,
      "learning_rate": 1.358910337820854e-06,
      "loss": 0.0141,
      "num_input_tokens_seen": 22576152,
      "step": 34540
    },
    {
      "epoch": 18.105345911949687,
      "grad_norm": 0.1674199104309082,
      "learning_rate": 1.3551943255312172e-06,
      "loss": 0.0154,
      "num_input_tokens_seen": 22579672,
      "step": 34545
    },
    {
      "epoch": 18.10796645702306,
      "grad_norm": 1.4580165147781372,
      "learning_rate": 1.3514832594651694e-06,
      "loss": 0.0049,
      "num_input_tokens_seen": 22583096,
      "step": 34550
    },
    {
      "epoch": 18.110587002096437,
      "grad_norm": 1.3000922203063965,
      "learning_rate": 1.347777140399026e-06,
      "loss": 0.0147,
      "num_input_tokens_seen": 22587256,
      "step": 34555
    },
    {
      "epoch": 18.11320754716981,
      "grad_norm": 0.29346397519111633,
      "learning_rate": 1.344075969108069e-06,
      "loss": 0.0221,
      "num_input_tokens_seen": 22591352,
      "step": 34560
    },
    {
      "epoch": 18.115828092243188,
      "grad_norm": 0.29072991013526917,
      "learning_rate": 1.3403797463665352e-06,
      "loss": 0.0019,
      "num_input_tokens_seen": 22594040,
      "step": 34565
    },
    {
      "epoch": 18.11844863731656,
      "grad_norm": 0.47525715827941895,
      "learning_rate": 1.3366884729476365e-06,
      "loss": 0.0988,
      "num_input_tokens_seen": 22597208,
      "step": 34570
    },
    {
      "epoch": 18.121069182389938,
      "grad_norm": 0.07536772638559341,
      "learning_rate": 1.3330021496235467e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 22600536,
      "step": 34575
    },
    {
      "epoch": 18.12368972746331,
      "grad_norm": 0.16678620874881744,
      "learning_rate": 1.3293207771653965e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 22603160,
      "step": 34580
    },
    {
      "epoch": 18.12631027253669,
      "grad_norm": 0.27102145552635193,
      "learning_rate": 1.3256443563432891e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 22606328,
      "step": 34585
    },
    {
      "epoch": 18.128930817610062,
      "grad_norm": 0.24729607999324799,
      "learning_rate": 1.3219728879262956e-06,
      "loss": 0.0064,
      "num_input_tokens_seen": 22609624,
      "step": 34590
    },
    {
      "epoch": 18.13155136268344,
      "grad_norm": 0.12348785996437073,
      "learning_rate": 1.3183063726824375e-06,
      "loss": 0.0242,
      "num_input_tokens_seen": 22612728,
      "step": 34595
    },
    {
      "epoch": 18.134171907756812,
      "grad_norm": 0.4628083109855652,
      "learning_rate": 1.3146448113787151e-06,
      "loss": 0.0353,
      "num_input_tokens_seen": 22615640,
      "step": 34600
    },
    {
      "epoch": 18.13679245283019,
      "grad_norm": 0.36884209513664246,
      "learning_rate": 1.3109882047810773e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 22618360,
      "step": 34605
    },
    {
      "epoch": 18.139412997903563,
      "grad_norm": 0.5065954327583313,
      "learning_rate": 1.307336553654448e-06,
      "loss": 0.006,
      "num_input_tokens_seen": 22621976,
      "step": 34610
    },
    {
      "epoch": 18.14203354297694,
      "grad_norm": 0.9341822862625122,
      "learning_rate": 1.303689858762716e-06,
      "loss": 0.0066,
      "num_input_tokens_seen": 22626008,
      "step": 34615
    },
    {
      "epoch": 18.144654088050313,
      "grad_norm": 0.14504922926425934,
      "learning_rate": 1.3000481208687188e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22629368,
      "step": 34620
    },
    {
      "epoch": 18.14727463312369,
      "grad_norm": 0.17777355015277863,
      "learning_rate": 1.2964113407342743e-06,
      "loss": 0.0053,
      "num_input_tokens_seen": 22632440,
      "step": 34625
    },
    {
      "epoch": 18.149895178197063,
      "grad_norm": 0.35774531960487366,
      "learning_rate": 1.292779519120152e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 22635160,
      "step": 34630
    },
    {
      "epoch": 18.15251572327044,
      "grad_norm": 2.6096532344818115,
      "learning_rate": 1.2891526567860856e-06,
      "loss": 0.0183,
      "num_input_tokens_seen": 22639896,
      "step": 34635
    },
    {
      "epoch": 18.155136268343817,
      "grad_norm": 0.28412067890167236,
      "learning_rate": 1.2855307544907825e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 22642968,
      "step": 34640
    },
    {
      "epoch": 18.15775681341719,
      "grad_norm": 0.0664735957980156,
      "learning_rate": 1.281913812991889e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 22646456,
      "step": 34645
    },
    {
      "epoch": 18.160377358490567,
      "grad_norm": 1.0278761386871338,
      "learning_rate": 1.2783018330460389e-06,
      "loss": 0.0109,
      "num_input_tokens_seen": 22650648,
      "step": 34650
    },
    {
      "epoch": 18.16299790356394,
      "grad_norm": 1.2236319780349731,
      "learning_rate": 1.2746948154088111e-06,
      "loss": 0.0017,
      "num_input_tokens_seen": 22654296,
      "step": 34655
    },
    {
      "epoch": 18.165618448637318,
      "grad_norm": 0.8725722432136536,
      "learning_rate": 1.2710927608347605e-06,
      "loss": 0.0045,
      "num_input_tokens_seen": 22658072,
      "step": 34660
    },
    {
      "epoch": 18.16823899371069,
      "grad_norm": 0.2945808172225952,
      "learning_rate": 1.2674956700773872e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 22661752,
      "step": 34665
    },
    {
      "epoch": 18.170859538784068,
      "grad_norm": 0.3413853049278259,
      "learning_rate": 1.2639035438891666e-06,
      "loss": 0.0105,
      "num_input_tokens_seen": 22665112,
      "step": 34670
    },
    {
      "epoch": 18.17348008385744,
      "grad_norm": 1.0875779390335083,
      "learning_rate": 1.260316383021526e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22667640,
      "step": 34675
    },
    {
      "epoch": 18.17610062893082,
      "grad_norm": 0.21057485044002533,
      "learning_rate": 1.2567341882248645e-06,
      "loss": 0.0076,
      "num_input_tokens_seen": 22670840,
      "step": 34680
    },
    {
      "epoch": 18.178721174004192,
      "grad_norm": 0.4096708297729492,
      "learning_rate": 1.2531569602485305e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 22673720,
      "step": 34685
    },
    {
      "epoch": 18.18134171907757,
      "grad_norm": 0.22991327941417694,
      "learning_rate": 1.2495846998408445e-06,
      "loss": 0.0081,
      "num_input_tokens_seen": 22676984,
      "step": 34690
    },
    {
      "epoch": 18.183962264150942,
      "grad_norm": 0.1948736011981964,
      "learning_rate": 1.246017407749084e-06,
      "loss": 0.0141,
      "num_input_tokens_seen": 22680216,
      "step": 34695
    },
    {
      "epoch": 18.18658280922432,
      "grad_norm": 0.30875328183174133,
      "learning_rate": 1.2424550847194739e-06,
      "loss": 0.0023,
      "num_input_tokens_seen": 22683448,
      "step": 34700
    },
    {
      "epoch": 18.189203354297693,
      "grad_norm": 0.17240677773952484,
      "learning_rate": 1.2388977314972238e-06,
      "loss": 0.0273,
      "num_input_tokens_seen": 22689912,
      "step": 34705
    },
    {
      "epoch": 18.19182389937107,
      "grad_norm": 0.21828225255012512,
      "learning_rate": 1.235345348826486e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 22692696,
      "step": 34710
    },
    {
      "epoch": 18.194444444444443,
      "grad_norm": 0.1888377070426941,
      "learning_rate": 1.2317979374503825e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 22695032,
      "step": 34715
    },
    {
      "epoch": 18.19706498951782,
      "grad_norm": 0.4113215506076813,
      "learning_rate": 1.228255498110989e-06,
      "loss": 0.0041,
      "num_input_tokens_seen": 22698712,
      "step": 34720
    },
    {
      "epoch": 18.199685534591197,
      "grad_norm": 0.11481380462646484,
      "learning_rate": 1.224718031549338e-06,
      "loss": 0.0133,
      "num_input_tokens_seen": 22702200,
      "step": 34725
    },
    {
      "epoch": 18.20230607966457,
      "grad_norm": 0.24290409684181213,
      "learning_rate": 1.2211855385054344e-06,
      "loss": 0.0015,
      "num_input_tokens_seen": 22704888,
      "step": 34730
    },
    {
      "epoch": 18.204926624737947,
      "grad_norm": 0.23717369139194489,
      "learning_rate": 1.217658019718232e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22708280,
      "step": 34735
    },
    {
      "epoch": 18.20754716981132,
      "grad_norm": 0.5660781860351562,
      "learning_rate": 1.2141354759256484e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22711384,
      "step": 34740
    },
    {
      "epoch": 18.210167714884697,
      "grad_norm": 1.2674179077148438,
      "learning_rate": 1.2106179078645641e-06,
      "loss": 0.0115,
      "num_input_tokens_seen": 22714712,
      "step": 34745
    },
    {
      "epoch": 18.21278825995807,
      "grad_norm": 0.4999176263809204,
      "learning_rate": 1.2071053162708096e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 22717432,
      "step": 34750
    },
    {
      "epoch": 18.215408805031448,
      "grad_norm": 0.32615000009536743,
      "learning_rate": 1.203597701879175e-06,
      "loss": 0.0033,
      "num_input_tokens_seen": 22720312,
      "step": 34755
    },
    {
      "epoch": 18.21802935010482,
      "grad_norm": 0.16134986281394958,
      "learning_rate": 1.2000950654234177e-06,
      "loss": 0.0284,
      "num_input_tokens_seen": 22727928,
      "step": 34760
    },
    {
      "epoch": 18.220649895178198,
      "grad_norm": 0.28144752979278564,
      "learning_rate": 1.1965974076362491e-06,
      "loss": 0.0027,
      "num_input_tokens_seen": 22730456,
      "step": 34765
    },
    {
      "epoch": 18.22327044025157,
      "grad_norm": 0.07743538916110992,
      "learning_rate": 1.193104729249342e-06,
      "loss": 0.0104,
      "num_input_tokens_seen": 22733240,
      "step": 34770
    },
    {
      "epoch": 18.22589098532495,
      "grad_norm": 0.3147968351840973,
      "learning_rate": 1.189617030993323e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22736344,
      "step": 34775
    },
    {
      "epoch": 18.228511530398322,
      "grad_norm": 2.330739736557007,
      "learning_rate": 1.1861343135977753e-06,
      "loss": 0.0043,
      "num_input_tokens_seen": 22739320,
      "step": 34780
    },
    {
      "epoch": 18.2311320754717,
      "grad_norm": 0.4123118817806244,
      "learning_rate": 1.182656577791244e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 22743000,
      "step": 34785
    },
    {
      "epoch": 18.233752620545072,
      "grad_norm": 0.23245805501937866,
      "learning_rate": 1.1791838243012359e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22745752,
      "step": 34790
    },
    {
      "epoch": 18.23637316561845,
      "grad_norm": 0.4943154454231262,
      "learning_rate": 1.1757160538542117e-06,
      "loss": 0.0705,
      "num_input_tokens_seen": 22748504,
      "step": 34795
    },
    {
      "epoch": 18.238993710691823,
      "grad_norm": 0.29665595293045044,
      "learning_rate": 1.1722532671755882e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22751704,
      "step": 34800
    },
    {
      "epoch": 18.2416142557652,
      "grad_norm": 0.2370486557483673,
      "learning_rate": 1.168795464989736e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 22755256,
      "step": 34805
    },
    {
      "epoch": 18.244234800838573,
      "grad_norm": 0.19567374885082245,
      "learning_rate": 1.165342648019993e-06,
      "loss": 0.0244,
      "num_input_tokens_seen": 22758616,
      "step": 34810
    },
    {
      "epoch": 18.24685534591195,
      "grad_norm": 0.06521941721439362,
      "learning_rate": 1.1618948169886457e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22761432,
      "step": 34815
    },
    {
      "epoch": 18.249475890985323,
      "grad_norm": 0.5070905089378357,
      "learning_rate": 1.1584519726169446e-06,
      "loss": 0.016,
      "num_input_tokens_seen": 22764376,
      "step": 34820
    },
    {
      "epoch": 18.2520964360587,
      "grad_norm": 0.05666021630167961,
      "learning_rate": 1.1550141156250942e-06,
      "loss": 0.0074,
      "num_input_tokens_seen": 22768056,
      "step": 34825
    },
    {
      "epoch": 18.254716981132077,
      "grad_norm": 1.341241478919983,
      "learning_rate": 1.1515812467322556e-06,
      "loss": 0.0229,
      "num_input_tokens_seen": 22771128,
      "step": 34830
    },
    {
      "epoch": 18.25733752620545,
      "grad_norm": 0.25635847449302673,
      "learning_rate": 1.1481533666565375e-06,
      "loss": 0.0018,
      "num_input_tokens_seen": 22774360,
      "step": 34835
    },
    {
      "epoch": 18.259958071278827,
      "grad_norm": 1.3800873756408691,
      "learning_rate": 1.144730476115019e-06,
      "loss": 0.0051,
      "num_input_tokens_seen": 22778712,
      "step": 34840
    },
    {
      "epoch": 18.2625786163522,
      "grad_norm": 0.2964054346084595,
      "learning_rate": 1.1413125758237304e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 22781368,
      "step": 34845
    },
    {
      "epoch": 18.265199161425578,
      "grad_norm": 3.0024595260620117,
      "learning_rate": 1.1378996664976577e-06,
      "loss": 0.0021,
      "num_input_tokens_seen": 22784120,
      "step": 34850
    },
    {
      "epoch": 18.26781970649895,
      "grad_norm": 0.8333958387374878,
      "learning_rate": 1.134491748850741e-06,
      "loss": 0.01,
      "num_input_tokens_seen": 22787160,
      "step": 34855
    },
    {
      "epoch": 18.270440251572328,
      "grad_norm": 0.21412965655326843,
      "learning_rate": 1.13108882359588e-06,
      "loss": 0.0185,
      "num_input_tokens_seen": 22790584,
      "step": 34860
    },
    {
      "epoch": 18.2730607966457,
      "grad_norm": 0.2943325936794281,
      "learning_rate": 1.127690891444924e-06,
      "loss": 0.0028,
      "num_input_tokens_seen": 22794168,
      "step": 34865
    },
    {
      "epoch": 18.27568134171908,
      "grad_norm": 0.24111467599868774,
      "learning_rate": 1.1242979531086829e-06,
      "loss": 0.0035,
      "num_input_tokens_seen": 22796600,
      "step": 34870
    },
    {
      "epoch": 18.278301886792452,
      "grad_norm": 1.649788498878479,
      "learning_rate": 1.1209100092969244e-06,
      "loss": 0.0058,
      "num_input_tokens_seen": 22799896,
      "step": 34875
    },
    {
      "epoch": 18.28092243186583,
      "grad_norm": 1.6500567197799683,
      "learning_rate": 1.1175270607183625e-06,
      "loss": 0.0131,
      "num_input_tokens_seen": 22803416,
      "step": 34880
    },
    {
      "epoch": 18.283542976939202,
      "grad_norm": 0.2789023220539093,
      "learning_rate": 1.114149108080678e-06,
      "loss": 0.0054,
      "num_input_tokens_seen": 22806552,
      "step": 34885
    },
    {
      "epoch": 18.28616352201258,
      "grad_norm": 0.7218884825706482,
      "learning_rate": 1.1107761520904925e-06,
      "loss": 0.0082,
      "num_input_tokens_seen": 22809496,
      "step": 34890
    },
    {
      "epoch": 18.288784067085953,
      "grad_norm": 0.2973540127277374,
      "learning_rate": 1.1074081934533908e-06,
      "loss": 0.0108,
      "num_input_tokens_seen": 22812664,
      "step": 34895
    },
    {
      "epoch": 18.29140461215933,
      "grad_norm": 0.15853741765022278,
      "learning_rate": 1.1040452328739153e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 22815992,
      "step": 34900
    },
    {
      "epoch": 18.294025157232703,
      "grad_norm": 0.34993231296539307,
      "learning_rate": 1.1006872710555615e-06,
      "loss": 0.0042,
      "num_input_tokens_seen": 22818968,
      "step": 34905
    },
    {
      "epoch": 18.29664570230608,
      "grad_norm": 0.06014801934361458,
      "learning_rate": 1.0973343087007703e-06,
      "loss": 0.0031,
      "num_input_tokens_seen": 22821624,
      "step": 34910
    },
    {
      "epoch": 18.299266247379457,
      "grad_norm": 2.633699417114258,
      "learning_rate": 1.0939863465109473e-06,
      "loss": 0.0236,
      "num_input_tokens_seen": 22824696,
      "step": 34915
    },
    {
      "epoch": 18.30188679245283,
      "grad_norm": 0.1852319985628128,
      "learning_rate": 1.0906433851864433e-06,
      "loss": 0.0029,
      "num_input_tokens_seen": 22827864,
      "step": 34920
    },
    {
      "epoch": 18.304507337526207,
      "grad_norm": 1.160319209098816,
      "learning_rate": 1.0873054254265713e-06,
      "loss": 0.0198,
      "num_input_tokens_seen": 22831384,
      "step": 34925
    },
    {
      "epoch": 18.30712788259958,
      "grad_norm": 1.3502720594406128,
      "learning_rate": 1.0839724679295976e-06,
      "loss": 0.0142,
      "num_input_tokens_seen": 22835928,
      "step": 34930
    },
    {
      "epoch": 18.309748427672957,
      "grad_norm": 0.30053338408470154,
      "learning_rate": 1.0806445133927311e-06,
      "loss": 0.005,
      "num_input_tokens_seen": 22838776,
      "step": 34935
    },
    {
      "epoch": 18.31236897274633,
      "grad_norm": 0.13783033192157745,
      "learning_rate": 1.0773215625121457e-06,
      "loss": 0.0038,
      "num_input_tokens_seen": 22842008,
      "step": 34940
    },
    {
      "epoch": 18.314989517819708,
      "grad_norm": 0.45460811257362366,
      "learning_rate": 1.0740036159829715e-06,
      "loss": 0.0189,
      "num_input_tokens_seen": 22846040,
      "step": 34945
    },
    {
      "epoch": 18.31761006289308,
      "grad_norm": 0.10327528417110443,
      "learning_rate": 1.0706906744992724e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22850328,
      "step": 34950
    },
    {
      "epoch": 18.320230607966458,
      "grad_norm": 0.4006335735321045,
      "learning_rate": 1.0673827387540857e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 22853624,
      "step": 34955
    },
    {
      "epoch": 18.32285115303983,
      "grad_norm": 1.1146870851516724,
      "learning_rate": 1.0640798094393916e-06,
      "loss": 0.0136,
      "num_input_tokens_seen": 22857240,
      "step": 34960
    },
    {
      "epoch": 18.32547169811321,
      "grad_norm": 0.1375211924314499,
      "learning_rate": 1.060781887246126e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 22860216,
      "step": 34965
    },
    {
      "epoch": 18.328092243186582,
      "grad_norm": 0.10917413234710693,
      "learning_rate": 1.057488972864179e-06,
      "loss": 0.0011,
      "num_input_tokens_seen": 22864920,
      "step": 34970
    },
    {
      "epoch": 18.33071278825996,
      "grad_norm": 0.34044018387794495,
      "learning_rate": 1.0542010669823855e-06,
      "loss": 0.0047,
      "num_input_tokens_seen": 22867928,
      "step": 34975
    },
    {
      "epoch": 18.333333333333332,
      "grad_norm": 0.813618540763855,
      "learning_rate": 1.0509181702885395e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22871256,
      "step": 34980
    },
    {
      "epoch": 18.33595387840671,
      "grad_norm": 0.2414018213748932,
      "learning_rate": 1.047640283469392e-06,
      "loss": 0.0172,
      "num_input_tokens_seen": 22875544,
      "step": 34985
    },
    {
      "epoch": 18.338574423480082,
      "grad_norm": 0.6976620554924011,
      "learning_rate": 1.0443674072106308e-06,
      "loss": 0.0119,
      "num_input_tokens_seen": 22878456,
      "step": 34990
    },
    {
      "epoch": 18.34119496855346,
      "grad_norm": 0.30456891655921936,
