{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.0, "eval_steps": 500, "global_step": 1575, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0019065776930409914, "grad_norm": 17.3579213184435, "learning_rate": 0.0, "loss": 4.3981, "step": 1 }, { "epoch": 0.0038131553860819827, "grad_norm": 17.90962181746105, "learning_rate": 6.329113924050633e-08, "loss": 4.3929, "step": 2 }, { "epoch": 0.005719733079122974, "grad_norm": 15.853496486640772, "learning_rate": 1.2658227848101266e-07, "loss": 4.2726, "step": 3 }, { "epoch": 0.0076263107721639654, "grad_norm": 18.616149923548736, "learning_rate": 1.89873417721519e-07, "loss": 4.6267, "step": 4 }, { "epoch": 0.009532888465204958, "grad_norm": 15.329949451228464, "learning_rate": 2.5316455696202533e-07, "loss": 4.2601, "step": 5 }, { "epoch": 0.011439466158245948, "grad_norm": 17.269413612762943, "learning_rate": 3.164556962025317e-07, "loss": 4.5666, "step": 6 }, { "epoch": 0.01334604385128694, "grad_norm": 16.423266229880248, "learning_rate": 3.79746835443038e-07, "loss": 4.3638, "step": 7 }, { "epoch": 0.015252621544327931, "grad_norm": 16.152685297099758, "learning_rate": 4.4303797468354435e-07, "loss": 4.3427, "step": 8 }, { "epoch": 0.017159199237368923, "grad_norm": 16.576169926701326, "learning_rate": 5.063291139240507e-07, "loss": 4.356, "step": 9 }, { "epoch": 0.019065776930409915, "grad_norm": 16.38368367449629, "learning_rate": 5.69620253164557e-07, "loss": 4.3008, "step": 10 }, { "epoch": 0.020972354623450904, "grad_norm": 16.121484737493308, "learning_rate": 6.329113924050634e-07, "loss": 4.3412, "step": 11 }, { "epoch": 0.022878932316491896, "grad_norm": 16.271252548841257, "learning_rate": 6.962025316455696e-07, "loss": 4.2293, "step": 12 }, { "epoch": 0.02478551000953289, "grad_norm": 16.19703086887933, "learning_rate": 7.59493670886076e-07, "loss": 4.3037, "step": 13 }, { "epoch": 0.02669208770257388, "grad_norm": 17.08020017617865, "learning_rate": 8.227848101265823e-07, "loss": 4.3703, "step": 14 }, { "epoch": 0.028598665395614873, "grad_norm": 16.428588811077084, "learning_rate": 8.860759493670887e-07, "loss": 4.446, "step": 15 }, { "epoch": 0.030505243088655862, "grad_norm": 16.76220806874717, "learning_rate": 9.493670886075951e-07, "loss": 4.2385, "step": 16 }, { "epoch": 0.032411820781696854, "grad_norm": 16.396276400292464, "learning_rate": 1.0126582278481013e-06, "loss": 4.3383, "step": 17 }, { "epoch": 0.034318398474737846, "grad_norm": 16.022367674791376, "learning_rate": 1.0759493670886077e-06, "loss": 4.2976, "step": 18 }, { "epoch": 0.03622497616777884, "grad_norm": 14.615409589985376, "learning_rate": 1.139240506329114e-06, "loss": 4.2518, "step": 19 }, { "epoch": 0.03813155386081983, "grad_norm": 15.453484674676906, "learning_rate": 1.2025316455696204e-06, "loss": 4.3496, "step": 20 }, { "epoch": 0.04003813155386082, "grad_norm": 16.120758319551253, "learning_rate": 1.2658227848101267e-06, "loss": 4.3757, "step": 21 }, { "epoch": 0.04194470924690181, "grad_norm": 14.814743516282718, "learning_rate": 1.3291139240506329e-06, "loss": 4.1737, "step": 22 }, { "epoch": 0.0438512869399428, "grad_norm": 14.76799262048467, "learning_rate": 1.3924050632911392e-06, "loss": 4.2685, "step": 23 }, { "epoch": 0.04575786463298379, "grad_norm": 13.106451410532546, "learning_rate": 1.4556962025316456e-06, "loss": 4.1796, "step": 24 }, { "epoch": 0.047664442326024785, "grad_norm": 14.508606089057851, "learning_rate": 1.518987341772152e-06, "loss": 4.1135, "step": 25 }, { "epoch": 0.04957102001906578, "grad_norm": 14.598838059669685, "learning_rate": 1.5822784810126585e-06, "loss": 4.2128, "step": 26 }, { "epoch": 0.05147759771210677, "grad_norm": 12.630113135826011, "learning_rate": 1.6455696202531647e-06, "loss": 3.9401, "step": 27 }, { "epoch": 0.05338417540514776, "grad_norm": 12.520456299950732, "learning_rate": 1.708860759493671e-06, "loss": 3.9678, "step": 28 }, { "epoch": 0.055290753098188754, "grad_norm": 12.185114039824388, "learning_rate": 1.7721518987341774e-06, "loss": 4.0077, "step": 29 }, { "epoch": 0.057197330791229746, "grad_norm": 12.51242672233334, "learning_rate": 1.8354430379746838e-06, "loss": 4.0827, "step": 30 }, { "epoch": 0.05910390848427073, "grad_norm": 12.101344008153692, "learning_rate": 1.8987341772151901e-06, "loss": 3.9888, "step": 31 }, { "epoch": 0.061010486177311724, "grad_norm": 10.17878937547436, "learning_rate": 1.9620253164556965e-06, "loss": 3.7357, "step": 32 }, { "epoch": 0.06291706387035272, "grad_norm": 10.491773492079057, "learning_rate": 2.0253164556962026e-06, "loss": 3.97, "step": 33 }, { "epoch": 0.06482364156339371, "grad_norm": 11.137460937899041, "learning_rate": 2.088607594936709e-06, "loss": 3.9309, "step": 34 }, { "epoch": 0.0667302192564347, "grad_norm": 8.31518226227282, "learning_rate": 2.1518987341772153e-06, "loss": 3.9813, "step": 35 }, { "epoch": 0.06863679694947569, "grad_norm": 9.982734441400883, "learning_rate": 2.2151898734177215e-06, "loss": 3.7861, "step": 36 }, { "epoch": 0.07054337464251668, "grad_norm": 8.431998063340147, "learning_rate": 2.278481012658228e-06, "loss": 3.8552, "step": 37 }, { "epoch": 0.07244995233555768, "grad_norm": 