      "learning_rate": 1.041099542196905e-06,
      "loss": 0.0084,
      "num_input_tokens_seen": 22882168,
      "step": 34995
    },
    {
      "epoch": 18.343815513626833,
      "grad_norm": 0.8285443186759949,
      "learning_rate": 1.0378366891118234e-06,
      "loss": 0.0036,
      "num_input_tokens_seen": 22884824,
      "step": 35000
    },
    {
      "epoch": 18.34643605870021,
      "grad_norm": 0.7952296137809753,
      "learning_rate": 1.0345788486379265e-06,
      "loss": 0.004,
      "num_input_tokens_seen": 22887864,
      "step": 35005
    },
    {
      "epoch": 18.349056603773583,
      "grad_norm": 2.125551223754883,
      "learning_rate": 1.0313260214567272e-06,
      "loss": 0.0466,
      "num_input_tokens_seen": 22892184,
      "step": 35010
    },
    {
      "epoch": 18.35167714884696,
      "grad_norm": 0.18247565627098083,
      "learning_rate": 1.0280782082486729e-06,
      "loss": 0.0073,
      "num_input_tokens_seen": 22895096,
      "step": 35015
    },
    {
      "epoch": 18.354297693920337,
      "grad_norm": 0.07963327318429947,
      "learning_rate": 1.02483540969317e-06,
      "loss": 0.0026,
      "num_input_tokens_seen": 22898168,
      "step": 35020
    },
    {
      "epoch": 18.35691823899371,
      "grad_norm": 0.42359837889671326,
      "learning_rate": 1.0215976264685763e-06,
      "loss": 0.0016,
      "num_input_tokens_seen": 22901368,
      "step": 35025
    },
    {
      "epoch": 18.359538784067087,
      "grad_norm": 0.9724813103675842,
      "learning_rate": 1.0183648592522021e-06,
      "loss": 0.0068,
      "num_input_tokens_seen": 22905048,
      "step": 35030
    },
    {
      "epoch": 18.36215932914046,
      "grad_norm": 0.2227775901556015,
      "learning_rate": 1.0151371087203015e-06,
      "loss": 0.0034,
      "num_input_tokens_seen": 22907800,
      "step": 35035
    },
    {
      "epoch": 18.364779874213838,
      "grad_norm": 1.2130396366119385,
      "learning_rate": 1.0119143755480815e-06,
      "loss": 0.0071,
      "num_input_tokens_seen": 22911480,
      "step": 35040
    },
    {
      "epoch": 18.36740041928721,
      "grad_norm": 0.31754177808761597,
      "learning_rate": 1.0086966604097026e-06,
      "loss": 0.0046,
      "num_input_tokens_seen": 22914936,
      "step": 35045
    },
    {
      "epoch": 18.370020964360588,
      "grad_norm": 0.163587287068367,
      "learning_rate": 1.0054839639782766e-06,
      "loss": 0.0059,
      "num_input_tokens_seen": 22918424,
      "step": 35050
    },
    {
      "epoch": 18.37264150943396,
      "grad_norm": 0.23683707416057587,
      "learning_rate": 1.00227628692586e-06,
      "loss": 0.0048,
      "num_input_tokens_seen": 22921112,
      "step": 35055
    },
    {
      "epoch": 18.37526205450734,
      "grad_norm": 0.24610672891139984,
      "learning_rate": 9.990736299234692e-07,
      "loss": 0.0298,
      "num_input_tokens_seen": 22924024,
      "step": 35060
    },
    {
      "epoch": 18.377882599580712,
      "grad_norm": 0.9757353067398071,
      "learning_rate": 9.958759936410573e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 22927128,
      "step": 35065
    },
    {
      "epoch": 18.38050314465409,
      "grad_norm": 0.18245987594127655,
      "learning_rate": 9.926833787475304e-07,
      "loss": 0.0024,
      "num_input_tokens_seen": 22929912,
      "step": 35070
    },
    {
      "epoch": 18.383123689727462,
      "grad_norm": 0.7364073395729065,
      "learning_rate": 9.89495785910749e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 22932504,
      "step": 35075
    },
    {
      "epoch": 18.38574423480084,
      "grad_norm": 0.22007407248020172,
      "learning_rate": 9.86313215797524e-07,
      "loss": 0.0075,
      "num_input_tokens_seen": 22936600,
      "step": 35080
    },
    {
      "epoch": 18.388364779874212,
      "grad_norm": 1.4866063594818115,
      "learning_rate": 9.831356690736149e-07,
      "loss": 0.0199,
      "num_input_tokens_seen": 22939864,
      "step": 35085
    },
    {
      "epoch": 18.39098532494759,
      "grad_norm": 0.5283789038658142,
      "learning_rate": 9.799631464037256e-07,
      "loss": 0.0309,
      "num_input_tokens_seen": 22943480,
      "step": 35090
    },
    {
      "epoch": 18.393605870020963,
      "grad_norm": 0.5866502523422241,
      "learning_rate": 9.76795648451509e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 22946712,
      "step": 35095
    },
    {
      "epoch": 18.39622641509434,
      "grad_norm": 0.7649863958358765,
      "learning_rate": 9.73633175879568e-07,
      "loss": 0.0056,
      "num_input_tokens_seen": 22950392,
      "step": 35100
    },
    {
      "epoch": 18.398846960167713,
      "grad_norm": 1.7898738384246826,
      "learning_rate": 9.704757293494627e-07,
      "loss": 0.0177,
      "num_input_tokens_seen": 22954264,
      "step": 35105
    },
    {
      "epoch": 18.40146750524109,
      "grad_norm": 0.1280694603919983,
      "learning_rate": 9.673233095216922e-07,
      "loss": 0.0057,
      "num_input_tokens_seen": 22958232,
      "step": 35110
    },
    {
      "epoch": 18.404088050314467,
      "grad_norm": 0.13247761130332947,
      "learning_rate": 9.641759170557068e-07,
      "loss": 0.0181,
      "num_input_tokens_seen": 22961208,
      "step": 35115
    },
    {
      "epoch": 18.40670859538784,
      "grad_norm": 0.49860620498657227,
      "learning_rate": 9.610335526099024e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 22966296,
      "step": 35120
    },
    {
      "epoch": 18.409329140461217,
      "grad_norm": 0.21286845207214355,
      "learning_rate": 9.578962168416278e-07,
      "loss": 0.0012,
      "num_input_tokens_seen": 22968856,
      "step": 35125
    },
    {
      "epoch": 18.41194968553459,
      "grad_norm": 0.11239713430404663,
      "learning_rate": 9.547639104071775e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 22972056,
      "step": 35130
    },
    {
      "epoch": 18.414570230607968,
      "grad_norm": 0.08666863292455673,
      "learning_rate": 9.516366339617944e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 22974840,
      "step": 35135
    },
    {
      "epoch": 18.41719077568134,
      "grad_norm": 1.6774647235870361,
      "learning_rate": 9.485143881596714e-07,
      "loss": 0.0454,
      "num_input_tokens_seen": 22978520,
      "step": 35140
    },
    {
      "epoch": 18.419811320754718,
      "grad_norm": 0.5429823398590088,
      "learning_rate": 9.453971736539447e-07,
      "loss": 0.0085,
      "num_input_tokens_seen": 22982136,
      "step": 35145
    },
    {
      "epoch": 18.42243186582809,
      "grad_norm": 0.2790464460849762,
      "learning_rate": 9.42284991096698e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 22985304,
      "step": 35150
    },
    {
      "epoch": 18.42505241090147,
      "grad_norm": 0.9026950597763062,
      "learning_rate": 9.391778411389634e-07,
      "loss": 0.0059,
      "num_input_tokens_seen": 22988856,
      "step": 35155
    },
    {
      "epoch": 18.427672955974842,
      "grad_norm": 0.2530357241630554,
      "learning_rate": 9.360757244307267e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 22991896,
      "step": 35160
    },
    {
      "epoch": 18.43029350104822,
      "grad_norm": 1.1468243598937988,
      "learning_rate": 9.329786416209157e-07,
      "loss": 0.0141,
      "num_input_tokens_seen": 22994808,
      "step": 35165
    },
    {
      "epoch": 18.432914046121592,
      "grad_norm": 0.12294899672269821,
      "learning_rate": 9.298865933574014e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 22997560,
      "step": 35170
    },
    {
      "epoch": 18.43553459119497,
      "grad_norm": 0.24093420803546906,
      "learning_rate": 9.267995802870022e-07,
      "loss": 0.0421,
      "num_input_tokens_seen": 23000696,
      "step": 35175
    },
    {
      "epoch": 18.438155136268342,
      "grad_norm": 0.3548932373523712,
      "learning_rate": 9.237176030554878e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23003320,
      "step": 35180
    },
    {
      "epoch": 18.44077568134172,
      "grad_norm": 1.127220630645752,
      "learning_rate": 9.206406623075758e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23006456,
      "step": 35185
    },
    {
      "epoch": 18.443396226415093,
      "grad_norm": 0.44905349612236023,
      "learning_rate": 9.175687586869292e-07,
      "loss": 0.0047,
      "num_input_tokens_seen": 23010712,
      "step": 35190
    },
    {
      "epoch": 18.44601677148847,
      "grad_norm": 0.5538759231567383,
      "learning_rate": 9.145018928361476e-07,
      "loss": 0.0061,
      "num_input_tokens_seen": 23014232,
      "step": 35195
    },
    {
      "epoch": 18.448637316561843,
      "grad_norm": 1.1857805252075195,
      "learning_rate": 9.114400653967903e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23016952,
      "step": 35200
    },
    {
      "epoch": 18.45125786163522,
      "grad_norm": 0.37431347370147705,
      "learning_rate": 9.083832770093531e-07,
      "loss": 0.0318,
      "num_input_tokens_seen": 23019832,
      "step": 35205
    },
    {
      "epoch": 18.453878406708597,
      "grad_norm": 0.32276681065559387,
      "learning_rate": 9.053315283132802e-07,
      "loss": 0.0041,
      "num_input_tokens_seen": 23022264,
      "step": 35210
    },
    {
      "epoch": 18.45649895178197,
      "grad_norm": 0.6864737868309021,
      "learning_rate": 9.022848199469663e-07,
      "loss": 0.0063,
      "num_input_tokens_seen": 23025912,
      "step": 35215
    },
    {
      "epoch": 18.459119496855347,
      "grad_norm": 0.5980037450790405,
      "learning_rate": 8.992431525477518e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23029560,
      "step": 35220
    },
    {
      "epoch": 18.46174004192872,
      "grad_norm": 0.392734169960022,
      "learning_rate": 8.962065267519082e-07,
      "loss": 0.0075,
      "num_input_tokens_seen": 23033176,
      "step": 35225
    },
    {
      "epoch": 18.464360587002098,
      "grad_norm": 0.09552164375782013,
      "learning_rate": 8.931749431946718e-07,
      "loss": 0.0034,
      "num_input_tokens_seen": 23036568,
      "step": 35230
    },
    {
      "epoch": 18.46698113207547,
      "grad_norm": 0.32766011357307434,
      "learning_rate": 8.901484025102103e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 23039448,
      "step": 35235
    },
    {
      "epoch": 18.469601677148848,
      "grad_norm": 0.15843355655670166,
      "learning_rate": 8.871269053316422e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 23043384,
      "step": 35240
    },
    {
      "epoch": 18.47222222222222,
      "grad_norm": 0.11231224238872528,
      "learning_rate": 8.841104522910343e-07,
      "loss": 0.0472,
      "num_input_tokens_seen": 23047128,
      "step": 35245
    },
    {
      "epoch": 18.4748427672956,
      "grad_norm": 0.4144173264503479,
      "learning_rate": 8.8109904401939e-07,
      "loss": 0.0075,
      "num_input_tokens_seen": 23050072,
      "step": 35250
    },
    {
      "epoch": 18.47746331236897,
      "grad_norm": 1.0654329061508179,
      "learning_rate": 8.780926811466639e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23053688,
      "step": 35255
    },
    {
      "epoch": 18.48008385744235,
      "grad_norm": 0.34497129917144775,
      "learning_rate": 8.7509136430175e-07,
      "loss": 0.003,
      "num_input_tokens_seen": 23056440,
      "step": 35260
    },
    {
      "epoch": 18.482704402515722,
      "grad_norm": 0.3353269398212433,
      "learning_rate": 8.720950941124877e-07,
      "loss": 0.0064,
      "num_input_tokens_seen": 23060120,
      "step": 35265
    },
    {
      "epoch": 18.4853249475891,
      "grad_norm": 0.41008540987968445,
      "learning_rate": 8.691038712056732e-07,
      "loss": 0.0079,
      "num_input_tokens_seen": 23063544,
      "step": 35270
    },
    {
      "epoch": 18.487945492662472,
      "grad_norm": 0.11295972764492035,
      "learning_rate": 8.66117696207025e-07,
      "loss": 0.0031,
      "num_input_tokens_seen": 23067096,
      "step": 35275
    },
    {
      "epoch": 18.49056603773585,
      "grad_norm": 0.7101256847381592,
      "learning_rate": 8.631365697412186e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23070328,
      "step": 35280
    },
    {
      "epoch": 18.493186582809223,
      "grad_norm": 0.3585268557071686,
      "learning_rate": 8.601604924318802e-07,
      "loss": 0.0155,
      "num_input_tokens_seen": 23074648,
      "step": 35285
    },
    {
      "epoch": 18.4958071278826,
      "grad_norm": 0.7453505992889404,
      "learning_rate": 8.571894649015588e-07,
      "loss": 0.0037,
      "num_input_tokens_seen": 23077400,
      "step": 35290
    },
    {
      "epoch": 18.498427672955973,
      "grad_norm": 1.286337971687317,
      "learning_rate": 8.542234877717686e-07,
      "loss": 0.0108,
      "num_input_tokens_seen": 23079992,
      "step": 35295
    },
    {
      "epoch": 18.50104821802935,
      "grad_norm": 1.032251238822937,
      "learning_rate": 8.512625616629549e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 23082936,
      "step": 35300
    },
    {
      "epoch": 18.503668763102727,
      "grad_norm": 0.23518796265125275,
      "learning_rate": 8.483066871945083e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23085528,
      "step": 35305
    },
    {
      "epoch": 18.5062893081761,
      "grad_norm": 0.16007865965366364,
      "learning_rate": 8.453558649847676e-07,
      "loss": 0.0076,
      "num_input_tokens_seen": 23089016,
      "step": 35310
    },
    {
      "epoch": 18.508909853249477,
      "grad_norm": 0.10596518963575363,
      "learning_rate": 8.424100956510112e-07,
      "loss": 0.0074,
      "num_input_tokens_seen": 23091512,
      "step": 35315
    },
    {
      "epoch": 18.51153039832285,
      "grad_norm": 0.5738604664802551,
      "learning_rate": 8.394693798094544e-07,
      "loss": 0.0072,
      "num_input_tokens_seen": 23095096,
      "step": 35320
    },
    {
      "epoch": 18.514150943396228,
      "grad_norm": 0.21990089118480682,
      "learning_rate": 8.365337180752691e-07,
      "loss": 0.0047,
      "num_input_tokens_seen": 23098904,
      "step": 35325
    },
    {
      "epoch": 18.5167714884696,
      "grad_norm": 1.5056325197219849,
      "learning_rate": 8.336031110625558e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 23101944,
      "step": 35330
    },
    {
      "epoch": 18.519392033542978,
      "grad_norm": 0.30989518761634827,
      "learning_rate": 8.306775593843657e-07,
      "loss": 0.0061,
      "num_input_tokens_seen": 23105816,
      "step": 35335
    },
    {
      "epoch": 18.52201257861635,
      "grad_norm": 0.4884800612926483,
      "learning_rate": 8.277570636526955e-07,
      "loss": 0.0101,
      "num_input_tokens_seen": 23109304,
      "step": 35340
    },
    {
      "epoch": 18.52463312368973,
      "grad_norm": 0.41925522685050964,
      "learning_rate": 8.24841624478473e-07,
      "loss": 0.0062,
      "num_input_tokens_seen": 23112984,
      "step": 35345
    },
    {
      "epoch": 18.5272536687631,
      "grad_norm": 0.9829399585723877,
      "learning_rate": 8.219312424715825e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 23115832,
      "step": 35350
    },
    {
      "epoch": 18.52987421383648,
      "grad_norm": 0.3132520020008087,
      "learning_rate": 8.190259182408345e-07,
      "loss": 0.0031,
      "num_input_tokens_seen": 23118872,
      "step": 35355
    },
    {
      "epoch": 18.532494758909852,
      "grad_norm": 0.2316303253173828,
      "learning_rate": 8.161256523939925e-07,
      "loss": 0.0097,
      "num_input_tokens_seen": 23121880,
      "step": 35360
    },
    {
      "epoch": 18.53511530398323,
      "grad_norm": 0.039980221539735794,
      "learning_rate": 8.132304455377576e-07,
      "loss": 0.0112,
      "num_input_tokens_seen": 23125528,
      "step": 35365
    },
    {
      "epoch": 18.537735849056602,
      "grad_norm": 0.6288179755210876,
      "learning_rate": 8.103402982777813e-07,
      "loss": 0.0256,
      "num_input_tokens_seen": 23130360,
      "step": 35370
    },
    {
      "epoch": 18.54035639412998,
      "grad_norm": 0.1143823191523552,
      "learning_rate": 8.07455211218644e-07,
      "loss": 0.0091,
      "num_input_tokens_seen": 23133912,
      "step": 35375
    },
    {
      "epoch": 18.542976939203353,
      "grad_norm": 0.2120276242494583,
      "learning_rate": 8.045751849638738e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 23136440,
      "step": 35380
    },
    {
      "epoch": 18.54559748427673,
      "grad_norm": 1.052377700805664,
      "learning_rate": 8.017002201159363e-07,
      "loss": 0.0077,
      "num_input_tokens_seen": 23139544,
      "step": 35385
    },
    {
      "epoch": 18.548218029350103,
      "grad_norm": 0.509529709815979,
      "learning_rate": 7.988303172762418e-07,
      "loss": 0.0134,
      "num_input_tokens_seen": 23144408,
      "step": 35390
    },
    {
      "epoch": 18.55083857442348,
      "grad_norm": 1.0218267440795898,
      "learning_rate": 7.959654770451463e-07,
      "loss": 0.004,
      "num_input_tokens_seen": 23147960,
      "step": 35395
    },
    {
      "epoch": 18.553459119496857,
      "grad_norm": 0.19067193567752838,
      "learning_rate": 7.9310570002194e-07,
      "loss": 0.004,
      "num_input_tokens_seen": 23151096,
      "step": 35400
    },
    {
      "epoch": 18.55607966457023,
      "grad_norm": 0.2231169193983078,
      "learning_rate": 7.902509868048552e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23154424,
      "step": 35405
    },
    {
      "epoch": 18.558700209643607,
      "grad_norm": 0.3439536392688751,
      "learning_rate": 7.874013379910616e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23157048,
      "step": 35410
    },
    {
      "epoch": 18.56132075471698,
      "grad_norm": 0.4360572397708893,
      "learning_rate": 7.845567541766769e-07,
      "loss": 0.0078,
      "num_input_tokens_seen": 23160792,
      "step": 35415
    },
    {
      "epoch": 18.563941299790358,
      "grad_norm": 0.3118366003036499,
      "learning_rate": 7.817172359567526e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23164248,
      "step": 35420
    },
    {
      "epoch": 18.56656184486373,
      "grad_norm": 0.5751054883003235,
      "learning_rate": 7.788827839252888e-07,
      "loss": 0.0187,
      "num_input_tokens_seen": 23167832,
      "step": 35425
    },
    {
      "epoch": 18.569182389937108,
      "grad_norm": 0.2612093985080719,
      "learning_rate": 7.760533986752222e-07,
      "loss": 0.0025,
      "num_input_tokens_seen": 23172152,
      "step": 35430
    },
    {
      "epoch": 18.57180293501048,
      "grad_norm": 0.3876231610774994,
      "learning_rate": 7.732290807984155e-07,
      "loss": 0.0025,
      "num_input_tokens_seen": 23175000,
      "step": 35435
    },
    {
      "epoch": 18.57442348008386,
      "grad_norm": 0.40857070684432983,
      "learning_rate": 7.704098308856933e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23177976,
      "step": 35440
    },
    {
      "epoch": 18.57704402515723,
      "grad_norm": 0.2938285768032074,
      "learning_rate": 7.675956495268088e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23180792,
      "step": 35445
    },
    {
      "epoch": 18.57966457023061,
      "grad_norm": 1.1351532936096191,
      "learning_rate": 7.647865373104551e-07,
      "loss": 0.0093,
      "num_input_tokens_seen": 23184344,
      "step": 35450
    },
    {
      "epoch": 18.582285115303982,
      "grad_norm": 0.7115507125854492,
      "learning_rate": 7.619824948242732e-07,
      "loss": 0.003,
      "num_input_tokens_seen": 23187640,
      "step": 35455
    },
    {
      "epoch": 18.58490566037736,
      "grad_norm": 0.3817996084690094,
      "learning_rate": 7.5918352265483e-07,
      "loss": 0.0168,
      "num_input_tokens_seen": 23191256,
      "step": 35460
    },
    {
      "epoch": 18.587526205450732,
      "grad_norm": 0.15198493003845215,
      "learning_rate": 7.563896213876348e-07,
      "loss": 0.0058,
      "num_input_tokens_seen": 23193944,
      "step": 35465
    },
    {
      "epoch": 18.59014675052411,
      "grad_norm": 0.7207276225090027,
      "learning_rate": 7.53600791607148e-07,
      "loss": 0.0054,
      "num_input_tokens_seen": 23197464,
      "step": 35470
    },
    {
      "epoch": 18.592767295597483,
      "grad_norm": 0.23945282399654388,
      "learning_rate": 7.508170338967585e-07,
      "loss": 0.0068,
      "num_input_tokens_seen": 23200024,
      "step": 35475
    },
    {
      "epoch": 18.59538784067086,
      "grad_norm": 0.5155551433563232,
      "learning_rate": 7.480383488387976e-07,
      "loss": 0.0031,
      "num_input_tokens_seen": 23202968,
      "step": 35480
    },
    {
      "epoch": 18.598008385744233,
      "grad_norm": 0.29020193219184875,
      "learning_rate": 7.452647370145338e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 23206296,
      "step": 35485
    },
    {
      "epoch": 18.60062893081761,
      "grad_norm": 0.4896828234195709,
      "learning_rate": 7.424961990041696e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 23209304,
      "step": 35490
    },
    {
      "epoch": 18.603249475890987,
      "grad_norm": 0.1685401052236557,
      "learning_rate": 7.397327353868583e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23213080,
      "step": 35495
    },
    {
      "epoch": 18.60587002096436,
      "grad_norm": 0.21604041755199432,
      "learning_rate": 7.369743467406848e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23215928,
      "step": 35500
    },
    {
      "epoch": 18.608490566037737,
      "grad_norm": 0.41679567098617554,
      "learning_rate": 7.34221033642668e-07,
      "loss": 0.0049,
      "num_input_tokens_seen": 23219320,
      "step": 35505
    },
    {
      "epoch": 18.61111111111111,
      "grad_norm": 0.20631584525108337,
      "learning_rate": 7.314727966687751e-07,
      "loss": 0.0041,
      "num_input_tokens_seen": 23222296,
      "step": 35510
    },
    {
      "epoch": 18.613731656184488,
      "grad_norm": 0.2856248617172241,
      "learning_rate": 7.287296363939045e-07,
      "loss": 0.0086,
      "num_input_tokens_seen": 23224728,
      "step": 35515
    },
    {
      "epoch": 18.61635220125786,
      "grad_norm": 0.09770609438419342,
      "learning_rate": 7.259915533918888e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23228248,
      "step": 35520
    },
    {
      "epoch": 18.618972746331238,
      "grad_norm": 0.03778274357318878,
      "learning_rate": 7.232585482355086e-07,
      "loss": 0.0065,
      "num_input_tokens_seen": 23232600,
      "step": 35525
    },
    {
      "epoch": 18.62159329140461,
      "grad_norm": 0.6730189919471741,
      "learning_rate": 7.205306214964736e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 23235384,
      "step": 35530
    },
    {
      "epoch": 18.62421383647799,
      "grad_norm": 0.5322210788726807,
      "learning_rate": 7.178077737454436e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23238744,
      "step": 35535
    },
    {
      "epoch": 18.62683438155136,
      "grad_norm": 0.20017723739147186,
      "learning_rate": 7.150900055519993e-07,
      "loss": 0.0153,
      "num_input_tokens_seen": 23241720,
      "step": 35540
    },
    {
      "epoch": 18.62945492662474,
      "grad_norm": 0.7665067315101624,
      "learning_rate": 7.123773174846693e-07,
      "loss": 0.015,
      "num_input_tokens_seen": 23244696,
      "step": 35545
    },
    {
      "epoch": 18.632075471698112,
      "grad_norm": 0.3420517146587372,
      "learning_rate": 7.096697101109134e-07,
      "loss": 0.0062,
      "num_input_tokens_seen": 23248536,
      "step": 35550
    },
    {
      "epoch": 18.63469601677149,
      "grad_norm": 0.4619334936141968,
      "learning_rate": 7.069671839971371e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 23251352,
      "step": 35555
    },
    {
      "epoch": 18.637316561844862,
      "grad_norm": 1.401728868484497,
      "learning_rate": 7.042697397086795e-07,
      "loss": 0.0109,
      "num_input_tokens_seen": 23254520,
      "step": 35560
    },
    {
      "epoch": 18.63993710691824,
      "grad_norm": 0.22745710611343384,
      "learning_rate": 7.015773778098089e-07,
      "loss": 0.0234,
      "num_input_tokens_seen": 23261752,
      "step": 35565
    },
    {
      "epoch": 18.642557651991613,
      "grad_norm": 0.4394373595714569,
      "learning_rate": 6.988900988637415e-07,
      "loss": 0.0139,
      "num_input_tokens_seen": 23265176,
      "step": 35570
    },
    {
      "epoch": 18.64517819706499,
      "grad_norm": 0.4474996328353882,
      "learning_rate": 6.962079034326218e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 23267800,
      "step": 35575
    },
    {
      "epoch": 18.647798742138363,
      "grad_norm": 0.11471682041883469,
      "learning_rate": 6.935307920775346e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23270488,
      "step": 35580
    },
    {
      "epoch": 18.65041928721174,
      "grad_norm": 1.505387306213379,
      "learning_rate": 6.90858765358507e-07,
      "loss": 0.0099,
      "num_input_tokens_seen": 23274008,
      "step": 35585
    },
    {
      "epoch": 18.653039832285117,
      "grad_norm": 0.07625169306993484,
      "learning_rate": 6.881918238344915e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 23277240,
      "step": 35590
    },
    {
      "epoch": 18.65566037735849,
      "grad_norm": 0.28724396228790283,
      "learning_rate": 6.855299680633808e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 23279768,
      "step": 35595
    },
    {
      "epoch": 18.658280922431867,
      "grad_norm": 0.07611913233995438,
      "learning_rate": 6.828731986020104e-07,
      "loss": 0.0111,
      "num_input_tokens_seen": 23283128,
      "step": 35600
    },
    {
      "epoch": 18.66090146750524,
      "grad_norm": 0.28834420442581177,
      "learning_rate": 6.802215160061381e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23286424,
      "step": 35605
    },
    {
      "epoch": 18.663522012578618,
      "grad_norm": 0.6295123100280762,
      "learning_rate": 6.775749208304704e-07,
      "loss": 0.0139,
      "num_input_tokens_seen": 23289880,
      "step": 35610
    },
    {
      "epoch": 18.66614255765199,
      "grad_norm": 0.9818454384803772,
      "learning_rate": 6.749334136286478e-07,
      "loss": 0.0075,
      "num_input_tokens_seen": 23293528,
      "step": 35615
    },
    {
      "epoch": 18.668763102725368,
      "grad_norm": 0.5558608174324036,
      "learning_rate": 6.722969949532365e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23296952,
      "step": 35620
    },
    {
      "epoch": 18.67138364779874,
      "grad_norm": 0.4166063368320465,
      "learning_rate": 6.69665665355751e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23299384,
      "step": 35625
    },
    {
      "epoch": 18.67400419287212,
      "grad_norm": 0.8206596970558167,
      "learning_rate": 6.670394253866314e-07,
      "loss": 0.0057,
      "num_input_tokens_seen": 23306232,
      "step": 35630
    },
    {
      "epoch": 18.67662473794549,
      "grad_norm": 0.17619942128658295,
      "learning_rate": 6.644182755952605e-07,
      "loss": 0.0064,
      "num_input_tokens_seen": 23309144,
      "step": 35635
    },
    {
      "epoch": 18.67924528301887,
      "grad_norm": 0.4376901686191559,
      "learning_rate": 6.618022165299526e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23312472,
      "step": 35640
    },
    {
      "epoch": 18.681865828092242,
      "grad_norm": 0.19329971075057983,
      "learning_rate": 6.591912487379531e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23315320,
      "step": 35645
    },
    {
      "epoch": 18.68448637316562,
      "grad_norm": 0.05743539333343506,
      "learning_rate": 6.565853727654503e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 23318360,
      "step": 35650
    },
    {
      "epoch": 18.687106918238992,
      "grad_norm": 0.8150842189788818,
      "learning_rate": 6.539845891575636e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23320824,
      "step": 35655
    },
    {
      "epoch": 18.68972746331237,
      "grad_norm": 1.0764374732971191,
      "learning_rate": 6.513888984583466e-07,
      "loss": 0.0047,
      "num_input_tokens_seen": 23324216,
      "step": 35660
    },
    {
      "epoch": 18.692348008385743,
      "grad_norm": 0.24453520774841309,
      "learning_rate": 6.487983012107873e-07,
      "loss": 0.0153,
      "num_input_tokens_seen": 23326840,
      "step": 35665
    },
    {
      "epoch": 18.69496855345912,
      "grad_norm": 0.1364022195339203,
      "learning_rate": 6.462127979568133e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23329464,
      "step": 35670
    },
    {
      "epoch": 18.697589098532493,
      "grad_norm": 0.570811927318573,
      "learning_rate": 6.436323892372781e-07,
      "loss": 0.0042,
      "num_input_tokens_seen": 23332280,
      "step": 35675
    },
    {
      "epoch": 18.70020964360587,
      "grad_norm": 0.3795738220214844,
      "learning_rate": 6.410570755919721e-07,
      "loss": 0.0079,
      "num_input_tokens_seen": 23335384,
      "step": 35680
    },
    {
      "epoch": 18.702830188679247,
      "grad_norm": 0.17958052456378937,
      "learning_rate": 6.384868575596281e-07,
      "loss": 0.0025,
      "num_input_tokens_seen": 23338776,
      "step": 35685
    },
    {
      "epoch": 18.70545073375262,
      "grad_norm": 0.15999187529087067,
      "learning_rate": 6.359217356779024e-07,
      "loss": 0.0128,
      "num_input_tokens_seen": 23341624,
      "step": 35690
    },
    {
      "epoch": 18.708071278825997,
      "grad_norm": 0.3536481261253357,
      "learning_rate": 6.333617104833878e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 23345144,
      "step": 35695
    },
    {
      "epoch": 18.71069182389937,
      "grad_norm": 0.45281538367271423,
      "learning_rate": 6.308067825116142e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 23350328,
      "step": 35700
    },
    {
      "epoch": 18.713312368972748,
      "grad_norm": 0.10404880344867706,
      "learning_rate": 6.282569522970433e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23353240,
      "step": 35705
    },
    {
      "epoch": 18.71593291404612,
      "grad_norm": 0.8834526538848877,
      "learning_rate": 6.257122203730731e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 23356120,
      "step": 35710
    },
    {
      "epoch": 18.718553459119498,
      "grad_norm": 0.6283930540084839,
      "learning_rate": 6.231725872720251e-07,
      "loss": 0.0173,
      "num_input_tokens_seen": 23359032,
      "step": 35715
    },
    {
      "epoch": 18.72117400419287,
      "grad_norm": 0.3102629780769348,
      "learning_rate": 6.206380535251716e-07,
      "loss": 0.0056,
      "num_input_tokens_seen": 23363064,
      "step": 35720
    },
    {
      "epoch": 18.72379454926625,
      "grad_norm": 0.5515471696853638,
      "learning_rate": 6.181086196626995e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23365432,
      "step": 35725
    },
    {
      "epoch": 18.72641509433962,
      "grad_norm": 0.20850113034248352,
      "learning_rate": 6.155842862137384e-07,
      "loss": 0.0092,
      "num_input_tokens_seen": 23368120,
      "step": 35730
    },
    {
      "epoch": 18.729035639413,
      "grad_norm": 2.018277406692505,
      "learning_rate": 6.130650537063548e-07,
      "loss": 0.0079,
      "num_input_tokens_seen": 23371576,
      "step": 35735
    },
    {
      "epoch": 18.731656184486372,
      "grad_norm": 0.08969906717538834,
      "learning_rate": 6.105509226675438e-07,
      "loss": 0.0047,
      "num_input_tokens_seen": 23376216,
      "step": 35740
    },
    {
      "epoch": 18.73427672955975,
      "grad_norm": 0.1853961944580078,
      "learning_rate": 6.08041893623229e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 23379256,
      "step": 35745
    },
    {
      "epoch": 18.736897274633122,
      "grad_norm": 2.2855722904205322,
      "learning_rate": 6.055379670982713e-07,
      "loss": 0.0121,
      "num_input_tokens_seen": 23381912,
      "step": 35750
    },
    {
      "epoch": 18.7395178197065,
      "grad_norm": 0.3487964868545532,
      "learning_rate": 6.030391436164628e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 23384568,
      "step": 35755
    },
    {
      "epoch": 18.742138364779873,
      "grad_norm": 0.1591724157333374,
      "learning_rate": 6.005454237005298e-07,
      "loss": 0.0081,
      "num_input_tokens_seen": 23387320,
      "step": 35760
    },
    {
      "epoch": 18.74475890985325,
      "grad_norm": 0.36002326011657715,
      "learning_rate": 5.980568078721327e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 23390168,
      "step": 35765
    },
    {
      "epoch": 18.747379454926623,
      "grad_norm": 0.19259583950042725,
      "learning_rate": 5.955732966518606e-07,
      "loss": 0.0063,
      "num_input_tokens_seen": 23393080,
      "step": 35770
    },
    {
      "epoch": 18.75,
      "grad_norm": 0.33957478404045105,
      "learning_rate": 5.930948905592343e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23395800,
      "step": 35775
    },
    {
      "epoch": 18.752620545073377,
      "grad_norm": 0.44290614128112793,
      "learning_rate": 5.906215901127054e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 23398424,
      "step": 35780
    },
    {
      "epoch": 18.75524109014675,
      "grad_norm": 3.0107572078704834,
      "learning_rate": 5.881533958296631e-07,
      "loss": 0.0078,
      "num_input_tokens_seen": 23400984,
      "step": 35785
    },
    {
      "epoch": 18.757861635220127,
      "grad_norm": 1.265383243560791,
      "learning_rate": 5.856903082264248e-07,
      "loss": 0.0086,
      "num_input_tokens_seen": 23404952,
      "step": 35790
    },
    {
      "epoch": 18.7604821802935,
      "grad_norm": 0.21100136637687683,
      "learning_rate": 5.832323278182422e-07,
      "loss": 0.0081,
      "num_input_tokens_seen": 23409048,
      "step": 35795
    },
    {
      "epoch": 18.763102725366878,
      "grad_norm": 0.11223085969686508,
      "learning_rate": 5.807794551192986e-07,
      "loss": 0.0047,
      "num_input_tokens_seen": 23411864,
      "step": 35800
    },
    {
      "epoch": 18.76572327044025,
      "grad_norm": 0.4521251320838928,
      "learning_rate": 5.783316906427e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 23414936,
      "step": 35805
    },
    {
      "epoch": 18.768343815513628,
      "grad_norm": 0.8126404881477356,
      "learning_rate": 5.758890349004925e-07,
      "loss": 0.0082,
      "num_input_tokens_seen": 23417976,
      "step": 35810
    },
    {
      "epoch": 18.770964360587,
      "grad_norm": 0.17843510210514069,
      "learning_rate": 5.734514884036563e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23420984,
      "step": 35815
    },
    {
      "epoch": 18.77358490566038,
      "grad_norm": 0.26752769947052,
      "learning_rate": 5.710190516620945e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 23424120,
      "step": 35820
    },
    {
      "epoch": 18.77620545073375,
      "grad_norm": 1.1265877485275269,
      "learning_rate": 5.685917251846501e-07,
      "loss": 0.0296,
      "num_input_tokens_seen": 23426584,
      "step": 35825
    },
    {
      "epoch": 18.77882599580713,
      "grad_norm": 0.1473032683134079,
      "learning_rate": 5.661695094790892e-07,
      "loss": 0.003,
      "num_input_tokens_seen": 23429880,
      "step": 35830
    },
    {
      "epoch": 18.781446540880502,
      "grad_norm": 0.15023162961006165,
      "learning_rate": 5.637524050521064e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 23432728,
      "step": 35835
    },
    {
      "epoch": 18.78406708595388,
      "grad_norm": 0.3099850118160248,
      "learning_rate": 5.613404124093391e-07,
      "loss": 0.0182,
      "num_input_tokens_seen": 23435704,
      "step": 35840
    },
    {
      "epoch": 18.786687631027252,
      "grad_norm": 0.081308513879776,
      "learning_rate": 5.589335320553446e-07,
      "loss": 0.0042,
      "num_input_tokens_seen": 23439416,
      "step": 35845
    },
    {
      "epoch": 18.78930817610063,
      "grad_norm": 0.43441158533096313,
      "learning_rate": 5.565317644936202e-07,
      "loss": 0.0086,
      "num_input_tokens_seen": 23441880,
      "step": 35850
    },
    {
      "epoch": 18.791928721174003,
      "grad_norm": 0.46753638982772827,
      "learning_rate": 5.541351102265863e-07,
      "loss": 0.0067,
      "num_input_tokens_seen": 23445560,
      "step": 35855
    },
    {
      "epoch": 18.79454926624738,
      "grad_norm": 0.13540108501911163,
      "learning_rate": 5.51743569755589e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 23448760,
      "step": 35860
    },
    {
      "epoch": 18.797169811320753,
      "grad_norm": 0.13844454288482666,
      "learning_rate": 5.493571435809197e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 23451864,
      "step": 35865
    },
    {
      "epoch": 18.79979035639413,
      "grad_norm": 1.4495140314102173,
      "learning_rate": 5.469758322017849e-07,
      "loss": 0.0283,
      "num_input_tokens_seen": 23458552,
      "step": 35870
    },
    {
      "epoch": 18.802410901467507,
      "grad_norm": 1.42551589012146,
      "learning_rate": 5.445996361163358e-07,
      "loss": 0.0077,
      "num_input_tokens_seen": 23461656,
      "step": 35875
    },
    {
      "epoch": 18.80503144654088,
      "grad_norm": 0.393660306930542,
      "learning_rate": 5.422285558216389e-07,
      "loss": 0.0279,
      "num_input_tokens_seen": 23464664,
      "step": 35880
    },
    {
      "epoch": 18.807651991614257,
      "grad_norm": 0.20386822521686554,
      "learning_rate": 5.398625918137001e-07,
      "loss": 0.0076,
      "num_input_tokens_seen": 23468216,
      "step": 35885
    },
    {
      "epoch": 18.81027253668763,
      "grad_norm": 0.1551416516304016,
      "learning_rate": 5.375017445874486e-07,
      "loss": 0.0066,
      "num_input_tokens_seen": 23471064,
      "step": 35890
    },
    {
      "epoch": 18.812893081761008,
      "grad_norm": 0.7259486317634583,
      "learning_rate": 5.351460146367504e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23474264,
      "step": 35895
    },
    {
      "epoch": 18.81551362683438,
      "grad_norm": 0.3224949836730957,
      "learning_rate": 5.327954024543946e-07,
      "loss": 0.0064,
      "num_input_tokens_seen": 23477720,
      "step": 35900
    },
    {
      "epoch": 18.818134171907758,
      "grad_norm": 0.31487396359443665,
      "learning_rate": 5.304499085321074e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23480344,
      "step": 35905
    },
    {
      "epoch": 18.82075471698113,
      "grad_norm": 0.25371086597442627,
      "learning_rate": 5.281095333605352e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23483544,