8.128468512922282, "learning_rate": 2.341772151898734e-06, "loss": 3.7141, "step": 38 }, { "epoch": 0.07435653002859867, "grad_norm": 8.783979176278532, "learning_rate": 2.4050632911392408e-06, "loss": 3.7681, "step": 39 }, { "epoch": 0.07626310772163966, "grad_norm": 8.280586169356493, "learning_rate": 2.4683544303797473e-06, "loss": 3.6815, "step": 40 }, { "epoch": 0.07816968541468065, "grad_norm": 9.030445399973246, "learning_rate": 2.5316455696202535e-06, "loss": 3.7874, "step": 41 }, { "epoch": 0.08007626310772165, "grad_norm": 8.612584956323863, "learning_rate": 2.5949367088607596e-06, "loss": 3.6955, "step": 42 }, { "epoch": 0.08198284080076262, "grad_norm": 13.060783025627266, "learning_rate": 2.6582278481012658e-06, "loss": 3.7036, "step": 43 }, { "epoch": 0.08388941849380362, "grad_norm": 8.412239834544044, "learning_rate": 2.7215189873417724e-06, "loss": 3.654, "step": 44 }, { "epoch": 0.08579599618684461, "grad_norm": 9.113089990426985, "learning_rate": 2.7848101265822785e-06, "loss": 3.989, "step": 45 }, { "epoch": 0.0877025738798856, "grad_norm": 10.553481155001158, "learning_rate": 2.848101265822785e-06, "loss": 3.7476, "step": 46 }, { "epoch": 0.08960915157292659, "grad_norm": 8.119832796406197, "learning_rate": 2.9113924050632912e-06, "loss": 3.6488, "step": 47 }, { "epoch": 0.09151572926596759, "grad_norm": 7.807003344063545, "learning_rate": 2.9746835443037974e-06, "loss": 3.5814, "step": 48 }, { "epoch": 0.09342230695900858, "grad_norm": 8.473165927022706, "learning_rate": 3.037974683544304e-06, "loss": 3.6174, "step": 49 }, { "epoch": 0.09532888465204957, "grad_norm": 7.468219919118173, "learning_rate": 3.10126582278481e-06, "loss": 3.6454, "step": 50 }, { "epoch": 0.09723546234509056, "grad_norm": 7.174639070655079, "learning_rate": 3.164556962025317e-06, "loss": 3.6853, "step": 51 }, { "epoch": 0.09914204003813155, "grad_norm": 7.609006172826824, "learning_rate": 3.2278481012658232e-06, "loss": 3.6356, "step": 52 }, { "epoch": 0.10104861773117255, "grad_norm": 7.291837559368081, "learning_rate": 3.2911392405063294e-06, "loss": 3.5283, "step": 53 }, { "epoch": 0.10295519542421354, "grad_norm": 6.780388195867472, "learning_rate": 3.354430379746836e-06, "loss": 3.4469, "step": 54 }, { "epoch": 0.10486177311725453, "grad_norm": 7.3501648910030255, "learning_rate": 3.417721518987342e-06, "loss": 3.6395, "step": 55 }, { "epoch": 0.10676835081029552, "grad_norm": 7.018157547889191, "learning_rate": 3.4810126582278487e-06, "loss": 3.5147, "step": 56 }, { "epoch": 0.10867492850333652, "grad_norm": 7.546002506187327, "learning_rate": 3.544303797468355e-06, "loss": 3.6704, "step": 57 }, { "epoch": 0.11058150619637751, "grad_norm": 7.555078692302205, "learning_rate": 3.607594936708861e-06, "loss": 3.7882, "step": 58 }, { "epoch": 0.1124880838894185, "grad_norm": 6.622995201910215, "learning_rate": 3.6708860759493675e-06, "loss": 3.5018, "step": 59 }, { "epoch": 0.11439466158245949, "grad_norm": 6.795950630109369, "learning_rate": 3.7341772151898737e-06, "loss": 3.4035, "step": 60 }, { "epoch": 0.11630123927550047, "grad_norm": 6.618931220298999, "learning_rate": 3.7974683544303802e-06, "loss": 3.6038, "step": 61 }, { "epoch": 0.11820781696854146, "grad_norm": 6.3526205556559, "learning_rate": 3.860759493670886e-06, "loss": 3.562, "step": 62 }, { "epoch": 0.12011439466158245, "grad_norm": 6.325055761694145, "learning_rate": 3.924050632911393e-06, "loss": 3.3324, "step": 63 }, { "epoch": 0.12202097235462345, "grad_norm": 6.406230854395011, "learning_rate": 3.9873417721518995e-06, "loss": 3.5668, "step": 64 }, { "epoch": 0.12392755004766444, "grad_norm": 6.56802058258271, "learning_rate": 4.050632911392405e-06, "loss": 3.5915, "step": 65 }, { "epoch": 0.12583412774070543, "grad_norm": 6.663617594643165, "learning_rate": 4.113924050632912e-06, "loss": 3.5524, "step": 66 }, { "epoch": 0.12774070543374644, "grad_norm": 6.591202852890635, "learning_rate": 4.177215189873418e-06, "loss": 3.5956, "step": 67 }, { "epoch": 0.12964728312678742, "grad_norm": 5.9774373498605735, "learning_rate": 4.240506329113924e-06, "loss": 3.5193, "step": 68 }, { "epoch": 0.1315538608198284, "grad_norm": 6.261478997047101, "learning_rate": 4.303797468354431e-06, "loss": 3.4833, "step": 69 }, { "epoch": 0.1334604385128694, "grad_norm": 6.159968297186925, "learning_rate": 4.367088607594937e-06, "loss": 3.4184, "step": 70 }, { "epoch": 0.13536701620591038, "grad_norm": 6.23299688638621, "learning_rate": 4.430379746835443e-06, "loss": 3.5988, "step": 71 }, { "epoch": 0.13727359389895138, "grad_norm": 6.413529407925796, "learning_rate": 4.4936708860759495e-06, "loss": 3.5748, "step": 72 }, { "epoch": 0.13918017159199236, "grad_norm": 6.319664708213027, "learning_rate": 4.556962025316456e-06, "loss": 3.456, "step": 73 }, { "epoch": 0.14108674928503337, "grad_norm": 6.455163450997423, "learning_rate": 4.620253164556963e-06, "loss": 3.5894, "step": 74 }, { "epoch": 0.14299332697807435, "grad_norm": 6.620403687552936, "learning_rate": 4.683544303797468e-06, "loss": 3.3576, "step": 75 }, { "epoch": 0.14489990467111535, "grad_norm": 6.1298620382201445, "learning_rate": 