      "step": 35910
    },
    {
      "epoch": 18.82337526205451,
      "grad_norm": 0.41796615719795227,
      "learning_rate": 5.257742774292556e-07,
      "loss": 0.0031,
      "num_input_tokens_seen": 23486392,
      "step": 35915
    },
    {
      "epoch": 18.82599580712788,
      "grad_norm": 1.1255689859390259,
      "learning_rate": 5.234441412267837e-07,
      "loss": 0.0103,
      "num_input_tokens_seen": 23490200,
      "step": 35920
    },
    {
      "epoch": 18.82861635220126,
      "grad_norm": 0.28804194927215576,
      "learning_rate": 5.211191252405517e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 23493048,
      "step": 35925
    },
    {
      "epoch": 18.831236897274632,
      "grad_norm": 0.7802236080169678,
      "learning_rate": 5.187992299569316e-07,
      "loss": 0.0169,
      "num_input_tokens_seen": 23496440,
      "step": 35930
    },
    {
      "epoch": 18.83385744234801,
      "grad_norm": 0.12042004615068436,
      "learning_rate": 5.164844558612131e-07,
      "loss": 0.0053,
      "num_input_tokens_seen": 23499832,
      "step": 35935
    },
    {
      "epoch": 18.836477987421382,
      "grad_norm": 0.2779167592525482,
      "learning_rate": 5.141748034376254e-07,
      "loss": 0.0082,
      "num_input_tokens_seen": 23503416,
      "step": 35940
    },
    {
      "epoch": 18.83909853249476,
      "grad_norm": 0.32999658584594727,
      "learning_rate": 5.118702731693209e-07,
      "loss": 0.0034,
      "num_input_tokens_seen": 23506616,
      "step": 35945
    },
    {
      "epoch": 18.841719077568133,
      "grad_norm": 0.29631686210632324,
      "learning_rate": 5.095708655383751e-07,
      "loss": 0.0057,
      "num_input_tokens_seen": 23509912,
      "step": 35950
    },
    {
      "epoch": 18.84433962264151,
      "grad_norm": 0.8102484345436096,
      "learning_rate": 5.07276581025809e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23513272,
      "step": 35955
    },
    {
      "epoch": 18.846960167714883,
      "grad_norm": 0.5198948979377747,
      "learning_rate": 5.049874201115495e-07,
      "loss": 0.0055,
      "num_input_tokens_seen": 23515992,
      "step": 35960
    },
    {
      "epoch": 18.84958071278826,
      "grad_norm": 0.20415839552879333,
      "learning_rate": 5.027033832744693e-07,
      "loss": 0.0054,
      "num_input_tokens_seen": 23519032,
      "step": 35965
    },
    {
      "epoch": 18.852201257861637,
      "grad_norm": 0.2950422465801239,
      "learning_rate": 5.004244709923639e-07,
      "loss": 0.0046,
      "num_input_tokens_seen": 23522392,
      "step": 35970
    },
    {
      "epoch": 18.85482180293501,
      "grad_norm": 1.899869680404663,
      "learning_rate": 4.981506837419519e-07,
      "loss": 0.0046,
      "num_input_tokens_seen": 23525240,
      "step": 35975
    },
    {
      "epoch": 18.857442348008387,
      "grad_norm": 0.6420112252235413,
      "learning_rate": 4.958820219988836e-07,
      "loss": 0.0117,
      "num_input_tokens_seen": 23528568,
      "step": 35980
    },
    {
      "epoch": 18.86006289308176,
      "grad_norm": 0.4986006021499634,
      "learning_rate": 4.936184862377402e-07,
      "loss": 0.0065,
      "num_input_tokens_seen": 23531288,
      "step": 35985
    },
    {
      "epoch": 18.862683438155138,
      "grad_norm": 0.4077673852443695,
      "learning_rate": 4.913600769320237e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23534520,
      "step": 35990
    },
    {
      "epoch": 18.86530398322851,
      "grad_norm": 0.35713908076286316,
      "learning_rate": 4.891067945541728e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 23537880,
      "step": 35995
    },
    {
      "epoch": 18.867924528301888,
      "grad_norm": 0.19111528992652893,
      "learning_rate": 4.868586395755437e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23541464,
      "step": 36000
    },
    {
      "epoch": 18.87054507337526,
      "grad_norm": 0.44271495938301086,
      "learning_rate": 4.846156124664242e-07,
      "loss": 0.0067,
      "num_input_tokens_seen": 23544760,
      "step": 36005
    },
    {
      "epoch": 18.87316561844864,
      "grad_norm": 1.1795570850372314,
      "learning_rate": 4.823777136960362e-07,
      "loss": 0.0037,
      "num_input_tokens_seen": 23547832,
      "step": 36010
    },
    {
      "epoch": 18.87578616352201,
      "grad_norm": 0.29215705394744873,
      "learning_rate": 4.801449437325162e-07,
      "loss": 0.0013,
      "num_input_tokens_seen": 23550968,
      "step": 36015
    },
    {
      "epoch": 18.87840670859539,
      "grad_norm": 1.5703318119049072,
      "learning_rate": 4.779173030429379e-07,
      "loss": 0.0065,
      "num_input_tokens_seen": 23554680,
      "step": 36020
    },
    {
      "epoch": 18.881027253668762,
      "grad_norm": 0.48534029722213745,
      "learning_rate": 4.7569479209329804e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 23558168,
      "step": 36025
    },
    {
      "epoch": 18.88364779874214,
      "grad_norm": 0.3960895240306854,
      "learning_rate": 4.734774113485163e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 23560952,
      "step": 36030
    },
    {
      "epoch": 18.886268343815512,
      "grad_norm": 0.4136182367801666,
      "learning_rate": 4.7126516127244945e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 23563096,
      "step": 36035
    },
    {
      "epoch": 18.88888888888889,
      "grad_norm": 1.7822201251983643,
      "learning_rate": 4.690580423278718e-07,
      "loss": 0.0298,
      "num_input_tokens_seen": 23565400,
      "step": 36040
    },
    {
      "epoch": 18.891509433962263,
      "grad_norm": 0.7397239208221436,
      "learning_rate": 4.668560549764861e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 23569016,
      "step": 36045
    },
    {
      "epoch": 18.89412997903564,
      "grad_norm": 0.450708270072937,
      "learning_rate": 4.646591996789268e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23572216,
      "step": 36050
    },
    {
      "epoch": 18.896750524109013,
      "grad_norm": 0.26364096999168396,
      "learning_rate": 4.6246747689474847e-07,
      "loss": 0.0096,
      "num_input_tokens_seen": 23576056,
      "step": 36055
    },
    {
      "epoch": 18.89937106918239,
      "grad_norm": 0.407696396112442,
      "learning_rate": 4.6028088708243434e-07,
      "loss": 0.0147,
      "num_input_tokens_seen": 23579096,
      "step": 36060
    },
    {
      "epoch": 18.901991614255767,
      "grad_norm": 0.09530720114707947,
      "learning_rate": 4.580994306993991e-07,
      "loss": 0.0034,
      "num_input_tokens_seen": 23581720,
      "step": 36065
    },
    {
      "epoch": 18.90461215932914,
      "grad_norm": 1.3596447706222534,
      "learning_rate": 4.5592310820196936e-07,
      "loss": 0.0134,
      "num_input_tokens_seen": 23584824,
      "step": 36070
    },
    {
      "epoch": 18.907232704402517,
      "grad_norm": 0.2251741886138916,
      "learning_rate": 4.537519200454116e-07,
      "loss": 0.0024,
      "num_input_tokens_seen": 23587800,
      "step": 36075
    },
    {
      "epoch": 18.90985324947589,
      "grad_norm": 0.4439716935157776,
      "learning_rate": 4.515858666839151e-07,
      "loss": 0.0852,
      "num_input_tokens_seen": 23590584,
      "step": 36080
    },
    {
      "epoch": 18.912473794549268,
      "grad_norm": 2.2826435565948486,
      "learning_rate": 4.494249485705926e-07,
      "loss": 0.0629,
      "num_input_tokens_seen": 23594392,
      "step": 36085
    },
    {
      "epoch": 18.91509433962264,
      "grad_norm": 0.414969801902771,
      "learning_rate": 4.472691661574824e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23597176,
      "step": 36090
    },
    {
      "epoch": 18.917714884696018,
      "grad_norm": 0.4779299199581146,
      "learning_rate": 4.4511851989555156e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23600600,
      "step": 36095
    },
    {
      "epoch": 18.92033542976939,
      "grad_norm": 0.34718307852745056,
      "learning_rate": 4.429730102346874e-07,
      "loss": 0.0051,
      "num_input_tokens_seen": 23603480,
      "step": 36100
    },
    {
      "epoch": 18.92295597484277,
      "grad_norm": 0.3520158529281616,
      "learning_rate": 4.4083263762370873e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 23605784,
      "step": 36105
    },
    {
      "epoch": 18.92557651991614,
      "grad_norm": 0.4575716257095337,
      "learning_rate": 4.386974025103574e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 23609144,
      "step": 36110
    },
    {
      "epoch": 18.92819706498952,
      "grad_norm": 0.3999690115451813,
      "learning_rate": 4.365673053413011e-07,
      "loss": 0.0116,
      "num_input_tokens_seen": 23611960,
      "step": 36115
    },
    {
      "epoch": 18.930817610062892,
      "grad_norm": 0.18292111158370972,
      "learning_rate": 4.34442346562125e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 23614680,
      "step": 36120
    },
    {
      "epoch": 18.93343815513627,
      "grad_norm": 4.699762344360352,
      "learning_rate": 4.323225266173542e-07,
      "loss": 0.0098,
      "num_input_tokens_seen": 23617592,
      "step": 36125
    },
    {
      "epoch": 18.936058700209642,
      "grad_norm": 0.6277283430099487,
      "learning_rate": 4.302078459504283e-07,
      "loss": 0.0068,
      "num_input_tokens_seen": 23620440,
      "step": 36130
    },
    {
      "epoch": 18.93867924528302,
      "grad_norm": 0.07100146263837814,
      "learning_rate": 4.2809830500371294e-07,
      "loss": 0.0014,
      "num_input_tokens_seen": 23623320,
      "step": 36135
    },
    {
      "epoch": 18.941299790356393,
      "grad_norm": 0.05317072570323944,
      "learning_rate": 4.2599390421850236e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 23626168,
      "step": 36140
    },
    {
      "epoch": 18.94392033542977,
      "grad_norm": 3.851801633834839,
      "learning_rate": 4.2389464403501113e-07,
      "loss": 0.0202,
      "num_input_tokens_seen": 23629240,
      "step": 36145
    },
    {
      "epoch": 18.946540880503143,
      "grad_norm": 0.3243536055088043,
      "learning_rate": 4.218005248923823e-07,
      "loss": 0.0043,
      "num_input_tokens_seen": 23632472,
      "step": 36150
    },
    {
      "epoch": 18.94916142557652,
      "grad_norm": 0.7833017110824585,
      "learning_rate": 4.1971154722867946e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23635864,
      "step": 36155
    },
    {
      "epoch": 18.951781970649897,
      "grad_norm": 0.6005925536155701,
      "learning_rate": 4.176277114808946e-07,
      "loss": 0.006,
      "num_input_tokens_seen": 23638712,
      "step": 36160
    },
    {
      "epoch": 18.95440251572327,
      "grad_norm": 0.5512354969978333,
      "learning_rate": 4.155490180849431e-07,
      "loss": 0.004,
      "num_input_tokens_seen": 23641592,
      "step": 36165
    },
    {
      "epoch": 18.957023060796647,
      "grad_norm": 0.3388539254665375,
      "learning_rate": 4.1347546747566304e-07,
      "loss": 0.0081,
      "num_input_tokens_seen": 23645912,
      "step": 36170
    },
    {
      "epoch": 18.95964360587002,
      "grad_norm": 0.12665516138076782,
      "learning_rate": 4.114070600868131e-07,
      "loss": 0.0033,
      "num_input_tokens_seen": 23648824,
      "step": 36175
    },
    {
      "epoch": 18.962264150943398,
      "grad_norm": 0.10681846737861633,
      "learning_rate": 4.093437963510888e-07,
      "loss": 0.0013,
      "num_input_tokens_seen": 23651096,
      "step": 36180
    },
    {
      "epoch": 18.96488469601677,
      "grad_norm": 0.49872153997421265,
      "learning_rate": 4.072856767000949e-07,
      "loss": 0.0083,
      "num_input_tokens_seen": 23654616,
      "step": 36185
    },
    {
      "epoch": 18.967505241090148,
      "grad_norm": 0.08104350417852402,
      "learning_rate": 4.052327015643703e-07,
      "loss": 0.0122,
      "num_input_tokens_seen": 23658200,
      "step": 36190
    },
    {
      "epoch": 18.97012578616352,
      "grad_norm": 0.1472814381122589,
      "learning_rate": 4.031848713733744e-07,
      "loss": 0.0041,
      "num_input_tokens_seen": 23660856,
      "step": 36195
    },
    {
      "epoch": 18.9727463312369,
      "grad_norm": 1.2066487073898315,
      "learning_rate": 4.011421865554837e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 23664056,
      "step": 36200
    },
    {
      "epoch": 18.97536687631027,
      "grad_norm": 0.375305712223053,
      "learning_rate": 3.9910464753801224e-07,
      "loss": 0.0057,
      "num_input_tokens_seen": 23667256,
      "step": 36205
    },
    {
      "epoch": 18.97798742138365,
      "grad_norm": 1.5756958723068237,
      "learning_rate": 3.9707225474718557e-07,
      "loss": 0.0089,
      "num_input_tokens_seen": 23670456,
      "step": 36210
    },
    {
      "epoch": 18.980607966457022,
      "grad_norm": 0.42798465490341187,
      "learning_rate": 3.950450086081553e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23673656,
      "step": 36215
    },
    {
      "epoch": 18.9832285115304,
      "grad_norm": 0.4754306674003601,
      "learning_rate": 3.930229095450072e-07,
      "loss": 0.0154,
      "num_input_tokens_seen": 23676824,
      "step": 36220
    },
    {
      "epoch": 18.985849056603772,
      "grad_norm": 0.4617580473423004,
      "learning_rate": 3.9100595798073345e-07,
      "loss": 0.0061,
      "num_input_tokens_seen": 23679832,
      "step": 36225
    },
    {
      "epoch": 18.98846960167715,
      "grad_norm": 0.17825569212436676,
      "learning_rate": 3.8899415433725493e-07,
      "loss": 0.0067,
      "num_input_tokens_seen": 23682808,
      "step": 36230
    },
    {
      "epoch": 18.991090146750523,
      "grad_norm": 0.23199114203453064,
      "learning_rate": 3.8698749903542666e-07,
      "loss": 0.0043,
      "num_input_tokens_seen": 23685976,
      "step": 36235
    },
    {
      "epoch": 18.9937106918239,
      "grad_norm": 0.22475916147232056,
      "learning_rate": 3.8498599249501e-07,
      "loss": 0.0116,
      "num_input_tokens_seen": 23688952,
      "step": 36240
    },
    {
      "epoch": 18.996331236897273,
      "grad_norm": 0.26651719212532043,
      "learning_rate": 3.8298963513470345e-07,
      "loss": 0.0051,
      "num_input_tokens_seen": 23691832,
      "step": 36245
    },
    {
      "epoch": 18.99895178197065,
      "grad_norm": 0.2564548850059509,
      "learning_rate": 3.8099842737211734e-07,
      "loss": 0.0037,
      "num_input_tokens_seen": 23695544,
      "step": 36250
    },
    {
      "epoch": 19.0,
      "eval_loss": 1.618476152420044,
      "eval_runtime": 11.8083,
      "eval_samples_per_second": 71.814,
      "eval_steps_per_second": 17.953,
      "num_input_tokens_seen": 23696296,
      "step": 36252
    },
    {
      "epoch": 19.001572327044027,
      "grad_norm": 0.4149634838104248,
      "learning_rate": 3.7901236962379627e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 23697768,
      "step": 36255
    },
    {
      "epoch": 19.0041928721174,
      "grad_norm": 0.35710129141807556,
      "learning_rate": 3.7703146230519126e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23700072,
      "step": 36260
    },
    {
      "epoch": 19.006813417190777,
      "grad_norm": 0.09171761572360992,
      "learning_rate": 3.7505570583069306e-07,
      "loss": 0.0087,
      "num_input_tokens_seen": 23702856,
      "step": 36265
    },
    {
      "epoch": 19.00943396226415,
      "grad_norm": 1.0101423263549805,
      "learning_rate": 3.7308510061360435e-07,
      "loss": 0.0068,
      "num_input_tokens_seen": 23706184,
      "step": 36270
    },
    {
      "epoch": 19.012054507337528,
      "grad_norm": 0.12499329447746277,
      "learning_rate": 3.7111964706615374e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 23709128,
      "step": 36275
    },
    {
      "epoch": 19.0146750524109,
      "grad_norm": 1.3158345222473145,
      "learning_rate": 3.691593455994929e-07,
      "loss": 0.0155,
      "num_input_tokens_seen": 23712808,
      "step": 36280
    },
    {
      "epoch": 19.017295597484278,
      "grad_norm": 0.5661552548408508,
      "learning_rate": 3.672041966236911e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 23716168,
      "step": 36285
    },
    {
      "epoch": 19.01991614255765,
      "grad_norm": 0.3635348081588745,
      "learning_rate": 3.6525420054774884e-07,
      "loss": 0.0021,
      "num_input_tokens_seen": 23719688,
      "step": 36290
    },
    {
      "epoch": 19.02253668763103,
      "grad_norm": 0.3976983428001404,
      "learning_rate": 3.6330935777957597e-07,
      "loss": 0.0071,
      "num_input_tokens_seen": 23723752,
      "step": 36295
    },
    {
      "epoch": 19.0251572327044,
      "grad_norm": 0.3606381416320801,
      "learning_rate": 3.613696687260193e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 23727432,
      "step": 36300
    },
    {
      "epoch": 19.02777777777778,
      "grad_norm": 0.15094886720180511,
      "learning_rate": 3.59435133792832e-07,
      "loss": 0.0019,
      "num_input_tokens_seen": 23730312,
      "step": 36305
    },
    {
      "epoch": 19.030398322851152,
      "grad_norm": 0.6061564087867737,
      "learning_rate": 3.575057533847043e-07,
      "loss": 0.0037,
      "num_input_tokens_seen": 23733448,
      "step": 36310
    },
    {
      "epoch": 19.03301886792453,
      "grad_norm": 0.27502143383026123,
      "learning_rate": 3.555815279052327e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23737320,
      "step": 36315
    },
    {
      "epoch": 19.035639412997902,
      "grad_norm": 0.7836617827415466,
      "learning_rate": 3.5366245775694805e-07,
      "loss": 0.0034,
      "num_input_tokens_seen": 23739752,
      "step": 36320
    },
    {
      "epoch": 19.03825995807128,
      "grad_norm": 0.30883222818374634,
      "learning_rate": 3.517485433412987e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 23742152,
      "step": 36325
    },
    {
      "epoch": 19.040880503144653,
      "grad_norm": 1.5290712118148804,
      "learning_rate": 3.4983978505865044e-07,
      "loss": 0.0094,
      "num_input_tokens_seen": 23745448,
      "step": 36330
    },
    {
      "epoch": 19.04350104821803,
      "grad_norm": 0.368207722902298,
      "learning_rate": 3.47936183308295e-07,
      "loss": 0.0159,
      "num_input_tokens_seen": 23749864,
      "step": 36335
    },
    {
      "epoch": 19.046121593291403,
      "grad_norm": 0.6345289349555969,
      "learning_rate": 3.460377384884472e-07,
      "loss": 0.0106,
      "num_input_tokens_seen": 23752552,
      "step": 36340
    },
    {
      "epoch": 19.04874213836478,
      "grad_norm": 0.9096505641937256,
      "learning_rate": 3.4414445099623384e-07,
      "loss": 0.0134,
      "num_input_tokens_seen": 23755528,
      "step": 36345
    },
    {
      "epoch": 19.051362683438157,
      "grad_norm": 0.12680111825466156,
      "learning_rate": 3.42256321227713e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 23758600,
      "step": 36350
    },
    {
      "epoch": 19.05398322851153,
      "grad_norm": 0.591676652431488,
      "learning_rate": 3.403733495778605e-07,
      "loss": 0.0024,
      "num_input_tokens_seen": 23761480,
      "step": 36355
    },
    {
      "epoch": 19.056603773584907,
      "grad_norm": 0.3908744752407074,
      "learning_rate": 3.3849553644056955e-07,
      "loss": 0.0065,
      "num_input_tokens_seen": 23764168,