4.746835443037975e-06, "loss": 3.5176, "step": 76 }, { "epoch": 0.14680648236415633, "grad_norm": 6.080308849269763, "learning_rate": 4.8101265822784815e-06, "loss": 3.6558, "step": 77 }, { "epoch": 0.14871306005719734, "grad_norm": 5.7438410447058725, "learning_rate": 4.873417721518987e-06, "loss": 3.4243, "step": 78 }, { "epoch": 0.15061963775023832, "grad_norm": 6.341235766908976, "learning_rate": 4.936708860759495e-06, "loss": 3.4782, "step": 79 }, { "epoch": 0.15252621544327932, "grad_norm": 6.451130584617015, "learning_rate": 5e-06, "loss": 3.4834, "step": 80 }, { "epoch": 0.1544327931363203, "grad_norm": 6.579622670427806, "learning_rate": 5.063291139240507e-06, "loss": 3.5267, "step": 81 }, { "epoch": 0.1563393708293613, "grad_norm": 6.3063142200405835, "learning_rate": 5.126582278481013e-06, "loss": 3.5069, "step": 82 }, { "epoch": 0.15824594852240229, "grad_norm": 5.59367893050122, "learning_rate": 5.189873417721519e-06, "loss": 3.3871, "step": 83 }, { "epoch": 0.1601525262154433, "grad_norm": 5.981161856469303, "learning_rate": 5.253164556962026e-06, "loss": 3.4395, "step": 84 }, { "epoch": 0.16205910390848427, "grad_norm": 5.8642047300711555, "learning_rate": 5.3164556962025316e-06, "loss": 3.4745, "step": 85 }, { "epoch": 0.16396568160152525, "grad_norm": 7.820208884857309, "learning_rate": 5.379746835443038e-06, "loss": 3.4555, "step": 86 }, { "epoch": 0.16587225929456625, "grad_norm": 6.408567983821371, "learning_rate": 5.443037974683545e-06, "loss": 3.5786, "step": 87 }, { "epoch": 0.16777883698760723, "grad_norm": 5.631033128333004, "learning_rate": 5.506329113924051e-06, "loss": 3.4328, "step": 88 }, { "epoch": 0.16968541468064824, "grad_norm": 5.7601521861325, "learning_rate": 5.569620253164557e-06, "loss": 3.3946, "step": 89 }, { "epoch": 0.17159199237368922, "grad_norm": 6.037200774807998, "learning_rate": 5.6329113924050636e-06, "loss": 3.4503, "step": 90 }, { "epoch": 0.17349857006673022, "grad_norm": 5.803262456635073, "learning_rate": 5.69620253164557e-06, "loss": 3.4889, "step": 91 }, { "epoch": 0.1754051477597712, "grad_norm": 6.152022954431213, "learning_rate": 5.759493670886076e-06, "loss": 3.3851, "step": 92 }, { "epoch": 0.1773117254528122, "grad_norm": 5.634031839993615, "learning_rate": 5.8227848101265824e-06, "loss": 3.3532, "step": 93 }, { "epoch": 0.17921830314585319, "grad_norm": 5.583971458574191, "learning_rate": 5.886075949367089e-06, "loss": 3.3878, "step": 94 }, { "epoch": 0.1811248808388942, "grad_norm": 5.954344970261813, "learning_rate": 5.949367088607595e-06, "loss": 3.3169, "step": 95 }, { "epoch": 0.18303145853193517, "grad_norm": 6.177213109772923, "learning_rate": 6.012658227848101e-06, "loss": 3.4694, "step": 96 }, { "epoch": 0.18493803622497618, "grad_norm": 5.685170155827507, "learning_rate": 6.075949367088608e-06, "loss": 3.5685, "step": 97 }, { "epoch": 0.18684461391801716, "grad_norm": 5.793914025240259, "learning_rate": 6.139240506329115e-06, "loss": 3.4196, "step": 98 }, { "epoch": 0.18875119161105816, "grad_norm": 6.165547254998855, "learning_rate": 6.20253164556962e-06, "loss": 3.5488, "step": 99 }, { "epoch": 0.19065776930409914, "grad_norm": 5.8259858366228565, "learning_rate": 6.265822784810128e-06, "loss": 3.4873, "step": 100 }, { "epoch": 0.19256434699714015, "grad_norm": 5.7703322561182135, "learning_rate": 6.329113924050634e-06, "loss": 3.38, "step": 101 }, { "epoch": 0.19447092469018112, "grad_norm": 5.811920650397745, "learning_rate": 6.392405063291139e-06, "loss": 3.5567, "step": 102 }, { "epoch": 0.19637750238322213, "grad_norm": 5.7006499489458236, "learning_rate": 6.4556962025316464e-06, "loss": 3.5391, "step": 103 }, { "epoch": 0.1982840800762631, "grad_norm": 5.335594399446883, "learning_rate": 6.518987341772153e-06, "loss": 3.3219, "step": 104 }, { "epoch": 0.2001906577693041, "grad_norm": 6.050129312227366, "learning_rate": 6.582278481012659e-06, "loss": 3.4023, "step": 105 }, { "epoch": 0.2020972354623451, "grad_norm": 6.829445090955656, "learning_rate": 6.645569620253165e-06, "loss": 3.5391, "step": 106 }, { "epoch": 0.20400381315538607, "grad_norm": 8.401310658356655, "learning_rate": 6.708860759493672e-06, "loss": 3.393, "step": 107 }, { "epoch": 0.20591039084842708, "grad_norm": 6.020497276051294, "learning_rate": 6.772151898734178e-06, "loss": 3.2307, "step": 108 }, { "epoch": 0.20781696854146806, "grad_norm": 5.656624129438986, "learning_rate": 6.835443037974684e-06, "loss": 3.4658, "step": 109 }, { "epoch": 0.20972354623450906, "grad_norm": 6.126143558100418, "learning_rate": 6.898734177215191e-06, "loss": 3.3892, "step": 110 }, { "epoch": 0.21163012392755004, "grad_norm": 5.7784413806664325, "learning_rate": 6.962025316455697e-06, "loss": 3.2375, "step": 111 }, { "epoch": 0.21353670162059105, "grad_norm": 6.8364629460621655, "learning_rate": 7.025316455696203e-06, "loss": 3.4938, "step": 112 }, { "epoch": 0.21544327931363202, "grad_norm": 5.689658614728312, "learning_rate": 7.08860759493671e-06, "loss": 3.5292, "step": 113 }, { "epoch": 0.21734985700667303, "grad_norm": 5.882627158035053, "learning_rate": 7.151898734177216e-06, "loss": 3.3882, "step": 114 }, { "epoch": 0.219256434699714, "grad_norm": 5.84520647086525, "learning_rate": 7.215189873417722e-06, "loss": 3.491, "step": 115 }, { "epoch": 0.22116301239275502, "grad_norm": 5.694302807252288, "learning_rate": 7.2784810126582285e-06, "loss": 3.324, "step": 116 }, { "epoch": 0.223069590085796, "grad_norm": 5.107303182267903, "learning_rate": 7.341772151898735e-06, "loss": 3.4687, "step": 117 }, { "epoch": 0.224976167778837, "grad_norm": 5.818987592699087, "learning_rate": 7.405063291139241e-06, "loss": 3.4988, "step": 118 }, { "epoch": 0.22688274547187798, "grad_norm": 6.428543639310683, "learning_rate": 7.468354430379747e-06, "loss": 3.4768, "step": 119 }, { "epoch": 0.22878932316491898, "grad_norm": 5.9308623932430065, "learning_rate": 7.531645569620254e-06, "loss": 3.4424, "step": 120 }, { "epoch": 0.23069590085795996, "grad_norm": 5.484681204150636, "learning_rate": 7.5949367088607605e-06, "loss": 3.244, "step": 121 }, { "epoch": 0.23260247855100094, "grad_norm": 6.099887809078382, "learning_rate": 7.658227848101265e-06, "loss": 3.5535, "step": 122 }, { "epoch": 0.23450905624404195, "grad_norm": 5.637027332824803, "learning_rate": 7.721518987341773e-06, "loss": 3.4404, "step": 123 }, { "epoch": 0.23641563393708293, "grad_norm": 5.833388998055909, "learning_rate": 7.78481012658228e-06, "loss": 3.4847, "step": 124 }, { "epoch": 0.23832221163012393, "grad_norm": 5.726952334703264, "learning_rate": 7.848101265822786e-06, "loss": 3.4677, "step": 125 }, { "epoch": 0.2402287893231649, "grad_norm": 6.145940503627187, "learning_rate": 7.911392405063292e-06, "loss": 3.5078, "step": 126 }, { "epoch": 0.24213536701620592, "grad_norm": 5.952043106359386, "learning_rate": 7.974683544303799e-06, "loss": 3.5256, "step": 127 }, { "epoch": 0.2440419447092469, "grad_norm": 5.48605086069196, "learning_rate": 8.037974683544305e-06, "loss": 3.4695, "step": 128 }, { "epoch": 0.2459485224022879, "grad_norm": 6.2007142222227625, "learning_rate": 8.10126582278481e-06, "loss": 3.3211, "step": 129 }, { "epoch": 0.24785510009532888, "grad_norm": 5.323362359500249, "learning_rate": 8.164556962025318e-06, "loss": 3.2878, "step": 130 }, { "epoch": 0.24976167778836988, "grad_norm": 5.759231575434587, "learning_rate": 8.227848101265824e-06, "loss": 3.5145, "step": 131 }, { "epoch": 0.25166825548141086, "grad_norm": 5.860756781853601, "learning_rate": 8.29113924050633e-06, "loss": 3.4893, "step": 132 }, { "epoch": 0.25357483317445184, "grad_norm": 6.072290143422754, "learning_rate": 8.354430379746837e-06, "loss": 3.4224, "step": 133 }, { "epoch": 0.2554814108674929, "grad_norm": 5.660396506814077, "learning_rate": 8.417721518987342e-06, "loss": 3.315, "step": 134 }, { "epoch": 0.25738798856053385, "grad_norm": 5.621077948448685, "learning_rate": 8.481012658227848e-06, "loss": 3.2896, "step": 135 }, { "epoch": 0.25929456625357483, "grad_norm": 6.163178742368489, "learning_rate": 8.544303797468356e-06, "loss": 3.4167, "step": 136 }, { "epoch": 0.2612011439466158, "grad_norm": 6.072707121001799, "learning_rate": 8.607594936708861e-06, "loss": 3.6158, "step": 137 }, { "epoch": 0.2631077216396568, "grad_norm": 5.236539829676567, "learning_rate": 8.670886075949367e-06, "loss": 3.4655, "step": 138 }, { "epoch": 0.2650142993326978, "grad_norm": 5.738963030706353, "learning_rate": 8.734177215189874e-06, "loss": 3.4801, "step": 139 }, { "epoch": 0.2669208770257388, "grad_norm": 5.479543195153456, "learning_rate": 8.79746835443038e-06, "loss": 3.3512, "step": 140 }, { "epoch": 0.2688274547187798, "grad_norm": 5.922641215304641, "learning_rate": 8.860759493670886e-06, "loss": 3.5754, "step": 141 }, { "epoch": 0.27073403241182076, "grad_norm": 5.556900729896823, "learning_rate": 8.924050632911393e-06, "loss": 3.4915, "step": 142 }, { "epoch": 0.2726406101048618, "grad_norm": 6.010924106662949, "learning_rate": 8.987341772151899e-06, "loss": 3.3577, "step": 143 }, { "epoch": 0.27454718779790277, "grad_norm": 6.0038886604808495, "learning_rate": 9.050632911392407e-06, "loss": 3.2821, "step": 144 }, { "epoch": 0.27645376549094375, "grad_norm": 6.311937071605589, "learning_rate": 9.113924050632912e-06, "loss": 3.3382, "step": 145 }, { "epoch": 0.2783603431839847, "grad_norm": 5.329916084527985, "learning_rate": 9.177215189873418e-06, "loss": 3.3967, "step": 146 }, { "epoch": 0.28026692087702576, "grad_norm": 5.690073587109105, "learning_rate": 9.240506329113925e-06, "loss": 3.4967, "step": 147 }, { "epoch": 0.28217349857006674, "grad_norm": 5.676396915272239, "learning_rate": 9.303797468354431e-06, "loss": 3.2648, "step": 148 }, { "epoch": 0.2840800762631077, "grad_norm": 5.958406772727416, "learning_rate": 9.367088607594937e-06, "loss": 3.3294, "step": 149 }, { "epoch": 0.2859866539561487, "grad_norm": 5.767745227063881, "learning_rate": 9.430379746835444e-06, "loss": 3.2463, "step": 150 }, { "epoch": 0.28789323164918973, "grad_norm": 6.3173487391492795, "learning_rate": 9.49367088607595e-06, "loss": 3.4643, "step": 151 }, { "epoch": 0.2897998093422307, "grad_norm": 5.300769300071107, "learning_rate": 9.556962025316456e-06, "loss": 3.2682, "step": 152 }, { "epoch": 0.2917063870352717, "grad_norm": 5.51927831811083, "learning_rate": 