      "step": 36360
    },
    {
      "epoch": 19.05922431865828,
      "grad_norm": 0.272596538066864,
      "learning_rate": 3.36622882208662e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23766920,
      "step": 36365
    },
    {
      "epoch": 19.061844863731658,
      "grad_norm": 0.8917790651321411,
      "learning_rate": 3.347553872738718e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 23770056,
      "step": 36370
    },
    {
      "epoch": 19.06446540880503,
      "grad_norm": 0.38570982217788696,
      "learning_rate": 3.3289305202685583e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23773096,
      "step": 36375
    },
    {
      "epoch": 19.067085953878408,
      "grad_norm": 0.34737449884414673,
      "learning_rate": 3.310358768571997e-07,
      "loss": 0.0068,
      "num_input_tokens_seen": 23775496,
      "step": 36380
    },
    {
      "epoch": 19.06970649895178,
      "grad_norm": 0.6219481825828552,
      "learning_rate": 3.2918386215339827e-07,
      "loss": 0.0052,
      "num_input_tokens_seen": 23778600,
      "step": 36385
    },
    {
      "epoch": 19.072327044025158,
      "grad_norm": 0.24637463688850403,
      "learning_rate": 3.2733700830287205e-07,
      "loss": 0.0095,
      "num_input_tokens_seen": 23783688,
      "step": 36390
    },
    {
      "epoch": 19.07494758909853,
      "grad_norm": 0.2895199656486511,
      "learning_rate": 3.254953156919649e-07,
      "loss": 0.038,
      "num_input_tokens_seen": 23786440,
      "step": 36395
    },
    {
      "epoch": 19.07756813417191,
      "grad_norm": 0.16349390149116516,
      "learning_rate": 3.2365878470593527e-07,
      "loss": 0.0202,
      "num_input_tokens_seen": 23789768,
      "step": 36400
    },
    {
      "epoch": 19.080188679245282,
      "grad_norm": 0.48394575715065,
      "learning_rate": 3.2182741572896747e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23792264,
      "step": 36405
    },
    {
      "epoch": 19.08280922431866,
      "grad_norm": 0.5795532464981079,
      "learning_rate": 3.2000120914415786e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 23795816,
      "step": 36410
    },
    {
      "epoch": 19.085429769392032,
      "grad_norm": 1.0310546159744263,
      "learning_rate": 3.181801653335315e-07,
      "loss": 0.0429,
      "num_input_tokens_seen": 23798984,
      "step": 36415
    },
    {
      "epoch": 19.08805031446541,
      "grad_norm": 0.4660179316997528,
      "learning_rate": 3.163642846780307e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 23802056,
      "step": 36420
    },
    {
      "epoch": 19.090670859538783,
      "grad_norm": 0.18276773393154144,
      "learning_rate": 3.1455356755751565e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 23804424,
      "step": 36425
    },
    {
      "epoch": 19.09329140461216,
      "grad_norm": 0.14739353954792023,
      "learning_rate": 3.127480143507666e-07,
      "loss": 0.0046,
      "num_input_tokens_seen": 23807112,
      "step": 36430
    },
    {
      "epoch": 19.095911949685533,
      "grad_norm": 0.6243339776992798,
      "learning_rate": 3.1094762543548704e-07,
      "loss": 0.0118,
      "num_input_tokens_seen": 23810184,
      "step": 36435
    },
    {
      "epoch": 19.09853249475891,
      "grad_norm": 0.16763930022716522,
      "learning_rate": 3.091524011882951e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 23812776,
      "step": 36440
    },
    {
      "epoch": 19.101153039832287,
      "grad_norm": 0.6674632430076599,
      "learning_rate": 3.073623419847349e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 23816296,
      "step": 36445
    },
    {
      "epoch": 19.10377358490566,
      "grad_norm": 0.32836249470710754,
      "learning_rate": 3.0557744819926235e-07,
      "loss": 0.0025,
      "num_input_tokens_seen": 23819432,
      "step": 36450
    },
    {
      "epoch": 19.106394129979037,
      "grad_norm": 0.8351295590400696,
      "learning_rate": 3.037977202052594e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23822600,
      "step": 36455
    },
    {
      "epoch": 19.10901467505241,
      "grad_norm": 0.14332377910614014,
      "learning_rate": 3.020231583750255e-07,
      "loss": 0.0033,
      "num_input_tokens_seen": 23825512,
      "step": 36460
    },
    {
      "epoch": 19.111635220125788,
      "grad_norm": 0.3199911415576935,
      "learning_rate": 3.002537630797747e-07,
      "loss": 0.0051,
      "num_input_tokens_seen": 23828456,
      "step": 36465
    },
    {
      "epoch": 19.11425576519916,
      "grad_norm": 0.29722854495048523,
      "learning_rate": 2.984895346896499e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23831656,
      "step": 36470
    },
    {
      "epoch": 19.116876310272538,
      "grad_norm": 0.4000774323940277,
      "learning_rate": 2.967304735737031e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 23834312,
      "step": 36475
    },
    {
      "epoch": 19.11949685534591,
      "grad_norm": 0.35134831070899963,
      "learning_rate": 2.9497658009991214e-07,
      "loss": 0.0051,
      "num_input_tokens_seen": 23836968,
      "step": 36480
    },
    {
      "epoch": 19.122117400419288,
      "grad_norm": 0.2687779366970062,
      "learning_rate": 2.932278546351724e-07,
      "loss": 0.0183,
      "num_input_tokens_seen": 23840616,
      "step": 36485
    },
    {
      "epoch": 19.12473794549266,
      "grad_norm": 1.2375516891479492,
      "learning_rate": 2.9148429754529137e-07,
      "loss": 0.0062,
      "num_input_tokens_seen": 23844168,
      "step": 36490
    },
    {
      "epoch": 19.12735849056604,
      "grad_norm": 0.23954616487026215,
      "learning_rate": 2.897459091950078e-07,
      "loss": 0.0072,
      "num_input_tokens_seen": 23846888,
      "step": 36495
    },
    {
      "epoch": 19.129979035639412,
      "grad_norm": 3.8235714435577393,
      "learning_rate": 2.880126899479724e-07,
      "loss": 0.0063,
      "num_input_tokens_seen": 23850856,
      "step": 36500
    },
    {
      "epoch": 19.13259958071279,
      "grad_norm": 0.15506485104560852,
      "learning_rate": 2.862846401667507e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23853992,
      "step": 36505
    },
    {
      "epoch": 19.135220125786162,
      "grad_norm": 0.23244409263134003,
      "learning_rate": 2.8456176021283697e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23857864,
      "step": 36510
    },
    {
      "epoch": 19.13784067085954,
      "grad_norm": 0.48631998896598816,
      "learning_rate": 2.8284405044662897e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 23861480,
      "step": 36515
    },
    {
      "epoch": 19.140461215932913,
      "grad_norm": 0.7764803171157837,
      "learning_rate": 2.8113151122745874e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23864424,
      "step": 36520
    },
    {
      "epoch": 19.14308176100629,
      "grad_norm": 0.657668948173523,
      "learning_rate": 2.7942414291357023e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 23867272,
      "step": 36525
    },
    {
      "epoch": 19.145702306079663,
      "grad_norm": 0.8171764016151428,
      "learning_rate": 2.777219458621222e-07,
      "loss": 0.006,
      "num_input_tokens_seen": 23871272,
      "step": 36530
    },
    {
      "epoch": 19.14832285115304,
      "grad_norm": 0.04129332676529884,
      "learning_rate": 2.7602492042919923e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23874376,
      "step": 36535
    },
    {
      "epoch": 19.150943396226417,
      "grad_norm": 1.0714194774627686,
      "learning_rate": 2.7433306696979523e-07,
      "loss": 0.0119,
      "num_input_tokens_seen": 23877096,
      "step": 36540
    },
    {
      "epoch": 19.15356394129979,
      "grad_norm": 0.12494219839572906,
      "learning_rate": 2.7264638583782973e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 23880776,
      "step": 36545
    },
    {
      "epoch": 19.156184486373167,
      "grad_norm": 0.24118714034557343,
      "learning_rate": 2.7096487738613177e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 23885000,
      "step": 36550
    },
    {
      "epoch": 19.15880503144654,
      "grad_norm": 0.107294000685215,
      "learning_rate": 2.6928854196646147e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 23887720,
      "step": 36555
    },
    {
      "epoch": 19.161425576519918,
      "grad_norm": 0.33063575625419617,
      "learning_rate": 2.6761737992948565e-07,
      "loss": 0.0042,
      "num_input_tokens_seen": 23890760,
      "step": 36560
    },
    {
      "epoch": 19.16404612159329,
      "grad_norm": 0.17480133473873138,
      "learning_rate": 2.6595139162479134e-07,
      "loss": 0.0033,
      "num_input_tokens_seen": 23893640,
      "step": 36565
    },
    {
      "epoch": 19.166666666666668,
      "grad_norm": 0.27976369857788086,
      "learning_rate": 2.6429057740088303e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 23896584,
      "step": 36570
    },
    {
      "epoch": 19.16928721174004,
      "grad_norm": 0.7628280520439148,
      "learning_rate": 2.6263493760518564e-07,
      "loss": 0.0081,
      "num_input_tokens_seen": 23901480,
      "step": 36575
    },
    {
      "epoch": 19.171907756813418,
      "grad_norm": 0.20127354562282562,
      "learning_rate": 2.6098447258404157e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 23904680,
      "step": 36580
    },
    {
      "epoch": 19.17452830188679,
      "grad_norm": 0.08594129234552383,
      "learning_rate": 2.593391826827107e-07,
      "loss": 0.0015,
      "num_input_tokens_seen": 23907144,
      "step": 36585
    },
    {
      "epoch": 19.17714884696017,
      "grad_norm": 0.22783592343330383,
      "learning_rate": 2.5769906824536225e-07,
      "loss": 0.0062,
      "num_input_tokens_seen": 23910888,
      "step": 36590
    },
    {
      "epoch": 19.179769392033542,
      "grad_norm": 0.9323662519454956,
      "learning_rate": 2.560641296150967e-07,
      "loss": 0.011,
      "num_input_tokens_seen": 23914184,
      "step": 36595
    },
    {
      "epoch": 19.18238993710692,
      "grad_norm": 0.20564626157283783,
      "learning_rate": 2.544343671339211e-07,
      "loss": 0.0116,
      "num_input_tokens_seen": 23917608,
      "step": 36600
    },
    {
      "epoch": 19.185010482180292,
      "grad_norm": 0.3775920271873474,
      "learning_rate": 2.5280978114276264e-07,
      "loss": 0.0086,
      "num_input_tokens_seen": 23922312,
      "step": 36605
    },
    {
      "epoch": 19.18763102725367,
      "grad_norm": 0.3678293228149414,
      "learning_rate": 2.511903719814662e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 23925096,
      "step": 36610
    },
    {
      "epoch": 19.190251572327043,
      "grad_norm": 0.2788933217525482,
      "learning_rate": 2.4957613998879695e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 23927976,
      "step": 36615
    },
    {
      "epoch": 19.19287211740042,
      "grad_norm": 1.2782797813415527,
      "learning_rate": 2.479670855024291e-07,
      "loss": 0.0263,
      "num_input_tokens_seen": 23934760,
      "step": 36620
    },
    {
      "epoch": 19.195492662473793,
      "grad_norm": 0.6674389839172363,
      "learning_rate": 2.463632088589629e-07,
      "loss": 0.0077,
      "num_input_tokens_seen": 23937960,
      "step": 36625
    },
    {
      "epoch": 19.19811320754717,
      "grad_norm": 0.3309580683708191,
      "learning_rate": 2.447645103939078e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 23940904,
      "step": 36630
    },
    {
      "epoch": 19.200733752620547,
      "grad_norm": 0.39178353548049927,
      "learning_rate": 2.431709904416962e-07,
      "loss": 0.0017,
      "num_input_tokens_seen": 23943848,
      "step": 36635
    },
    {
      "epoch": 19.20335429769392,
      "grad_norm": 0.49224522709846497,
      "learning_rate": 2.4158264933567264e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 23947112,
      "step": 36640
    },
    {
      "epoch": 19.205974842767297,
      "grad_norm": 1.5472277402877808,
      "learning_rate": 2.399994874080963e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 23950056,
      "step": 36645
    },
    {
      "epoch": 19.20859538784067,
      "grad_norm": 0.7807113528251648,
      "learning_rate": 2.3842150499015504e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 23953320,
      "step": 36650
    },
    {
      "epoch": 19.211215932914047,
      "grad_norm": 0.17727603018283844,
      "learning_rate": 2.368487024119348e-07,
      "loss": 0.0014,
      "num_input_tokens_seen": 23956168,
      "step": 36655
    },
    {
      "epoch": 19.21383647798742,
      "grad_norm": 0.6717193722724915,
      "learning_rate": 2.3528108000245573e-07,
      "loss": 0.0043,
      "num_input_tokens_seen": 23959528,
      "step": 36660
    },
    {
      "epoch": 19.216457023060798,
      "grad_norm": 0.09781275689601898,
      "learning_rate": 2.3371863808964445e-07,
      "loss": 0.0035,
      "num_input_tokens_seen": 23962504,
      "step": 36665
    },
    {
      "epoch": 19.21907756813417,
      "grad_norm": 1.0238314867019653,
      "learning_rate": 2.32161377000345e-07,
      "loss": 0.0068,
      "num_input_tokens_seen": 23965736,
      "step": 36670
    },
    {
      "epoch": 19.221698113207548,
      "grad_norm": 0.5933910608291626,
      "learning_rate": 2.306092970603163e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 23968328,
      "step": 36675
    },
    {
      "epoch": 19.22431865828092,
      "grad_norm": 0.2605127692222595,
      "learning_rate": 2.2906239859424306e-07,
      "loss": 0.0056,
      "num_input_tokens_seen": 23972360,
      "step": 36680
    },
    {
      "epoch": 19.2269392033543,
      "grad_norm": 1.08118736743927,
      "learning_rate": 2.2752068192571084e-07,
      "loss": 0.0117,
      "num_input_tokens_seen": 23975688,
      "step": 36685
    },
    {
      "epoch": 19.229559748427672,
      "grad_norm": 0.8822554349899292,
      "learning_rate": 2.2598414737723396e-07,
      "loss": 0.0053,
      "num_input_tokens_seen": 23979656,
      "step": 36690
    },
    {
      "epoch": 19.23218029350105,
      "grad_norm": 0.3489609360694885,
      "learning_rate": 2.2445279527023578e-07,
      "loss": 0.0154,
      "num_input_tokens_seen": 23982568,
      "step": 36695
    },
    {
      "epoch": 19.234800838574422,
      "grad_norm": 0.0878843292593956,
      "learning_rate": 2.2292662592506007e-07,
      "loss": 0.0106,
      "num_input_tokens_seen": 23985640,
      "step": 36700
    },
    {
      "epoch": 19.2374213836478,
      "grad_norm": 1.1961859464645386,
      "learning_rate": 2.2140563966095985e-07,
      "loss": 0.0047,
      "num_input_tokens_seen": 23989000,
      "step": 36705
    },
    {
      "epoch": 19.240041928721173,
      "grad_norm": 0.21030229330062866,
      "learning_rate": 2.1988983679611386e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 23991784,
      "step": 36710
    },
    {
      "epoch": 19.24266247379455,
      "grad_norm": 0.5797088742256165,
      "learning_rate": 2.1837921764760462e-07,
      "loss": 0.0053,
      "num_input_tokens_seen": 23996328,
      "step": 36715
    },
    {
      "epoch": 19.245283018867923,
      "grad_norm": 0.06559602916240692,
      "learning_rate": 2.1687378253144043e-07,
      "loss": 0.0222,
      "num_input_tokens_seen": 23999816,
      "step": 36720
    },
    {
      "epoch": 19.2479035639413,
      "grad_norm": 0.21997253596782684,
      "learning_rate": 2.153735317625416e-07,
      "loss": 0.0227,
      "num_input_tokens_seen": 24002472,
      "step": 36725
    },
    {
      "epoch": 19.250524109014677,
      "grad_norm": 0.09457720071077347,
      "learning_rate": 2.1387846565474045e-07,
      "loss": 0.0009,
      "num_input_tokens_seen": 24005256,
      "step": 36730
    },
    {
      "epoch": 19.25314465408805,
      "grad_norm": 0.5845494866371155,
      "learning_rate": 2.123885845207896e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 24008776,
      "step": 36735
    },
    {
      "epoch": 19.255765199161427,
      "grad_norm": 0.2321147620677948,
      "learning_rate": 2.1090388867235356e-07,
      "loss": 0.0023,
      "num_input_tokens_seen": 24012904,
      "step": 36740
    },
    {
      "epoch": 19.2583857442348,
      "grad_norm": 0.3764892518520355,
      "learning_rate": 2.0942437842001728e-07,
      "loss": 0.0059,
      "num_input_tokens_seen": 24016040,
      "step": 36745
    },
    {
      "epoch": 19.261006289308177,
      "grad_norm": 0.030224550515413284,
      "learning_rate": 2.0795005407327483e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24018760,
      "step": 36750
    },
    {
      "epoch": 19.26362683438155,
      "grad_norm": 0.38675811886787415,
      "learning_rate": 2.0648091594053508e-07,
      "loss": 0.0156,
      "num_input_tokens_seen": 24022088,
      "step": 36755
    },
    {
      "epoch": 19.266247379454928,
      "grad_norm": 0.22594459354877472,
      "learning_rate": 2.0501696432913275e-07,
      "loss": 0.0027,
      "num_input_tokens_seen": 24025064,
      "step": 36760
    },
    {
      "epoch": 19.2688679245283,
      "grad_norm": 0.5313555598258972,
      "learning_rate": 2.035581995453034e-07,
      "loss": 0.0042,
      "num_input_tokens_seen": 24028264,
      "step": 36765
    },
    {
      "epoch": 19.271488469601678,
      "grad_norm": 0.44243064522743225,
      "learning_rate": 2.0210462189420577e-07,
      "loss": 0.0055,
      "num_input_tokens_seen": 24031560,
      "step": 36770
    },
    {
      "epoch": 19.27410901467505,
      "grad_norm": 1.2976008653640747,
      "learning_rate": 2.006562316799132e-07,
      "loss": 0.0104,
      "num_input_tokens_seen": 24035944,
      "step": 36775
    },
    {
      "epoch": 19.27672955974843,
      "grad_norm": 0.2090427726507187,
      "learning_rate": 1.992130292054084e-07,
      "loss": 0.0221,
      "num_input_tokens_seen": 24039912,
      "step": 36780
    },
    {
      "epoch": 19.279350104821802,
      "grad_norm": 0.08653858304023743,
      "learning_rate": 1.9777501477259708e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24042792,
      "step": 36785
    },
    {
      "epoch": 19.28197064989518,
      "grad_norm": 0.6318742632865906,
      "learning_rate": 1.9634218868229414e-07,
      "loss": 0.0085,
      "num_input_tokens_seen": 24046184,
      "step": 36790
    },
    {
      "epoch": 19.284591194968552,
      "grad_norm": 0.12265274673700333,
      "learning_rate": 1.9491455123423207e-07,
      "loss": 0.0066,
      "num_input_tokens_seen": 24049576,
      "step": 36795
    },
    {
      "epoch": 19.28721174004193,
      "grad_norm": 0.7216014862060547,
      "learning_rate": 1.9349210272705253e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 24051944,
      "step": 36800
    },
    {
      "epoch": 19.289832285115303,
      "grad_norm": 0.5682976841926575,
      "learning_rate": 1.920748434583175e-07,
      "loss": 0.0076,
      "num_input_tokens_seen": 24055112,
      "step": 36805
    },
    {
      "epoch": 19.29245283018868,
      "grad_norm": 0.6411077380180359,
      "learning_rate": 1.906627737245009e-07,
      "loss": 0.0043,
      "num_input_tokens_seen": 24058888,
      "step": 36810
    },
    {
      "epoch": 19.295073375262053,
      "grad_norm": 0.18595102429389954,
      "learning_rate": 1.8925589382099152e-07,
      "loss": 0.0041,
      "num_input_tokens_seen": 24062728,
      "step": 36815
    },
    {
      "epoch": 19.29769392033543,