9.620253164556963e-06, "loss": 3.4519, "step": 153 }, { "epoch": 0.29361296472831266, "grad_norm": 5.82083955776914, "learning_rate": 9.68354430379747e-06, "loss": 3.2912, "step": 154 }, { "epoch": 0.29551954242135364, "grad_norm": 5.228725361309231, "learning_rate": 9.746835443037975e-06, "loss": 3.3896, "step": 155 }, { "epoch": 0.2974261201143947, "grad_norm": 5.721472364650993, "learning_rate": 9.810126582278482e-06, "loss": 3.3877, "step": 156 }, { "epoch": 0.29933269780743565, "grad_norm": 5.5375195997756315, "learning_rate": 9.87341772151899e-06, "loss": 3.5309, "step": 157 }, { "epoch": 0.30123927550047663, "grad_norm": 5.741239251712303, "learning_rate": 9.936708860759493e-06, "loss": 3.4758, "step": 158 }, { "epoch": 0.3031458531935176, "grad_norm": 5.92358469629776, "learning_rate": 1e-05, "loss": 3.4039, "step": 159 }, { "epoch": 0.30505243088655865, "grad_norm": 5.74389849841475, "learning_rate": 9.999987711471653e-06, "loss": 3.4956, "step": 160 }, { "epoch": 0.3069590085795996, "grad_norm": 5.460186151638466, "learning_rate": 9.999950845947013e-06, "loss": 3.2716, "step": 161 }, { "epoch": 0.3088655862726406, "grad_norm": 6.4170161123261655, "learning_rate": 9.999889403607288e-06, "loss": 3.2515, "step": 162 }, { "epoch": 0.3107721639656816, "grad_norm": 6.662941222230564, "learning_rate": 9.999803384754496e-06, "loss": 3.4254, "step": 163 }, { "epoch": 0.3126787416587226, "grad_norm": 5.479579977412246, "learning_rate": 9.999692789811453e-06, "loss": 3.3131, "step": 164 }, { "epoch": 0.3145853193517636, "grad_norm": 5.603207921110108, "learning_rate": 9.999557619321777e-06, "loss": 3.4606, "step": 165 }, { "epoch": 0.31649189704480457, "grad_norm": 7.364356563059311, "learning_rate": 9.99939787394989e-06, "loss": 3.6557, "step": 166 }, { "epoch": 0.31839847473784555, "grad_norm": 5.860112962619253, "learning_rate": 9.999213554481006e-06, "loss": 3.3046, "step": 167 }, { "epoch": 0.3203050524308866, "grad_norm": 6.450466498372151, "learning_rate": 9.999004661821127e-06, "loss": 3.4876, "step": 168 }, { "epoch": 0.32221163012392756, "grad_norm": 5.922032547206523, "learning_rate": 9.99877119699705e-06, "loss": 3.3569, "step": 169 }, { "epoch": 0.32411820781696854, "grad_norm": 5.534845582982128, "learning_rate": 9.99851316115635e-06, "loss": 3.467, "step": 170 }, { "epoch": 0.3260247855100095, "grad_norm": 5.626710896508691, "learning_rate": 9.998230555567376e-06, "loss": 3.3399, "step": 171 }, { "epoch": 0.3279313632030505, "grad_norm": 5.571836944265983, "learning_rate": 9.997923381619257e-06, "loss": 3.3097, "step": 172 }, { "epoch": 0.32983794089609153, "grad_norm": 6.54604992019215, "learning_rate": 9.997591640821874e-06, "loss": 3.4222, "step": 173 }, { "epoch": 0.3317445185891325, "grad_norm": 6.10955574436862, "learning_rate": 9.997235334805873e-06, "loss": 3.3108, "step": 174 }, { "epoch": 0.3336510962821735, "grad_norm": 6.764988535428428, "learning_rate": 9.99685446532264e-06, "loss": 3.2327, "step": 175 }, { "epoch": 0.33555767397521447, "grad_norm": 6.3752901383036615, "learning_rate": 9.996449034244311e-06, "loss": 3.6006, "step": 176 }, { "epoch": 0.3374642516682555, "grad_norm": 6.0455167155749985, "learning_rate": 9.996019043563742e-06, "loss": 3.4197, "step": 177 }, { "epoch": 0.3393708293612965, "grad_norm": 5.812481830979919, "learning_rate": 9.995564495394517e-06, "loss": 3.3515, "step": 178 }, { "epoch": 0.34127740705433746, "grad_norm": 5.710322222400245, "learning_rate": 9.995085391970925e-06, "loss": 3.4744, "step": 179 }, { "epoch": 0.34318398474737843, "grad_norm": 6.586665063517891, "learning_rate": 9.994581735647958e-06, "loss": 3.2798, "step": 180 }, { "epoch": 0.34509056244041947, "grad_norm": 5.467627036918975, "learning_rate": 9.994053528901294e-06, "loss": 3.358, "step": 181 }, { "epoch": 0.34699714013346045, "grad_norm": 5.873410734003397, "learning_rate": 9.993500774327283e-06, "loss": 3.3463, "step": 182 }, { "epoch": 0.3489037178265014, "grad_norm": 5.621654427015682, "learning_rate": 9.992923474642945e-06, "loss": 3.3221, "step": 183 }, { "epoch": 0.3508102955195424, "grad_norm": 5.535494829898761, "learning_rate": 9.992321632685946e-06, "loss": 3.2349, "step": 184 }, { "epoch": 0.35271687321258344, "grad_norm": 5.782543404549457, "learning_rate": 9.991695251414584e-06, "loss": 3.3171, "step": 185 }, { "epoch": 0.3546234509056244, "grad_norm": 6.019641838382528, "learning_rate": 9.991044333907782e-06, "loss": 3.6293, "step": 186 }, { "epoch": 0.3565300285986654, "grad_norm": 5.351523481742335, "learning_rate": 9.990368883365066e-06, "loss": 3.2905, "step": 187 }, { "epoch": 0.35843660629170637, "grad_norm": 5.657818518227769, "learning_rate": 9.989668903106555e-06, "loss": 3.4371, "step": 188 }, { "epoch": 0.3603431839847474, "grad_norm": 5.791112419771409, "learning_rate": 9.988944396572939e-06, "loss": 3.4027, "step": 189 }, { "epoch": 0.3622497616777884, "grad_norm": 5.7380410883537465, "learning_rate": 9.988195367325467e-06, "loss": 3.3519, "step": 190 }, { "epoch": 0.36415633937082936, "grad_norm": 6.00764678564817, "learning_rate": 9.987421819045923e-06, "loss": 3.4783, "step": 191 }, { "epoch": 