      "grad_norm": 0.5114368796348572,
      "learning_rate": 1.8785420404209553e-07,
      "loss": 0.094,
      "num_input_tokens_seen": 24066504,
      "step": 36820
    },
    {
      "epoch": 19.300314465408803,
      "grad_norm": 0.14402474462985992,
      "learning_rate": 1.864577046810284e-07,
      "loss": 0.0064,
      "num_input_tokens_seen": 24069256,
      "step": 36825
    },
    {
      "epoch": 19.30293501048218,
      "grad_norm": 0.2599821984767914,
      "learning_rate": 1.8506639602992038e-07,
      "loss": 0.0019,
      "num_input_tokens_seen": 24071720,
      "step": 36830
    },
    {
      "epoch": 19.305555555555557,
      "grad_norm": 0.3695582151412964,
      "learning_rate": 1.836802783798164e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 24075720,
      "step": 36835
    },
    {
      "epoch": 19.30817610062893,
      "grad_norm": 0.6699880957603455,
      "learning_rate": 1.822993520206817e-07,
      "loss": 0.0058,
      "num_input_tokens_seen": 24078920,
      "step": 36840
    },
    {
      "epoch": 19.310796645702307,
      "grad_norm": 0.2972777485847473,
      "learning_rate": 1.8092361724138528e-07,
      "loss": 0.0044,
      "num_input_tokens_seen": 24082376,
      "step": 36845
    },
    {
      "epoch": 19.31341719077568,
      "grad_norm": 0.5184828042984009,
      "learning_rate": 1.7955307432971914e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 24085928,
      "step": 36850
    },
    {
      "epoch": 19.316037735849058,
      "grad_norm": 0.7888059616088867,
      "learning_rate": 1.7818772357238445e-07,
      "loss": 0.0061,
      "num_input_tokens_seen": 24088648,
      "step": 36855
    },
    {
      "epoch": 19.31865828092243,
      "grad_norm": 0.39409342408180237,
      "learning_rate": 1.7682756525499166e-07,
      "loss": 0.005,
      "num_input_tokens_seen": 24092136,
      "step": 36860
    },
    {
      "epoch": 19.321278825995808,
      "grad_norm": 1.3217964172363281,
      "learning_rate": 1.7547259966207708e-07,
      "loss": 0.0034,
      "num_input_tokens_seen": 24094632,
      "step": 36865
    },
    {
      "epoch": 19.32389937106918,
      "grad_norm": 0.4118402302265167,
      "learning_rate": 1.7412282707708062e-07,
      "loss": 0.0045,
      "num_input_tokens_seen": 24098248,
      "step": 36870
    },
    {
      "epoch": 19.32651991614256,
      "grad_norm": 1.5798981189727783,
      "learning_rate": 1.7277824778235974e-07,
      "loss": 0.0051,
      "num_input_tokens_seen": 24100872,
      "step": 36875
    },
    {
      "epoch": 19.329140461215932,
      "grad_norm": 0.09090428054332733,
      "learning_rate": 1.714388620591867e-07,
      "loss": 0.0095,
      "num_input_tokens_seen": 24104520,
      "step": 36880
    },
    {
      "epoch": 19.33176100628931,
      "grad_norm": 0.2128066122531891,
      "learning_rate": 1.701046701877429e-07,
      "loss": 0.0065,
      "num_input_tokens_seen": 24111048,
      "step": 36885
    },
    {
      "epoch": 19.334381551362682,
      "grad_norm": 0.2233431190252304,
      "learning_rate": 1.6877567244713011e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 24113448,
      "step": 36890
    },
    {
      "epoch": 19.33700209643606,
      "grad_norm": 0.08369435369968414,
      "learning_rate": 1.6745186911535372e-07,
      "loss": 0.0033,
      "num_input_tokens_seen": 24116264,
      "step": 36895
    },
    {
      "epoch": 19.339622641509433,
      "grad_norm": 0.12162046134471893,
      "learning_rate": 1.6613326046934497e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 24118888,
      "step": 36900
    },
    {
      "epoch": 19.34224318658281,
      "grad_norm": 0.156576469540596,
      "learning_rate": 1.64819846784936e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 24122056,
      "step": 36905
    },
    {
      "epoch": 19.344863731656183,
      "grad_norm": 0.2435082644224167,
      "learning_rate": 1.6351162833688204e-07,
      "loss": 0.0149,
      "num_input_tokens_seen": 24125288,
      "step": 36910
    },
    {
      "epoch": 19.34748427672956,
      "grad_norm": 0.15785981714725494,
      "learning_rate": 1.6220860539884475e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 24128904,
      "step": 36915
    },
    {
      "epoch": 19.350104821802937,
      "grad_norm": 0.25490739941596985,
      "learning_rate": 1.6091077824340327e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 24131880,
      "step": 36920
    },
    {
      "epoch": 19.35272536687631,
      "grad_norm": 0.06879004091024399,
      "learning_rate": 1.5961814714204605e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24134600,
      "step": 36925
    },
    {
      "epoch": 19.355345911949687,
      "grad_norm": 0.2618641257286072,
      "learning_rate": 1.5833071236518172e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 24137448,
      "step": 36930
    },
    {
      "epoch": 19.35796645702306,
      "grad_norm": 0.35270434617996216,
      "learning_rate": 1.5704847418212266e-07,
      "loss": 0.0138,
      "num_input_tokens_seen": 24141064,
      "step": 36935
    },
    {
      "epoch": 19.360587002096437,
      "grad_norm": 0.7382941842079163,
      "learning_rate": 1.5577143286109875e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 24143688,
      "step": 36940
    },
    {
      "epoch": 19.36320754716981,
      "grad_norm": 0.4330906271934509,
      "learning_rate": 1.5449958866925185e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 24146984,
      "step": 36945
    },
    {
      "epoch": 19.365828092243188,
      "grad_norm": 0.20832446217536926,
      "learning_rate": 1.532329418726386e-07,
      "loss": 0.0162,
      "num_input_tokens_seen": 24151784,
      "step": 36950
    },
    {
      "epoch": 19.36844863731656,
      "grad_norm": 0.7230214476585388,
      "learning_rate": 1.5197149273623036e-07,
      "loss": 0.0066,
      "num_input_tokens_seen": 24154792,
      "step": 36955
    },
    {
      "epoch": 19.371069182389938,
      "grad_norm": 0.862203061580658,
      "learning_rate": 1.5071524152389936e-07,
      "loss": 0.003,
      "num_input_tokens_seen": 24157512,
      "step": 36960
    },
    {
      "epoch": 19.37368972746331,
      "grad_norm": 0.3557215631008148,
      "learning_rate": 1.4946418849844934e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24160392,
      "step": 36965
    },
    {
      "epoch": 19.37631027253669,
      "grad_norm": 0.16979092359542847,
      "learning_rate": 1.4821833392157923e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 24163656,
      "step": 36970
    },
    {
      "epoch": 19.378930817610062,
      "grad_norm": 0.2295307070016861,
      "learning_rate": 1.4697767805390838e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24166632,
      "step": 36975
    },
    {
      "epoch": 19.38155136268344,
      "grad_norm": 0.11190370470285416,
      "learning_rate": 1.4574222115497084e-07,
      "loss": 0.0062,
      "num_input_tokens_seen": 24169320,
      "step": 36980
    },
    {
      "epoch": 19.384171907756812,
      "grad_norm": 0.41848039627075195,
      "learning_rate": 1.4451196348320706e-07,
      "loss": 0.0106,
      "num_input_tokens_seen": 24172520,
      "step": 36985
    },
    {
      "epoch": 19.38679245283019,
      "grad_norm": 0.5499804615974426,
      "learning_rate": 1.4328690529597232e-07,
      "loss": 0.0028,
      "num_input_tokens_seen": 24176232,
      "step": 36990
    },
    {
      "epoch": 19.389412997903563,
      "grad_norm": 1.2156320810317993,
      "learning_rate": 1.4206704684953943e-07,
      "loss": 0.0061,
      "num_input_tokens_seen": 24179400,
      "step": 36995
    },
    {
      "epoch": 19.39203354297694,
      "grad_norm": 1.0074368715286255,
      "learning_rate": 1.4085238839908476e-07,
      "loss": 0.0055,
      "num_input_tokens_seen": 24182568,
      "step": 37000
    },
    {
      "epoch": 19.394654088050313,
      "grad_norm": 0.21504908800125122,
      "learning_rate": 1.3964293019870235e-07,
      "loss": 0.0049,
      "num_input_tokens_seen": 24185640,
      "step": 37005
    },
    {
      "epoch": 19.39727463312369,
      "grad_norm": 0.1802881956100464,
      "learning_rate": 1.3843867250139807e-07,
      "loss": 0.0049,
      "num_input_tokens_seen": 24189256,
      "step": 37010
    },
    {
      "epoch": 19.399895178197063,
      "grad_norm": 0.4154653549194336,
      "learning_rate": 1.3723961555908715e-07,
      "loss": 0.0091,
      "num_input_tokens_seen": 24192744,
      "step": 37015
    },
    {
      "epoch": 19.40251572327044,
      "grad_norm": 0.16430869698524475,
      "learning_rate": 1.3604575962259947e-07,
      "loss": 0.027,
      "num_input_tokens_seen": 24197128,
      "step": 37020
    },
    {
      "epoch": 19.405136268343817,
      "grad_norm": 0.2639158368110657,
      "learning_rate": 1.3485710494167692e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 24200552,
      "step": 37025
    },
    {
      "epoch": 19.40775681341719,
      "grad_norm": 0.30419668555259705,
      "learning_rate": 1.3367365176496783e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 24203880,
      "step": 37030
    },
    {
      "epoch": 19.410377358490567,
      "grad_norm": 0.2216290682554245,
      "learning_rate": 1.324954003400436e-07,
      "loss": 0.0055,
      "num_input_tokens_seen": 24207752,
      "step": 37035
    },
    {
      "epoch": 19.41299790356394,
      "grad_norm": 0.29804736375808716,
      "learning_rate": 1.313223509133793e-07,
      "loss": 0.0066,
      "num_input_tokens_seen": 24211752,
      "step": 37040
    },
    {
      "epoch": 19.415618448637318,
      "grad_norm": 0.057701367884874344,
      "learning_rate": 1.301545037303592e-07,
      "loss": 0.0039,
      "num_input_tokens_seen": 24213992,
      "step": 37045
    },
    {
      "epoch": 19.41823899371069,
      "grad_norm": 2.4888696670532227,
      "learning_rate": 1.2899185903529066e-07,
      "loss": 0.02,
      "num_input_tokens_seen": 24216584,
      "step": 37050
    },
    {
      "epoch": 19.420859538784068,
      "grad_norm": 0.27223271131515503,
      "learning_rate": 1.278344170713791e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 24219336,
      "step": 37055
    },
    {
      "epoch": 19.42348008385744,
      "grad_norm": 0.09444055706262589,
      "learning_rate": 1.2668217808075033e-07,
      "loss": 0.0048,
      "num_input_tokens_seen": 24222024,
      "step": 37060
    },
    {
      "epoch": 19.42610062893082,
      "grad_norm": 0.18052731454372406,
      "learning_rate": 1.255351423044393e-07,
      "loss": 0.0029,
      "num_input_tokens_seen": 24225384,
      "step": 37065
    },
    {
      "epoch": 19.428721174004192,
      "grad_norm": 0.1407448649406433,
      "learning_rate": 1.2439330998239572e-07,
      "loss": 0.0026,
      "num_input_tokens_seen": 24228232,
      "step": 37070
    },
    {
      "epoch": 19.43134171907757,
      "grad_norm": 0.14916135370731354,
      "learning_rate": 1.2325668135347578e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24231432,
      "step": 37075
    },
    {
      "epoch": 19.433962264150942,
      "grad_norm": 0.7073591351509094,
      "learning_rate": 1.221252566554476e-07,
      "loss": 0.0121,
      "num_input_tokens_seen": 24234952,
      "step": 37080
    },
    {
      "epoch": 19.43658280922432,
      "grad_norm": 0.18676412105560303,
      "learning_rate": 1.2099903612499685e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 24237768,
      "step": 37085
    },
    {
      "epoch": 19.439203354297693,
      "grad_norm": 0.5402892231941223,
      "learning_rate": 1.198780199977101e-07,
      "loss": 0.0098,
      "num_input_tokens_seen": 24241032,
      "step": 37090
    },
    {
      "epoch": 19.44182389937107,
      "grad_norm": 0.05773334950208664,
      "learning_rate": 1.1876220850809416e-07,
      "loss": 0.0067,
      "num_input_tokens_seen": 24245416,
      "step": 37095
    },
    {
      "epoch": 19.444444444444443,
      "grad_norm": 0.19122633337974548,
      "learning_rate": 1.1765160188956514e-07,
      "loss": 0.0034,
      "num_input_tokens_seen": 24249128,
      "step": 37100
    },
    {
      "epoch": 19.44706498951782,
      "grad_norm": 0.17280955612659454,
      "learning_rate": 1.1654620037445107e-07,
      "loss": 0.002,
      "num_input_tokens_seen": 24252200,
      "step": 37105
    },
    {
      "epoch": 19.449685534591197,
      "grad_norm": 0.36196908354759216,
      "learning_rate": 1.1544600419398088e-07,
      "loss": 0.0037,
      "num_input_tokens_seen": 24255112,
      "step": 37110
    },
    {
      "epoch": 19.45230607966457,
      "grad_norm": 0.5354392528533936,
      "learning_rate": 1.1435101357831213e-07,
      "loss": 0.0099,
      "num_input_tokens_seen": 24258504,
      "step": 37115
    },
    {
      "epoch": 19.454926624737947,
      "grad_norm": 0.10881184041500092,
      "learning_rate": 1.1326122875650047e-07,
      "loss": 0.0018,
      "num_input_tokens_seen": 24260776,
      "step": 37120
    },
    {
      "epoch": 19.45754716981132,
      "grad_norm": 0.1980711966753006,
      "learning_rate": 1.1217664995651633e-07,
      "loss": 0.0049,
      "num_input_tokens_seen": 24263976,
      "step": 37125
    },
    {
      "epoch": 19.460167714884697,
      "grad_norm": 0.4032990634441376,
      "learning_rate": 1.1109727740523935e-07,
      "loss": 0.0022,
      "num_input_tokens_seen": 24267144,
      "step": 37130
    },
    {
      "epoch": 19.46278825995807,
      "grad_norm": 0.22313547134399414,
      "learning_rate": 1.1002311132846944e-07,
      "loss": 0.0031,
      "num_input_tokens_seen": 24270312,
      "step": 37135
    },
    {
      "epoch": 19.465408805031448,
      "grad_norm": 0.2915721833705902,
      "learning_rate": 1.0895415195090186e-07,
      "loss": 0.0032,
      "num_input_tokens_seen": 24274632,
      "step": 37140
    },
    {
      "epoch": 19.46802935010482,
      "grad_norm": 0.14504286646842957,
      "learning_rate": 1.0789039949615221e-07,
      "loss": 0.021,
      "num_input_tokens_seen": 24279144,
      "step": 37145
    },
    {
      "epoch": 19.470649895178198,
      "grad_norm": 0.1933012306690216,
      "learning_rate": 1.0683185418674801e-07,
      "loss": 0.0036,
      "num_input_tokens_seen": 24282024,
      "step": 37150
    },
    {
      "epoch": 19.47327044025157,
      "grad_norm": 0.2632312476634979,
      "learning_rate": 1.0577851624412604e-07,
      "loss": 0.0073,
      "num_input_tokens_seen": 24285288,
      "step": 37155
    },
    {
      "epoch": 19.47589098532495,
      "grad_norm": 0.07663071900606155,
      "learning_rate": 1.0473038588862672e-07,
      "loss": 0.0055,
      "num_input_tokens_seen": 24289384,
      "step": 37160
    },
    {
      "epoch": 19.478511530398322,
      "grad_norm": 2.862516164779663,
      "learning_rate": 1.0368746333951351e-07,
      "loss": 0.0479,
      "num_input_tokens_seen": 24291848,
      "step": 37165
    },
    {
      "epoch": 19.4811320754717,
      "grad_norm": 0.17520315945148468,
      "learning_rate": 1.0264974881494805e-07,
      "loss": 0.0038,
      "num_input_tokens_seen": 24296296,
      "step": 37170
    },
    {
      "epoch": 19.483752620545072,
      "grad_norm": 0.1313394457101822,
      "learning_rate": 1.0161724253201222e-07,
      "loss": 0.0438,
      "num_input_tokens_seen": 24299176,
      "step": 37175
    },
    {
      "epoch": 19.48637316561845,
      "grad_norm": 0.33491480350494385,
      "learning_rate": 1.0058994470669436e-07,
      "loss": 0.0058,
      "num_input_tokens_seen": 24302216,
      "step": 37180
    },
    {
      "epoch": 19.488993710691823,
      "grad_norm": 4.111268043518066,
      "learning_rate": 9.956785555388926e-08,
      "loss": 0.0042,
      "num_input_tokens_seen": 24304648,
      "step": 37185
    },
    {
      "epoch": 19.4916142557652,
      "grad_norm": 0.08050554990768433,
      "learning_rate": 9.855097528741475e-08,
      "loss": 0.0055,
      "num_input_tokens_seen": 24308712,
      "step": 37190
    },
    {
      "epoch": 19.494234800838573,
      "grad_norm": 0.08888456225395203,
      "learning_rate": 9.753930411998402e-08,
      "loss": 0.0043,
      "num_input_tokens_seen": 24312072,
      "step": 37195
    },
    {
      "epoch": 19.49685534591195,
      "grad_norm": 0.3161020874977112,
      "learning_rate": 9.653284226322501e-08,
      "loss": 0.0053,
      "num_input_tokens_seen": 24318312,
      "step": 37200
    },
    {
      "epoch": 19.499475890985323,
      "grad_norm": 0.14791953563690186,
      "learning_rate": 9.553158992768318e-08,
      "loss": 0.0032,
      "num_input_tokens_seen": 24321128,
      "step": 37205
    },
    {
      "epoch": 19.5020964360587,
      "grad_norm": 0.4276447594165802,
      "learning_rate": 9.453554732281044e-08,
      "loss": 0.0032,
      "num_input_tokens_seen": 24324392,
      "step": 37210
    },
    {
      "epoch": 19.504716981132077,
      "grad_norm": 0.23689861595630646,
      "learning_rate": 9.354471465696236e-08,
      "loss": 0.0065,
      "num_input_tokens_seen": 24327560,
      "step": 37215
    },
    {
      "epoch": 19.50733752620545,
      "grad_norm": 0.1607091873884201,
      "learning_rate": 9.255909213741199e-08,
      "loss": 0.0422,
      "num_input_tokens_seen": 24330504,
      "step": 37220
    },
    {
      "epoch": 19.509958071278827,
      "grad_norm": 0.02914956398308277,
      "learning_rate": 9.157867997034164e-08,
      "loss": 0.0062,
      "num_input_tokens_seen": 24333800,
      "step": 37225
    },
    {
      "epoch": 19.5125786163522,
      "grad_norm": 0.44544729590415955,
      "learning_rate": 9.060347836084004e-08,
      "loss": 0.0066,
      "num_input_tokens_seen": 24336264,
      "step": 37230
    },
    {
      "epoch": 19.515199161425578,
      "grad_norm": 0.1775171309709549,
      "learning_rate": 8.96334875129079e-08,
      "loss": 0.0195,
      "num_input_tokens_seen": 24338696,
      "step": 37235
    },
    {
      "epoch": 19.51781970649895,
      "grad_norm": 0.21046939492225647,
      "learning_rate": 8.866870762946067e-08,
      "loss": 0.0024,
      "num_input_tokens_seen": 24342056,
      "step": 37240
    },
    {
      "epoch": 19.520440251572328,
      "grad_norm": 0.7975975871086121,
      "learning_rate": 8.77091389123147e-08,
      "loss": 0.0084,
      "num_input_tokens_seen": 24345480,
      "step": 37245
    },
    {
      "epoch": 19.5230607966457,
      "grad_norm": 1.3921375274658203,
      "learning_rate": 8.675478156220385e-08,
      "loss": 0.0138,
      "num_input_tokens_seen": 24348680,
      "step": 37250
    },
    {
      "epoch": 19.52568134171908,
      "grad_norm": 0.31284090876579285,
      "learning_rate": 8.580563577876566e-08,
      "loss": 0.0075,
      "num_input_tokens_seen": 24351624,
      "step": 37255
    },
    {
      "epoch": 19.528301886792452,
      "grad_norm": 0.27507516741752625,
      "learning_rate": 8.486170176055242e-08,
      "loss": 0.0089,
      "num_input_tokens_seen": 24356552,
      "step": 37260
    },
    {
      "epoch": 19.53092243186583,
      "grad_norm": 0.18976488709449768,
      "learning_rate": 8.392297970502282e-08,
      "loss": 0.0028,
      "num_input_tokens_seen": 24359688,
      "step": 37265
    },
    {
      "epoch": 19.533542976939202,
      "grad_norm": 0.4587344527244568,
      "learning_rate": 8.298946980855315e-08,
      "loss": 0.0078,
      "num_input_tokens_seen": 24364328,
      "step": 37270
    },
    {