0.36606291706387034, "grad_norm": 5.79950405816056, "learning_rate": 9.986623755536618e-06, "loss": 3.3799, "step": 192 }, { "epoch": 0.3679694947569113, "grad_norm": 6.1226100214398524, "learning_rate": 9.985801180720359e-06, "loss": 3.43, "step": 193 }, { "epoch": 0.36987607244995235, "grad_norm": 5.891470803445207, "learning_rate": 9.984954098640442e-06, "loss": 3.3824, "step": 194 }, { "epoch": 0.37178265014299333, "grad_norm": 6.2137025292485495, "learning_rate": 9.984082513460624e-06, "loss": 3.4992, "step": 195 }, { "epoch": 0.3736892278360343, "grad_norm": 5.696211807445619, "learning_rate": 9.983186429465106e-06, "loss": 3.4783, "step": 196 }, { "epoch": 0.3755958055290753, "grad_norm": 5.130677290320459, "learning_rate": 9.982265851058505e-06, "loss": 3.3914, "step": 197 }, { "epoch": 0.3775023832221163, "grad_norm": 5.885063642898856, "learning_rate": 9.981320782765847e-06, "loss": 3.2562, "step": 198 }, { "epoch": 0.3794089609151573, "grad_norm": 5.451378895226174, "learning_rate": 9.980351229232528e-06, "loss": 3.3897, "step": 199 }, { "epoch": 0.3813155386081983, "grad_norm": 6.100490230015147, "learning_rate": 9.979357195224303e-06, "loss": 3.3779, "step": 200 }, { "epoch": 0.38322211630123926, "grad_norm": 5.3361827071437125, "learning_rate": 9.978338685627263e-06, "loss": 3.3505, "step": 201 }, { "epoch": 0.3851286939942803, "grad_norm": 5.8394689667726425, "learning_rate": 9.977295705447793e-06, "loss": 3.3892, "step": 202 }, { "epoch": 0.38703527168732127, "grad_norm": 5.8309911018346074, "learning_rate": 9.976228259812577e-06, "loss": 3.3636, "step": 203 }, { "epoch": 0.38894184938036225, "grad_norm": 6.103052469016239, "learning_rate": 9.975136353968547e-06, "loss": 3.4916, "step": 204 }, { "epoch": 0.3908484270734032, "grad_norm": 5.743196691401166, "learning_rate": 9.974019993282869e-06, "loss": 3.4428, "step": 205 }, { "epoch": 0.39275500476644426, "grad_norm": 5.893687562151129, "learning_rate": 9.972879183242913e-06, "loss": 3.4217, "step": 206 }, { "epoch": 0.39466158245948524, "grad_norm": 6.110472027271644, "learning_rate": 9.971713929456233e-06, "loss": 3.2234, "step": 207 }, { "epoch": 0.3965681601525262, "grad_norm": 6.226167230763502, "learning_rate": 9.970524237650527e-06, "loss": 3.5347, "step": 208 }, { "epoch": 0.3984747378455672, "grad_norm": 5.776379369351125, "learning_rate": 9.969310113673624e-06, "loss": 3.4171, "step": 209 }, { "epoch": 0.4003813155386082, "grad_norm": 6.3259993790066, "learning_rate": 9.96807156349344e-06, "loss": 3.4349, "step": 210 }, { "epoch": 0.4022878932316492, "grad_norm": 5.448698522372244, "learning_rate": 9.966808593197959e-06, "loss": 3.2422, "step": 211 }, { "epoch": 0.4041944709246902, "grad_norm": 6.2048364852960685, "learning_rate": 9.9655212089952e-06, "loss": 3.3196, "step": 212 }, { "epoch": 0.40610104861773116, "grad_norm": 6.301263899061097, "learning_rate": 9.964209417213185e-06, "loss": 3.3874, "step": 213 }, { "epoch": 0.40800762631077214, "grad_norm": 5.8103645384829194, "learning_rate": 9.962873224299912e-06, "loss": 3.3627, "step": 214 }, { "epoch": 0.4099142040038132, "grad_norm": 4.931466441973721, "learning_rate": 9.961512636823316e-06, "loss": 3.2843, "step": 215 }, { "epoch": 0.41182078169685415, "grad_norm": 6.308334319845687, "learning_rate": 9.960127661471248e-06, "loss": 3.4392, "step": 216 }, { "epoch": 0.41372735938989513, "grad_norm": 5.4742653518060695, "learning_rate": 9.958718305051427e-06, "loss": 3.3159, "step": 217 }, { "epoch": 0.4156339370829361, "grad_norm": 5.6194855191049795, "learning_rate": 9.957284574491423e-06, "loss": 3.3955, "step": 218 }, { "epoch": 0.41754051477597715, "grad_norm": 5.789930214619397, "learning_rate": 9.955826476838609e-06, "loss": 3.2994, "step": 219 }, { "epoch": 0.4194470924690181, "grad_norm": 5.096833264822564, "learning_rate": 9.954344019260138e-06, "loss": 3.3818, "step": 220 }, { "epoch": 0.4213536701620591, "grad_norm": 6.104824309661346, "learning_rate": 9.952837209042893e-06, "loss": 3.5139, "step": 221 }, { "epoch": 0.4232602478551001, "grad_norm": 6.171533895083068, "learning_rate": 9.951306053593474e-06, "loss": 3.474, "step": 222 }, { "epoch": 0.4251668255481411, "grad_norm": 5.859894110495812, "learning_rate": 9.949750560438134e-06, "loss": 3.4561, "step": 223 }, { "epoch": 0.4270734032411821, "grad_norm": 6.0692427067059835, "learning_rate": 9.948170737222763e-06, "loss": 3.3501, "step": 224 }, { "epoch": 0.42897998093422307, "grad_norm": 5.175300512347947, "learning_rate": 9.946566591712844e-06, "loss": 3.4292, "step": 225 }, { "epoch": 0.43088655862726405, "grad_norm": 5.92480217873188, "learning_rate": 9.944938131793406e-06, "loss": 3.3914, "step": 226 }, { "epoch": 0.432793136320305, "grad_norm": 5.880195104008302, "learning_rate": 9.943285365469008e-06, "loss": 3.4757, "step": 227 }, { "epoch": 0.43469971401334606, "grad_norm": 5.909420163764389, "learning_rate": 9.94160830086367e-06, "loss": 3.5574, "step": 228 }, { "epoch": 0.43660629170638704, "grad_norm": 5.727567804435262, "learning_rate": 9.939906946220857e-06, "loss": 3.3308, "step": 229 }, { "epoch": 0.438512869399428, "grad_norm": 5.424238634720088, "learning_rate": 