      "epoch": 19.53616352201258,
      "grad_norm": 0.16099956631660461,
      "learning_rate": 8.206117226641497e-08,
      "loss": 0.0014,
      "num_input_tokens_seen": 24367752,
      "step": 37275
    },
    {
      "epoch": 19.538784067085953,
      "grad_norm": 0.714459240436554,
      "learning_rate": 8.113808727280015e-08,
      "loss": 0.0049,
      "num_input_tokens_seen": 24371208,
      "step": 37280
    },
    {
      "epoch": 19.54140461215933,
      "grad_norm": 0.3450953960418701,
      "learning_rate": 8.022021502080979e-08,
      "loss": 0.0065,
      "num_input_tokens_seen": 24374824,
      "step": 37285
    },
    {
      "epoch": 19.544025157232703,
      "grad_norm": 0.3144988417625427,
      "learning_rate": 7.930755570244863e-08,
      "loss": 0.0047,
      "num_input_tokens_seen": 24378408,
      "step": 37290
    },
    {
      "epoch": 19.54664570230608,
      "grad_norm": 1.2272160053253174,
      "learning_rate": 7.840010950863619e-08,
      "loss": 0.0049,
      "num_input_tokens_seen": 24381160,
      "step": 37295
    },
    {
      "epoch": 19.549266247379457,
      "grad_norm": 1.1606745719909668,
      "learning_rate": 7.749787662920116e-08,
      "loss": 0.0095,
      "num_input_tokens_seen": 24384072,
      "step": 37300
    },
    {
      "epoch": 19.55188679245283,
      "grad_norm": 0.14509065449237823,
      "learning_rate": 7.660085725288146e-08,
      "loss": 0.0035,
      "num_input_tokens_seen": 24386952,
      "step": 37305
    },
    {
      "epoch": 19.554507337526207,
      "grad_norm": 0.39499732851982117,
      "learning_rate": 7.570905156731867e-08,
      "loss": 0.0039,
      "num_input_tokens_seen": 24390600,
      "step": 37310
    },
    {
      "epoch": 19.55712788259958,
      "grad_norm": 0.2403491884469986,
      "learning_rate": 7.482245975907188e-08,
      "loss": 0.006,
      "num_input_tokens_seen": 24393416,
      "step": 37315
    },
    {
      "epoch": 19.559748427672957,
      "grad_norm": 0.6393749117851257,
      "learning_rate": 7.394108201360383e-08,
      "loss": 0.0028,
      "num_input_tokens_seen": 24396328,
      "step": 37320
    },
    {
      "epoch": 19.56236897274633,
      "grad_norm": 0.1356205940246582,
      "learning_rate": 7.306491851529206e-08,
      "loss": 0.0278,
      "num_input_tokens_seen": 24404040,
      "step": 37325
    },
    {
      "epoch": 19.564989517819708,
      "grad_norm": 0.8267420530319214,
      "learning_rate": 7.219396944741774e-08,
      "loss": 0.0122,
      "num_input_tokens_seen": 24407304,
      "step": 37330
    },
    {
      "epoch": 19.56761006289308,
      "grad_norm": 0.2338176816701889,
      "learning_rate": 7.132823499217123e-08,
      "loss": 0.0132,
      "num_input_tokens_seen": 24411496,
      "step": 37335
    },
    {
      "epoch": 19.570230607966458,
      "grad_norm": 0.5144206285476685,
      "learning_rate": 7.046771533065766e-08,
      "loss": 0.0019,
      "num_input_tokens_seen": 24414280,
      "step": 37340
    },
    {
      "epoch": 19.57285115303983,
      "grad_norm": 0.8209997415542603,
      "learning_rate": 6.961241064288859e-08,
      "loss": 0.0026,
      "num_input_tokens_seen": 24417352,
      "step": 37345
    },
    {
      "epoch": 19.57547169811321,
      "grad_norm": 0.2299492359161377,
      "learning_rate": 6.876232110778202e-08,
      "loss": 0.0052,
      "num_input_tokens_seen": 24421192,
      "step": 37350
    },
    {
      "epoch": 19.578092243186582,
      "grad_norm": 0.4161807596683502,
      "learning_rate": 6.791744690316515e-08,
      "loss": 0.0038,
      "num_input_tokens_seen": 24424744,
      "step": 37355
    },
    {
      "epoch": 19.58071278825996,
      "grad_norm": 1.4563555717468262,
      "learning_rate": 6.707778820577715e-08,
      "loss": 0.011,
      "num_input_tokens_seen": 24427592,
      "step": 37360
    },
    {
      "epoch": 19.583333333333332,
      "grad_norm": 0.9199591875076294,
      "learning_rate": 6.624334519126919e-08,
      "loss": 0.0049,
      "num_input_tokens_seen": 24430600,
      "step": 37365
    },
    {
      "epoch": 19.58595387840671,
      "grad_norm": 0.2343871295452118,
      "learning_rate": 6.541411803419329e-08,
      "loss": 0.0075,
      "num_input_tokens_seen": 24433160,
      "step": 37370
    },
    {
      "epoch": 19.588574423480082,
      "grad_norm": 0.12170210480690002,
      "learning_rate": 6.459010690801626e-08,
      "loss": 0.0025,
      "num_input_tokens_seen": 24435656,
      "step": 37375
    },
    {
      "epoch": 19.59119496855346,
      "grad_norm": 1.6749132871627808,
      "learning_rate": 6.377131198510855e-08,
      "loss": 0.0086,
      "num_input_tokens_seen": 24438440,
      "step": 37380
    },
    {
      "epoch": 19.593815513626833,
      "grad_norm": 0.8695160150527954,
      "learning_rate": 6.295773343675537e-08,
      "loss": 0.0039,
      "num_input_tokens_seen": 24441416,
      "step": 37385
    },
    {
      "epoch": 19.59643605870021,
      "grad_norm": 0.1730542629957199,
      "learning_rate": 6.214937143314837e-08,
      "loss": 0.0022,
      "num_input_tokens_seen": 24444328,
      "step": 37390
    },
    {
      "epoch": 19.599056603773583,
      "grad_norm": 0.20985764265060425,
      "learning_rate": 6.134622614338559e-08,
      "loss": 0.0037,
      "num_input_tokens_seen": 24446472,
      "step": 37395
    },
    {
      "epoch": 19.60167714884696,
      "grad_norm": 0.7747572660446167,
      "learning_rate": 6.05482977354771e-08,
      "loss": 0.013,
      "num_input_tokens_seen": 24449544,
      "step": 37400
    },
    {
      "epoch": 19.604297693920337,
      "grad_norm": 0.5871946215629578,
      "learning_rate": 5.975558637634215e-08,
      "loss": 0.0111,
      "num_input_tokens_seen": 24452776,
      "step": 37405
    },
    {
      "epoch": 19.60691823899371,
      "grad_norm": 0.22111593186855316,
      "learning_rate": 5.896809223180366e-08,
      "loss": 0.0024,
      "num_input_tokens_seen": 24455560,
      "step": 37410
    },
    {
      "epoch": 19.609538784067087,
      "grad_norm": 0.09516289830207825,
      "learning_rate": 5.8185815466596516e-08,
      "loss": 0.0019,
      "num_input_tokens_seen": 24458408,
      "step": 37415
    },
    {
      "epoch": 19.61215932914046,
      "grad_norm": 0.137550488114357,
      "learning_rate": 5.7408756244370365e-08,
      "loss": 0.0061,
      "num_input_tokens_seen": 24461576,
      "step": 37420
    },
    {
      "epoch": 19.614779874213838,
      "grad_norm": 0.3078208863735199,
      "learning_rate": 5.663691472766741e-08,
      "loss": 0.005,
      "num_input_tokens_seen": 24464840,
      "step": 37425
    },
    {
      "epoch": 19.61740041928721,
      "grad_norm": 0.1716417819261551,
      "learning_rate": 5.587029107795572e-08,
      "loss": 0.0018,
      "num_input_tokens_seen": 24468200,
      "step": 37430
    },
    {
      "epoch": 19.620020964360588,
      "grad_norm": 0.268918514251709,
      "learning_rate": 5.510888545559867e-08,
      "loss": 0.0026,
      "num_input_tokens_seen": 24471176,
      "step": 37435
    },
    {
      "epoch": 19.62264150943396,
      "grad_norm": 0.2808757424354553,
      "learning_rate": 5.435269801987997e-08,
      "loss": 0.002,
      "num_input_tokens_seen": 24473960,
      "step": 37440
    },
    {
      "epoch": 19.62526205450734,
      "grad_norm": 0.6203911900520325,
      "learning_rate": 5.360172892897863e-08,
      "loss": 0.0043,
      "num_input_tokens_seen": 24476776,
      "step": 37445
    },
    {
      "epoch": 19.627882599580712,
      "grad_norm": 0.2179444283246994,
      "learning_rate": 5.285597833999401e-08,
      "loss": 0.0053,
      "num_input_tokens_seen": 24480456,
      "step": 37450
    },
    {
      "epoch": 19.63050314465409,
      "grad_norm": 0.2537921965122223,
      "learning_rate": 5.2115446408926295e-08,
      "loss": 0.0031,
      "num_input_tokens_seen": 24483400,
      "step": 37455
    },
    {
      "epoch": 19.633123689727462,
      "grad_norm": 0.12092757970094681,
      "learning_rate": 5.138013329068769e-08,
      "loss": 0.0115,
      "num_input_tokens_seen": 24486536,
      "step": 37460
    },
    {
      "epoch": 19.63574423480084,
      "grad_norm": 0.7975555658340454,
      "learning_rate": 5.065003913909405e-08,
      "loss": 0.0057,
      "num_input_tokens_seen": 24490056,
      "step": 37465
    },
    {
      "epoch": 19.638364779874212,
      "grad_norm": 0.03799581900238991,
      "learning_rate": 4.9925164106875975e-08,
      "loss": 0.004,
      "num_input_tokens_seen": 24493160,
      "step": 37470
    },
    {
      "epoch": 19.64098532494759,
      "grad_norm": 0.7188948392868042,
      "learning_rate": 4.9205508345670506e-08,
      "loss": 0.0054,
      "num_input_tokens_seen": 24496200,
      "step": 37475
    },
    {
      "epoch": 19.643605870020963,
      "grad_norm": 0.26680150628089905,
      "learning_rate": 4.8491072006015546e-08,
      "loss": 0.0079,
      "num_input_tokens_seen": 24500680,
      "step": 37480
    },
    {
      "epoch": 19.64622641509434,
      "grad_norm": 0.1861567497253418,
      "learning_rate": 4.778185523736933e-08,
      "loss": 0.0632,
      "num_input_tokens_seen": 24503720,
      "step": 37485
    },
    {
      "epoch": 19.648846960167717,
      "grad_norm": 0.6314684152603149,
      "learning_rate": 4.7077858188085414e-08,
      "loss": 0.0063,
      "num_input_tokens_seen": 24506440,
      "step": 37490
    },
    {
      "epoch": 19.65146750524109,
      "grad_norm": 0.2396109253168106,
      "learning_rate": 4.6379081005437644e-08,
      "loss": 0.0037,
      "num_input_tokens_seen": 24509384,
      "step": 37495
    },
    {
      "epoch": 19.654088050314467,
      "grad_norm": 0.5037866830825806,
      "learning_rate": 4.568552383560076e-08,
      "loss": 0.0045,
      "num_input_tokens_seen": 24511720,
      "step": 37500
    },
    {
      "epoch": 19.65670859538784,
      "grad_norm": 0.18340222537517548,
      "learning_rate": 4.4997186823655945e-08,
      "loss": 0.0027,
      "num_input_tokens_seen": 24514472,
      "step": 37505
    },
    {
      "epoch": 19.659329140461217,
      "grad_norm": 0.23432253301143646,
      "learning_rate": 4.431407011360189e-08,
      "loss": 0.003,
      "num_input_tokens_seen": 24517224,
      "step": 37510
    },
    {
      "epoch": 19.66194968553459,
      "grad_norm": 6.093330383300781,
      "learning_rate": 4.3636173848332627e-08,
      "loss": 0.0056,
      "num_input_tokens_seen": 24519944,
      "step": 37515
    },
    {
      "epoch": 19.664570230607968,
      "grad_norm": 0.32191285490989685,
      "learning_rate": 4.2963498169656946e-08,
      "loss": 0.01,
      "num_input_tokens_seen": 24524840,
      "step": 37520
    },
    {
      "epoch": 19.66719077568134,
      "grad_norm": 0.36787667870521545,
      "learning_rate": 4.229604321829561e-08,
      "loss": 0.0025,
      "num_input_tokens_seen": 24527912,
      "step": 37525
    },
    {
      "epoch": 19.669811320754718,
      "grad_norm": 0.2608681917190552,
      "learning_rate": 4.16338091338675e-08,
      "loss": 0.0047,
      "num_input_tokens_seen": 24530696,
      "step": 37530
    },
    {
      "epoch": 19.67243186582809,
      "grad_norm": 0.19873139262199402,
      "learning_rate": 4.097679605490623e-08,
      "loss": 0.0011,
      "num_input_tokens_seen": 24533256,
      "step": 37535
    },
    {
      "epoch": 19.67505241090147,
      "grad_norm": 0.16801980137825012,
      "learning_rate": 4.032500411885465e-08,
      "loss": 0.0049,
      "num_input_tokens_seen": 24536552,
      "step": 37540
    },
    {
      "epoch": 19.677672955974842,
      "grad_norm": 0.08663252741098404,
      "learning_rate": 3.967843346205369e-08,
      "loss": 0.0044,
      "num_input_tokens_seen": 24539464,
      "step": 37545
    },
    {
      "epoch": 19.68029350104822,
      "grad_norm": 0.5671118497848511,
      "learning_rate": 3.903708421976182e-08,
      "loss": 0.0021,
      "num_input_tokens_seen": 24542312,
      "step": 37550
    },
    {
      "epoch": 19.682914046121592,
      "grad_norm": 0.6065060496330261,
      "learning_rate": 3.840095652614395e-08,
      "loss": 0.0024,
      "num_input_tokens_seen": 24546056,
      "step": 37555
    },
    {
      "epoch": 19.68553459119497,
      "grad_norm": 0.7284513115882874,
      "learning_rate": 3.777005051426863e-08,
      "loss": 0.0064,
      "num_input_tokens_seen": 24549512,
      "step": 37560
    },
    {
      "epoch": 19.688155136268342,
      "grad_norm": 0.30521515011787415,
      "learning_rate": 3.71443663161164e-08,
      "loss": 0.003,
      "num_input_tokens_seen": 24553000,
      "step": 37565
    },
    {
      "epoch": 19.69077568134172,
      "grad_norm": 0.43618884682655334,
      "learning_rate": 3.652390406256867e-08,
      "loss": 0.1698,
      "num_input_tokens_seen": 24555944,
      "step": 37570
    },
    {
      "epoch": 19.693396226415093,
      "grad_norm": 0.23612584173679352,
      "learning_rate": 3.5908663883421624e-08,
      "loss": 0.0034,
      "num_input_tokens_seen": 24558600,
      "step": 37575
    },
    {
      "epoch": 19.69601677148847,
      "grad_norm": 0.1733483523130417,
      "learning_rate": 3.529864590737786e-08,
      "loss": 0.0032,
      "num_input_tokens_seen": 24562440,
      "step": 37580
    },
    {
      "epoch": 19.698637316561843,
      "grad_norm": 0.5714046955108643,
      "learning_rate": 3.4693850262046415e-08,
      "loss": 0.0303,
      "num_input_tokens_seen": 24565736,
      "step": 37585
    },
    {
      "epoch": 19.70125786163522,
      "grad_norm": 0.3204423487186432,
      "learning_rate": 3.409427707394275e-08,
      "loss": 0.0039,
      "num_input_tokens_seen": 24569064,
      "step": 37590
    },
    {
      "epoch": 19.703878406708597,
      "grad_norm": 0.12052324414253235,
      "learning_rate": 3.349992646848599e-08,
      "loss": 0.0039,
      "num_input_tokens_seen": 24572584,
      "step": 37595
    },
    {
      "epoch": 19.70649895178197,
      "grad_norm": 0.17404261231422424,
      "learning_rate": 3.291079857001556e-08,
      "loss": 0.0255,
      "num_input_tokens_seen": 24575240,
      "step": 37600
    },
    {
      "epoch": 19.709119496855347,
      "grad_norm": 0.10259600728750229,
      "learning_rate": 3.2326893501763436e-08,
      "loss": 0.0346,
      "num_input_tokens_seen": 24578408,
      "step": 37605
    },
    {
      "epoch": 19.71174004192872,
      "grad_norm": 0.22222478687763214,
      "learning_rate": 3.174821138588191e-08,
      "loss": 0.0092,
      "num_input_tokens_seen": 24581064,
      "step": 37610
    },
    {
      "epoch": 19.714360587002098,
      "grad_norm": 0.4799582362174988,
      "learning_rate": 3.11747523434186e-08,
      "loss": 0.004,
      "num_input_tokens_seen": 24584200,
      "step": 37615
    },
    {
      "epoch": 19.71698113207547,
      "grad_norm": 0.6826513409614563,
      "learning_rate": 3.060651649434143e-08,
      "loss": 0.0045,
      "num_input_tokens_seen": 24587400,
      "step": 37620
    },
    {
      "epoch": 19.719601677148848,
      "grad_norm": 0.4262351989746094,
      "learning_rate": 3.004350395751088e-08,
      "loss": 0.0033,
      "num_input_tokens_seen": 24590088,
      "step": 37625
    },
    {
      "epoch": 19.72222222222222,
      "grad_norm": 0.13130918145179749,
      "learning_rate": 2.94857148507105e-08,
      "loss": 0.0163,
      "num_input_tokens_seen": 24593064,
      "step": 37630
    },
    {
      "epoch": 19.7248427672956,
      "grad_norm": 0.3358103930950165,
      "learning_rate": 2.893314929061641e-08,
      "loss": 0.0037,
      "num_input_tokens_seen": 24595816,
      "step": 37635
    },
    {
      "epoch": 19.72746331236897,
      "grad_norm": 1.1281322240829468,
      "learning_rate": 2.8385807392825037e-08,
      "loss": 0.0051,
      "num_input_tokens_seen": 24598824,
      "step": 37640
    },
    {
      "epoch": 19.73008385744235,
      "grad_norm": 0.23546718060970306,
      "learning_rate": 2.7843689271830897e-08,
      "loss": 0.0133,
      "num_input_tokens_seen": 24601480,
      "step": 37645
    },
    {
      "epoch": 19.732704402515722,
      "grad_norm": 0.2573755085468292,
      "learning_rate": 2.7306795041040502e-08,
      "loss": 0.0021,
      "num_input_tokens_seen": 24604296,
      "step": 37650
    },
    {
      "epoch": 19.7353249475891,
      "grad_norm": 0.44457775354385376,
      "learning_rate": 2.677512481276401e-08,
      "loss": 0.0048,
      "num_input_tokens_seen": 24607336,
      "step": 37655
    },
    {
      "epoch": 19.737945492662472,
      "grad_norm": 0.3038486838340759,
      "learning_rate": 2.6248678698220785e-08,
      "loss": 0.0101,
      "num_input_tokens_seen": 24610984,
      "step": 37660
    },
    {
      "epoch": 19.74056603773585,
      "grad_norm": 0.4293043613433838,
      "learning_rate": 2.5727456807542183e-08,
      "loss": 0.0013,
      "num_input_tokens_seen": 24613800,
      "step": 37665
    },
    {
      "epoch": 19.743186582809223,
      "grad_norm": 0.8280525803565979,
      "learning_rate": 2.5211459249754875e-08,
      "loss": 0.0042,
      "num_input_tokens_seen": 24617064,
      "step": 37670
    },
    {
      "epoch": 19.7458071278826,
      "grad_norm": 0.2146199494600296,
      "learning_rate": 2.4700686132803076e-08,
      "loss": 0.0033,
      "num_input_tokens_seen": 24621416,
      "step": 37675
    },
    {
      "epoch": 19.748427672955973,
      "grad_norm": 0.11804354190826416,
      "learning_rate": 2.4195137563534643e-08,
      "loss": 0.0185,
      "num_input_tokens_seen": 24625160,
      "step": 37680
    },
    {
      "epoch": 19.75104821802935,
      "grad_norm": 0.34254780411720276,
      "learning_rate": 2.369481364770387e-08,
      "loss": 0.0072,
      "num_input_tokens_seen": 24628520,
      "step": 37685
    },
    {
      "epoch": 19.753668763102727,
      "grad_norm": 0.16243623197078705,
      "learning_rate": 2.3199714489974266e-08,
      "loss": 0.0408,
      "num_input_tokens_seen": 24631528,
      "step": 37690
    },
    {
      "epoch": 19.7562893081761,
      "grad_norm": 0.23340649902820587,
      "learning_rate": 2.2709840193912978e-08,
      "loss": 0.0028,
      "num_input_tokens_seen": 24634440,
      "step": 37695
    },
    {
      "epoch": 19.758909853249477,
      "grad_norm": 0.1067904531955719,
      "learning_rate": 2.2225190861999146e-08,
      "loss": 0.0026,
      "num_input_tokens_seen": 24637704,
      "step": 37700
    },
    {
      "epoch": 19.76153039832285,
      "grad_norm": 0.4788852334022522,
      "learning_rate": 2.1745766595612783e-08,
      "loss": 0.0045,
      "num_input_tokens_seen": 24640776,
      "step": 37705
    },
    {
      "epoch": 19.764150943396228,
      "grad_norm": 0.07667921483516693,
      "learning_rate": 2.1271567495045886e-08,
      "loss": 0.0018,
      "num_input_tokens_seen": 24643624,
      "step": 37710
    },
    {
      "epoch": 19.7667714884696,
      "grad_norm": 2.100447177886963,
      "learning_rate": 2.0802593659491333e-08,
      "loss": 0.0074,
      "num_input_tokens_seen": 24647496,
      "step": 37715
    },
    {
      "epoch": 19.769392033542978,
      "grad_norm": 0.20699171721935272,
      "learning_rate": 2.0338845187059528e-08,
      "loss": 0.0073,
      "num_input_tokens_seen": 24651048,
      "step": 37720
    },
    {
      "epoch": 19.77201257861635,
      "grad_norm": 0.401872456073761,
      "learning_rate": 1.988032217475899e-08,
      "loss": 0.0327,
      "num_input_tokens_seen": 24653736,
      "step": 37725
    },
    {
      "epoch": 19.77463312368973,
      "grad_norm": 0.24431128799915314,
      "learning_rate": 1.942702471850466e-08,
      "loss": 0.0159,
      "num_input_tokens_seen": 24657032,
      "step": 37730
    },
    {
      "epoch": 19.7772536687631,
      "grad_norm": 0.8134015202522278,
      "learning_rate": 1.897895291312346e-08,
      "loss": 0.0061,
      "num_input_tokens_seen": 24660136,
      "step": 37735
    },
    {
      "epoch": 19.77987421383648,
      "grad_norm": 0.3568041920661926,
      "learning_rate": 1.8536106852348746e-08,
      "loss": 0.0234,
      "num_input_tokens_seen": 24663048,
      "step": 37740
    },
    {
      "epoch": 19.782494758909852,
      "grad_norm": 0.6352342367172241,
      "learning_rate": 1.8098486628814748e-08,
      "loss": 0.0146,
      "num_input_tokens_seen": 24667016,
      "step": 37745
    },
    {
      "epoch": 19.78511530398323,
      "grad_norm": 0.36260291934013367,
      "learning_rate": 1.7666092334067684e-08,
      "loss": 0.0108,
      "num_input_tokens_seen": 24670568,
      "step": 37750
    },
    {
      "epoch": 19.787735849056602,
      "grad_norm": 0.38773152232170105,
      "learning_rate": 1.7238924058562977e-08,
      "loss": 0.007,
      "num_input_tokens_seen": 24674024,
      "step": 37755
    },
    {
      "epoch": 19.79035639412998,
      "grad_norm": 0.20187117159366608,
      "learning_rate": 1.6816981891656924e-08,
      "loss": 0.0079,
      "num_input_tokens_seen": 24677320,
      "step": 37760
    },
    {
      "epoch": 19.792976939203353,
      "grad_norm": 0.26705583930015564,
      "learning_rate": 1.6400265921615032e-08,
      "loss": 0.0027,
      "num_input_tokens_seen": 24681096,
      "step": 37765
    },
    {
      "epoch": 19.79559748427673,
      "grad_norm": 0.32366856932640076,
      "learning_rate": 1.598877623560924e-08,
      "loss": 0.0028,
      "num_input_tokens_seen": 24684168,
      "step": 37770
    },
    {
      "epoch": 19.798218029350103,
      "grad_norm": 0.1367257982492447,
      "learning_rate": 1.5582512919717908e-08,
      "loss": 0.0024,
      "num_input_tokens_seen": 24687464,
      "step": 37775
    },
    {
      "epoch": 19.80083857442348,
      "grad_norm": 0.28097522258758545,
      "learning_rate": 1.5181476058928613e-08,
      "loss": 0.0265,
      "num_input_tokens_seen": 24690920,
      "step": 37780
    },
    {
      "epoch": 19.803459119496857,
      "grad_norm": 0.8140471577644348,
      "learning_rate": 1.4785665737129805e-08,
      "loss": 0.0027,
      "num_input_tokens_seen": 24693768,
      "step": 37785
    },
    {
      "epoch": 19.80607966457023,
      "grad_norm": 0.22836291790008545,
      "learning_rate": 1.4395082037127471e-08,
      "loss": 0.0102,
      "num_input_tokens_seen": 24696552,
      "step": 37790
    },
    {
      "epoch": 19.808700209643607,
      "grad_norm": 0.40534698963165283,
      "learning_rate": 1.4009725040620148e-08,
      "loss": 0.0043,
      "num_input_tokens_seen": 24699656,
      "step": 37795
    },
    {
      "epoch": 19.81132075471698,
      "grad_norm": 0.22412222623825073,
      "learning_rate": 1.3629594828223902e-08,
      "loss": 0.0103,
      "num_input_tokens_seen": 24703496,
      "step": 37800
    },
    {
      "epoch": 19.813941299790358,
      "grad_norm": 0.43334394693374634,
      "learning_rate": 1.3254691479458458e-08,
      "loss": 0.0128,
      "num_input_tokens_seen": 24706760,
      "step": 37805
    },
    {
      "epoch": 19.81656184486373,
      "grad_norm": 0.5445120334625244,
      "learning_rate": 1.2885015072747197e-08,
      "loss": 0.012,
      "num_input_tokens_seen": 24710952,
      "step": 37810
    },
    {
      "epoch": 19.819182389937108,
      "grad_norm": 0.4867153465747833,
      "learning_rate": 1.2520565685419927e-08,
      "loss": 0.0032,
      "num_input_tokens_seen": 24714088,
      "step": 37815
    },
    {
      "epoch": 19.82180293501048,
      "grad_norm": 0.3436506390571594,
      "learning_rate": 1.2161343393721214e-08,
      "loss": 0.0048,
      "num_input_tokens_seen": 24717064,
      "step": 37820
    },
    {
      "epoch": 19.82442348008386,
      "grad_norm": 0.34123146533966064,
      "learning_rate": 1.1807348272790952e-08,
      "loss": 0.0149,
      "num_input_tokens_seen": 24720712,
      "step": 37825
    },
    {
      "epoch": 19.82704402515723,
      "grad_norm": 0.2528479993343353,
      "learning_rate": 1.1458580396683794e-08,
      "loss": 0.0121,
      "num_input_tokens_seen": 24723560,
      "step": 37830
    },
    {
      "epoch": 19.82966457023061,
      "grad_norm": 0.3605547547340393,
      "learning_rate": 1.1115039838358044e-08,
      "loss": 0.0241,
      "num_input_tokens_seen": 24731656,
      "step": 37835
    },
    {
      "epoch": 19.832285115303982,
      "grad_norm": 0.1394100785255432,
      "learning_rate": 1.0776726669678439e-08,
      "loss": 0.0149,
      "num_input_tokens_seen": 24734280,
      "step": 37840
    },
    {
      "epoch": 19.83490566037736,
      "grad_norm": 0.2612559199333191,
      "learning_rate": 1.044364096141337e-08,
      "loss": 0.0063,
      "num_input_tokens_seen": 24736840,
      "step": 37845
    },
    {
      "epoch": 19.837526205450732,
      "grad_norm": 0.20393411815166473,
      "learning_rate": 1.0115782783243211e-08,
      "loss": 0.0034,
      "num_input_tokens_seen": 24740200,
      "step": 37850
    },
    {
      "epoch": 19.84014675052411,
      "grad_norm": 0.24586912989616394,
      "learning_rate": 9.793152203751988e-09,
      "loss": 0.0065,
      "num_input_tokens_seen": 24743592,
      "step": 37855
    },
    {
      "epoch": 19.842767295597483,
      "grad_norm": 0.12652327120304108,
      "learning_rate": 9.475749290430158e-09,
      "loss": 0.0114,
      "num_input_tokens_seen": 24746344,
      "step": 37860
    },
    {
      "epoch": 19.84538784067086,
      "grad_norm": 0.2790052592754364,
      "learning_rate": 9.16357410967461e-09,
      "loss": 0.0028,
      "num_input_tokens_seen": 24748712,
      "step": 37865
    },
    {
      "epoch": 19.848008385744233,
      "grad_norm": 0.25705432891845703,
      "learning_rate": 8.856626726788664e-09,
      "loss": 0.0039,
      "num_input_tokens_seen": 24752264,
      "step": 37870
    },
    {
      "epoch": 19.85062893081761,
      "grad_norm": 0.24813322722911835,
      "learning_rate": 8.554907205982066e-09,
      "loss": 0.0016,
      "num_input_tokens_seen": 24755816,
      "step": 37875
    },
    {
      "epoch": 19.853249475890987,
      "grad_norm": 0.24419040977954865,
      "learning_rate": 8.258415610373772e-09,
      "loss": 0.0044,
      "num_input_tokens_seen": 24758792,
      "step": 37880
    },
    {
      "epoch": 19.85587002096436,
      "grad_norm": 1.217791199684143,
      "learning_rate": 7.967152001980838e-09,
      "loss": 0.0092,
      "num_input_tokens_seen": 24762504,
      "step": 37885
    },
    {
      "epoch": 19.858490566037737,
      "grad_norm": 0.16553057730197906,
      "learning_rate": 7.681116441737856e-09,
      "loss": 0.0816,
      "num_input_tokens_seen": 24766760,
      "step": 37890
    },
    {
      "epoch": 19.86111111111111,
      "grad_norm": 0.31572213768959045,
      "learning_rate": 7.4003089894747464e-09,
      "loss": 0.0023,
      "num_input_tokens_seen": 24770536,
      "step": 37895
    },
    {
      "epoch": 19.863731656184488,
      "grad_norm": 0.32128310203552246,
      "learning_rate": 7.12472970393896e-09,
      "loss": 0.0016,
      "num_input_tokens_seen": 24773960,
      "step": 37900
    },
    {
      "epoch": 19.86635220125786,
      "grad_norm": 0.36885344982147217,
      "learning_rate": 6.854378642776049e-09,
      "loss": 0.0032,
      "num_input_tokens_seen": 24776744,
      "step": 37905
    },
    {
      "epoch": 19.868972746331238,
      "grad_norm": 0.16170088946819305,
      "learning_rate": 6.589255862538002e-09,
      "loss": 0.0083,
      "num_input_tokens_seen": 24781128,
      "step": 37910
    },
    {
      "epoch": 19.87159329140461,
      "grad_norm": 0.1514916718006134,
      "learning_rate": 6.32936141869156e-09,
      "loss": 0.0015,
      "num_input_tokens_seen": 24784072,
      "step": 37915
    },
    {
      "epoch": 19.87421383647799,
      "grad_norm": 1.5983796119689941,
      "learning_rate": 6.074695365596017e-09,
      "loss": 0.0198,
      "num_input_tokens_seen": 24787304,
      "step": 37920
    },
    {
      "epoch": 19.87683438155136,
      "grad_norm": 0.3017413020133972,
      "learning_rate": 5.8252577565309775e-09,
      "loss": 0.0028,
      "num_input_tokens_seen": 24790216,
      "step": 37925
    },
    {
      "epoch": 19.87945492662474,
      "grad_norm": 0.6858842968940735,
      "learning_rate": 5.581048643671371e-09,
      "loss": 0.008,
      "num_input_tokens_seen": 24794024,
      "step": 37930
    },
    {
      "epoch": 19.882075471698112,
      "grad_norm": 1.537444829940796,
      "learning_rate": 5.342068078106888e-09,
      "loss": 0.003,
      "num_input_tokens_seen": 24797032,
      "step": 37935
    },
    {
      "epoch": 19.88469601677149,
      "grad_norm": 0.2559589743614197,
      "learning_rate": 5.1083161098253175e-09,
      "loss": 0.0029,
      "num_input_tokens_seen": 24799592,
      "step": 37940
    },
    {
      "epoch": 19.887316561844862,
      "grad_norm": 0.432963490486145,
      "learning_rate": 4.87979278772921e-09,
      "loss": 0.0068,
      "num_input_tokens_seen": 24803944,
      "step": 37945
    },
    {
      "epoch": 19.88993710691824,
      "grad_norm": 0.461001992225647,
      "learning_rate": 4.656498159619216e-09,
      "loss": 0.0035,
      "num_input_tokens_seen": 24807816,
      "step": 37950
    },
    {
      "epoch": 19.892557651991613,
      "grad_norm": 0.07177762687206268,
      "learning_rate": 4.438432272210746e-09,
      "loss": 0.0031,
      "num_input_tokens_seen": 24811272,
      "step": 37955
    },
    {
      "epoch": 19.89517819706499,
      "grad_norm": 0.1887536495923996,
      "learning_rate": 4.225595171114538e-09,
      "loss": 0.0025,
      "num_input_tokens_seen": 24815048,
      "step": 37960
    },
    {
      "epoch": 19.897798742138363,
      "grad_norm": 0.2731162905693054,
      "learning_rate": 4.017986900858861e-09,
      "loss": 0.0119,
      "num_input_tokens_seen": 24818120,
      "step": 37965
    },
    {
      "epoch": 19.90041928721174,
      "grad_norm": 0.1446433663368225,
      "learning_rate": 3.815607504870089e-09,
      "loss": 0.005,
      "num_input_tokens_seen": 24822280,
      "step": 37970
    },
    {
      "epoch": 19.903039832285117,
      "grad_norm": 0.17531874775886536,
      "learning_rate": 3.618457025486577e-09,
      "loss": 0.0042,
      "num_input_tokens_seen": 24826568,
      "step": 37975
    },
    {
      "epoch": 19.90566037735849,
      "grad_norm": 0.9238464832305908,
      "learning_rate": 3.426535503947559e-09,
      "loss": 0.0039,
      "num_input_tokens_seen": 24829768,
      "step": 37980
    },
    {
      "epoch": 19.908280922431867,
      "grad_norm": 2.81716251373291,
      "learning_rate": 3.2398429803986997e-09,
      "loss": 0.0076,
      "num_input_tokens_seen": 24832392,
      "step": 37985
    },
    {
      "epoch": 19.91090146750524,
      "grad_norm": 0.4500378966331482,
      "learning_rate": 3.058379493900421e-09,
      "loss": 0.0067,
      "num_input_tokens_seen": 24834856,
      "step": 37990
    },
    {
      "epoch": 19.913522012578618,
      "grad_norm": 0.905928909778595,
      "learning_rate": 2.882145082405696e-09,
      "loss": 0.0166,
      "num_input_tokens_seen": 24838248,
      "step": 37995
    },
    {
      "epoch": 19.91614255765199,
      "grad_norm": 0.0531507171690464,
      "learning_rate": 2.711139782785033e-09,
      "loss": 0.0028,
      "num_input_tokens_seen": 24841864,
      "step": 38000
    },
    {
      "epoch": 19.918763102725368,
      "grad_norm": 0.27128151059150696,
      "learning_rate": 2.5453636308098163e-09,
      "loss": 0.003,
      "num_input_tokens_seen": 24844776,
      "step": 38005
    },
    {
      "epoch": 19.92138364779874,
      "grad_norm": 0.2424280345439911,
      "learning_rate": 2.3848166611606383e-09,
      "loss": 0.0055,
      "num_input_tokens_seen": 24848360,
      "step": 38010
    },
    {
      "epoch": 19.92400419287212,
      "grad_norm": 0.1713266223669052,
      "learning_rate": 2.2294989074161943e-09,
      "loss": 0.0086,
      "num_input_tokens_seen": 24851496,
      "step": 38015
    },
    {
      "epoch": 19.92662473794549,
      "grad_norm": 0.5410776734352112,
      "learning_rate": 2.0794104020727125e-09,
      "loss": 0.0069,
      "num_input_tokens_seen": 24855144,
      "step": 38020
    },
    {
      "epoch": 19.92924528301887,
      "grad_norm": 0.3285772204399109,
      "learning_rate": 1.9345511765272993e-09,
      "loss": 0.0056,
      "num_input_tokens_seen": 24858120,
      "step": 38025
    },
    {
      "epoch": 19.931865828092242,
      "grad_norm": 0.794620931148529,
      "learning_rate": 1.794921261077942e-09,
      "loss": 0.0052,
      "num_input_tokens_seen": 24862248,
      "step": 38030
    },
    {
      "epoch": 19.93448637316562,
      "grad_norm": 0.3323573172092438,
      "learning_rate": 1.6605206849373833e-09,
      "loss": 0.0047,
      "num_input_tokens_seen": 24864776,
      "step": 38035
    },
    {
      "epoch": 19.937106918238992,
      "grad_norm": 0.40328311920166016,
      "learning_rate": 1.531349476219246e-09,
      "loss": 0.003,
      "num_input_tokens_seen": 24868200,
      "step": 38040
    },
    {
      "epoch": 19.93972746331237,
      "grad_norm": 0.24331410229206085,
      "learning_rate": 1.4074076619463584e-09,
      "loss": 0.0038,
      "num_input_tokens_seen": 24870824,
      "step": 38045
    },
    {
      "epoch": 19.942348008385743,
      "grad_norm": 0.6631301641464233,
      "learning_rate": 1.2886952680452036e-09,
      "loss": 0.0027,
      "num_input_tokens_seen": 24874056,
      "step": 38050
    },
    {
      "epoch": 19.94496855345912,
      "grad_norm": 0.5863845944404602,
      "learning_rate": 1.1752123193459197e-09,
      "loss": 0.0097,
      "num_input_tokens_seen": 24877096,
      "step": 38055
    },
    {
      "epoch": 19.947589098532493,
      "grad_norm": 0.42329102754592896,
      "learning_rate": 1.0669588395934017e-09,
      "loss": 0.0038,
      "num_input_tokens_seen": 24880104,
      "step": 38060
    },
    {
      "epoch": 19.95020964360587,
      "grad_norm": 0.2924177944660187,
      "learning_rate": 9.639348514278722e-10,
      "loss": 0.0108,
      "num_input_tokens_seen": 24883112,
      "step": 38065
    },
    {
      "epoch": 19.952830188679247,
      "grad_norm": 0.3748258352279663,
      "learning_rate": 8.661403764043119e-10,
      "loss": 0.0074,
      "num_input_tokens_seen": 24886824,
      "step": 38070
    },
    {
      "epoch": 19.95545073375262,
      "grad_norm": 0.499105304479599,
      "learning_rate": 7.735754349785795e-10,
      "loss": 0.0029,
      "num_input_tokens_seen": 24889096,
      "step": 38075
    },
    {
      "epoch": 19.958071278825997,
      "grad_norm": 0.16933554410934448,
      "learning_rate": 6.862400465157403e-10,
      "loss": 0.0131,
      "num_input_tokens_seen": 24893640,
      "step": 38080
    },
    {
      "epoch": 19.96069182389937,
      "grad_norm": 0.808255672454834,
      "learning_rate": 6.041342292817387e-10,
      "loss": 0.0065,
      "num_input_tokens_seen": 24896264,
      "step": 38085
    },
    {
      "epoch": 19.963312368972748,
      "grad_norm": 0.3157031238079071,
      "learning_rate": 5.272580004572758e-10,
      "loss": 0.0057,
      "num_input_tokens_seen": 24899080,
      "step": 38090
    },
    {
      "epoch": 19.96593291404612,
      "grad_norm": 0.9376400113105774,
      "learning_rate": 4.556113761211567e-10,
      "loss": 0.0059,
      "num_input_tokens_seen": 24902024,
      "step": 38095
    },
    {
      "epoch": 19.968553459119498,
      "grad_norm": 0.2959236800670624,
      "learning_rate": 3.8919437125861657e-10,
      "loss": 0.0034,
      "num_input_tokens_seen": 24906088,
      "step": 38100
    },
    {
      "epoch": 19.97117400419287,
      "grad_norm": 0.3948329985141754,
      "learning_rate": 3.280069997696478e-10,
      "loss": 0.0062,
      "num_input_tokens_seen": 24908968,
      "step": 38105
    },
    {
      "epoch": 19.97379454926625,
      "grad_norm": 0.8977921605110168,
      "learning_rate": 2.720492744467951e-10,
      "loss": 0.022,
      "num_input_tokens_seen": 24912104,
      "step": 38110
    },
    {
      "epoch": 19.97641509433962,
      "grad_norm": 0.08678987622261047,
      "learning_rate": 2.2132120700291136e-10,
      "loss": 0.0049,
      "num_input_tokens_seen": 24915464,
      "step": 38115
    },
    {
      "epoch": 19.979035639413,
      "grad_norm": 1.133821725845337,
      "learning_rate": 1.7582280804340213e-10,
      "loss": 0.0032,
      "num_input_tokens_seen": 24918632,
      "step": 38120
    },
    {
      "epoch": 19.981656184486372,
      "grad_norm": 1.1083035469055176,
      "learning_rate": 1.3555408708842977e-10,
      "loss": 0.0043,
      "num_input_tokens_seen": 24921032,
      "step": 38125
    },
    {
      "epoch": 19.98427672955975,
      "grad_norm": 0.1886010468006134,
      "learning_rate": 1.0051505256458704e-10,
      "loss": 0.0017,
      "num_input_tokens_seen": 24923912,
      "step": 38130
    },
    {
      "epoch": 19.986897274633122,
      "grad_norm": 0.4642328917980194,
      "learning_rate": 7.070571179657037e-11,
      "loss": 0.0067,
      "num_input_tokens_seen": 24926952,
      "step": 38135
    },
    {
      "epoch": 19.9895178197065,
      "grad_norm": 0.8514014482498169,
      "learning_rate": 4.612607102383315e-11,
      "loss": 0.0068,
      "num_input_tokens_seen": 24930984,
      "step": 38140
    },
    {
      "epoch": 19.992138364779873,
      "grad_norm": 0.80704665184021,
      "learning_rate": 2.677613538670798e-11,
      "loss": 0.0053,
      "num_input_tokens_seen": 24933832,
      "step": 38145
    },
    {
      "epoch": 19.99475890985325,
      "grad_norm": 0.7158359885215759,
      "learning_rate": 1.2655908931957782e-11,
      "loss": 0.003,
      "num_input_tokens_seen": 24936584,
      "step": 38150
    },
    {
      "epoch": 19.997379454926623,
      "grad_norm": 0.07662786543369293,
      "learning_rate": 3.765394615551366e-12,
      "loss": 0.0028,
      "num_input_tokens_seen": 24939432,
      "step": 38155
    },
    {
      "epoch": 20.0,
      "grad_norm": 0.2535932660102844,
      "learning_rate": 1.0459429711229262e-13,
      "loss": 0.0408,
      "num_input_tokens_seen": 24941912,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "eval_loss": 1.629987120628357,
      "eval_runtime": 11.7534,
      "eval_samples_per_second": 72.149,
      "eval_steps_per_second": 18.037,
      "num_input_tokens_seen": 24941912,
      "step": 38160
    },
    {
      "epoch": 20.0,
      "num_input_tokens_seen": 24941912,
      "step": 38160,
      "total_flos": 1.1253199176005714e+18,
      "train_loss": 0.18923116609550802,
      "train_runtime": 8804.0767,
      "train_samples_per_second": 17.331,
      "train_steps_per_second": 4.334
    }
  ],
  "logging_steps": 5,
  "max_steps": 38160,
  "num_input_tokens_seen": 24941912,
  "num_train_epochs": 20,
  "save_steps": 1908,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1253199176005714e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}