9.938181309903427e-06, "loss": 3.558, "step": 230 }, { "epoch": 0.440419447092469, "grad_norm": 5.617727890757919, "learning_rate": 9.936431400393589e-06, "loss": 3.3073, "step": 231 }, { "epoch": 0.44232602478551003, "grad_norm": 5.025374049493658, "learning_rate": 9.934657226292874e-06, "loss": 3.362, "step": 232 }, { "epoch": 0.444232602478551, "grad_norm": 5.661390614737721, "learning_rate": 9.932858796322072e-06, "loss": 3.4254, "step": 233 }, { "epoch": 0.446139180171592, "grad_norm": 5.877397353045783, "learning_rate": 9.93103611932121e-06, "loss": 3.3923, "step": 234 }, { "epoch": 0.44804575786463297, "grad_norm": 5.918176094869725, "learning_rate": 9.929189204249491e-06, "loss": 3.531, "step": 235 }, { "epoch": 0.449952335557674, "grad_norm": 5.427744581544994, "learning_rate": 9.927318060185265e-06, "loss": 3.3973, "step": 236 }, { "epoch": 0.451858913250715, "grad_norm": 5.942368423706001, "learning_rate": 9.925422696325976e-06, "loss": 3.5039, "step": 237 }, { "epoch": 0.45376549094375596, "grad_norm": 6.461421760512872, "learning_rate": 9.923503121988114e-06, "loss": 3.3694, "step": 238 }, { "epoch": 0.45567206863679693, "grad_norm": 6.019838054576566, "learning_rate": 9.92155934660718e-06, "loss": 3.4018, "step": 239 }, { "epoch": 0.45757864632983797, "grad_norm": 5.75868887856665, "learning_rate": 9.919591379737626e-06, "loss": 3.1834, "step": 240 }, { "epoch": 0.45948522402287895, "grad_norm": 5.985076569750373, "learning_rate": 9.91759923105282e-06, "loss": 3.4928, "step": 241 }, { "epoch": 0.4613918017159199, "grad_norm": 5.625138866996842, "learning_rate": 9.915582910344992e-06, "loss": 3.3577, "step": 242 }, { "epoch": 0.4632983794089609, "grad_norm": 5.493018076481541, "learning_rate": 9.913542427525188e-06, "loss": 3.2531, "step": 243 }, { "epoch": 0.4652049571020019, "grad_norm": 5.548341447909812, "learning_rate": 9.91147779262322e-06, "loss": 3.4939, "step": 244 }, { "epoch": 0.4671115347950429, "grad_norm": 5.3212588669360255, "learning_rate": 9.90938901578762e-06, "loss": 3.1868, "step": 245 }, { "epoch": 0.4690181124880839, "grad_norm": 6.398204520006437, "learning_rate": 9.907276107285585e-06, "loss": 3.4114, "step": 246 }, { "epoch": 0.47092469018112487, "grad_norm": 5.7804919620577655, "learning_rate": 9.905139077502923e-06, "loss": 3.5325, "step": 247 }, { "epoch": 0.47283126787416585, "grad_norm": 5.605283642108553, "learning_rate": 9.902977936944023e-06, "loss": 3.3909, "step": 248 }, { "epoch": 0.4747378455672069, "grad_norm": 5.304356481504183, "learning_rate": 9.900792696231775e-06, "loss": 3.1974, "step": 249 }, { "epoch": 0.47664442326024786, "grad_norm": 5.145263285847987, "learning_rate": 9.898583366107539e-06, "loss": 3.1958, "step": 250 }, { "epoch": 0.47855100095328884, "grad_norm": 6.062826107905853, "learning_rate": 9.896349957431077e-06, "loss": 3.3653, "step": 251 }, { "epoch": 0.4804575786463298, "grad_norm": 5.934858455234602, "learning_rate": 9.894092481180515e-06, "loss": 3.2834, "step": 252 }, { "epoch": 0.48236415633937085, "grad_norm": 6.094957800197048, "learning_rate": 9.891810948452276e-06, "loss": 3.5766, "step": 253 }, { "epoch": 0.48427073403241183, "grad_norm": 5.714221156586713, "learning_rate": 9.889505370461033e-06, "loss": 3.4469, "step": 254 }, { "epoch": 0.4861773117254528, "grad_norm": 5.795897363949467, "learning_rate": 9.887175758539648e-06, "loss": 3.5477, "step": 255 }, { "epoch": 0.4880838894184938, "grad_norm": 5.4339593684696785, "learning_rate": 9.884822124139123e-06, "loss": 3.4346, "step": 256 }, { "epoch": 0.4899904671115348, "grad_norm": 5.923043223142063, "learning_rate": 9.88244447882854e-06, "loss": 3.3534, "step": 257 }, { "epoch": 0.4918970448045758, "grad_norm": 6.033003122890893, "learning_rate": 9.880042834295005e-06, "loss": 3.3717, "step": 258 }, { "epoch": 0.4938036224976168, "grad_norm": 5.857574884203947, "learning_rate": 9.877617202343584e-06, "loss": 3.4349, "step": 259 }, { "epoch": 0.49571020019065776, "grad_norm": 5.321739268912914, "learning_rate": 9.87516759489726e-06, "loss": 3.1882, "step": 260 }, { "epoch": 0.49761677788369874, "grad_norm": 5.67431793204277, "learning_rate": 9.87269402399686e-06, "loss": 3.6208, "step": 261 }, { "epoch": 0.49952335557673977, "grad_norm": 5.513405518867633, "learning_rate": 9.870196501801002e-06, "loss": 3.4229, "step": 262 }, { "epoch": 0.5014299332697807, "grad_norm": 5.643766072490395, "learning_rate": 9.867675040586035e-06, "loss": 3.4123, "step": 263 }, { "epoch": 0.5033365109628217, "grad_norm": 5.709228167544829, "learning_rate": 9.865129652745977e-06, "loss": 3.2949, "step": 264 }, { "epoch": 0.5052430886558628, "grad_norm": 5.749346298334859, "learning_rate": 9.86256035079246e-06, "loss": 3.4238, "step": 265 }, { "epoch": 0.5071496663489037, "grad_norm": 5.726140477931862, "learning_rate": 9.859967147354656e-06, "loss": 3.2949, "step": 266 }, { "epoch": 0.5090562440419447, "grad_norm": 6.319370756654015, "learning_rate": 9.857350055179228e-06, "loss": 3.1751, "step": 267 }, { "epoch": 0.5109628217349858, "grad_norm": 5.780806153916819, "learning_rate": 9.854709087130261e-06, "loss": 3.2388, "step": 268 }, { "epoch": 0.5128693994280267, "grad_norm": 5.556287296561824, "learning_rate": 9.8520442561892e-06, "loss": 3.2969, "step": 269 }, { "epoch": 0.5147759771210677, "grad_norm": 5.4738634059025655, "learning_rate": 9.849355575454782e-06, "loss": 3.3176, "step": 270 }, { "epoch": 0.5166825548141086, "grad_norm": 5.84168848368528, "learning_rate": 9.846643058142982e-06, "loss": 3.4679, "step": 271 }, { "epoch": 0.5185891325071497, "grad_norm": 5.464961718581959, "learning_rate": 9.843906717586937e-06, "loss": 3.3617, "step": 272 }, { "epoch": 0.5204957102001907, "grad_norm": 5.542352870603014, "learning_rate": 9.841146567236888e-06, "loss": 3.2572, "step": 273 }, { "epoch": 0.5224022878932316, "grad_norm": 5.721326112320646, "learning_rate": 9.838362620660106e-06, "loss": 3.3157, "step": 274 }, { "epoch": 0.5243088655862727, "grad_norm": 5.560147351168661, "learning_rate": 9.835554891540836e-06, "loss": 3.3391, "step": 275 }, { "epoch": 0.5262154432793136, "grad_norm": 5.177812450128499, "learning_rate": 9.832723393680222e-06, "loss": 3.2879, "step": 276 }, { "epoch": 0.5281220209723546, "grad_norm": 5.073172551707531, "learning_rate": 9.82986814099624e-06, "loss": 3.1986, "step": 277 }, { "epoch": 0.5300285986653956, "grad_norm": 5.520530089011374, "learning_rate": 9.82698914752363e-06, "loss": 3.406, "step": 278 }, { "epoch": 0.5319351763584366, "grad_norm": 6.4038626297908685, "learning_rate": 9.824086427413831e-06, "loss": 3.5419, "step": 279 }, { "epoch": 0.5338417540514776, "grad_norm": 5.543324660721873, "learning_rate": 9.821159994934906e-06, "loss": 3.4908, "step": 280 }, { "epoch": 0.5357483317445186, "grad_norm": 5.466441281662752, "learning_rate": 9.818209864471473e-06, "loss": 3.4308, "step": 281 }, { "epoch": 0.5376549094375596, "grad_norm": 5.9433518426255185, "learning_rate": 9.815236050524638e-06, "loss": 3.3644, "step": 282 }, { "epoch": 0.5395614871306006, "grad_norm": 5.798204806013681, "learning_rate": 9.81223856771192e-06, "loss": 3.3465, "step": 283 }, { "epoch": 0.5414680648236415, "grad_norm": 5.553093606990968, "learning_rate": 9.809217430767181e-06, "loss": 3.4975, "step": 284 }, { "epoch": 0.5433746425166825, "grad_norm": 5.213504335102452, "learning_rate": 9.806172654540549e-06, "loss": 3.23, "step": 285 }, { "epoch": 0.5452812202097236, "grad_norm": 5.410206781097647, "learning_rate": 9.803104253998354e-06, "loss": 3.1448, "step": 286 }, { "epoch": 0.5471877979027645, "grad_norm": 5.710859150424546, "learning_rate": 9.800012244223045e-06, "loss": 3.4752, "step": 287 }, { "epoch": 0.5490943755958055, "grad_norm": 4.916620075050709, "learning_rate": 9.796896640413122e-06, "loss": 3.3544, "step": 288 }, { "epoch": 0.5510009532888466, "grad_norm": 6.2077542666031755, "learning_rate": 9.793757457883062e-06, "loss": 3.3664, "step": 289 }, { "epoch": 0.5529075309818875, "grad_norm": 5.518298244583745, "learning_rate": 9.790594712063235e-06, "loss": 3.3776, "step": 290 }, { "epoch": 0.5548141086749285, "grad_norm": 5.6418871945608045, "learning_rate": 9.78740841849984e-06, "loss": 3.3793, "step": 291 }, { "epoch": 0.5567206863679695, "grad_norm": 5.677850185865345, "learning_rate": 9.784198592854819e-06, "loss": 3.4836, "step": 292 }, { "epoch": 0.5586272640610105, "grad_norm": 5.3393848210318415, "learning_rate": 9.780965250905787e-06, "loss": 3.2805, "step": 293 }, { "epoch": 0.5605338417540515, "grad_norm": 5.186809171807981, "learning_rate": 9.777708408545947e-06, "loss": 3.2603, "step": 294 }, { "epoch": 0.5624404194470924, "grad_norm": 5.914161910102725, "learning_rate": 9.774428081784022e-06, "loss": 3.2596, "step": 295 }, { "epoch": 0.5643469971401335, "grad_norm": 5.3131031867243514, "learning_rate": 9.771124286744166e-06, "loss": 3.3781, "step": 296 }, { "epoch": 0.5662535748331744, "grad_norm": 6.469860784202754, "learning_rate": 9.767797039665889e-06, "loss": 3.4081, "step": 297 }, { "epoch": 0.5681601525262154, "grad_norm": 7.124365776417729, "learning_rate": 9.764446356903983e-06, "loss": 3.4687, "step": 298 }, { "epoch": 0.5700667302192565, "grad_norm": 5.637856873398375, "learning_rate": 9.761072254928427e-06, "loss": 3.43, "step": 299 }, { "epoch": 0.5719733079122974, "grad_norm": 5.629024174000757, "learning_rate": 9.757674750324325e-06, "loss": 3.4701, "step": 300 }, { "epoch": 0.5738798856053384, "grad_norm": 6.697291805329228, "learning_rate": 9.754253859791805e-06, "loss": 3.4596, "step": 301 }, { "epoch": 0.5757864632983795, "grad_norm": 5.233255863510588, "learning_rate": 9.750809600145955e-06, "loss": 3.2807, "step": 302 }, { "epoch": 0.5776930409914204, "grad_norm": 6.438748626010119, "learning_rate": 9.747341988316726e-06, "loss": 3.3037, "step": 303 }, { "epoch": 0.5795996186844614, "grad_norm": 5.1619988172628775, "learning_rate": 9.743851041348855e-06, "loss": 3.2916, "step": 304 }, { "epoch": 0.5815061963775023, "grad_norm": 5.570294459192208, "learning_rate": 9.740336776401786e-06, "loss": 3.4076, "step": 305 }, { "epoch": 0.5834127740705434, "grad_norm": 5.744553429715351, "learning_rate": 9.736799210749575e-06, "loss": 3.2208, "step": 306 }, { "epoch": 0.5853193517635844, "grad_norm": 5.265912859724632, "learning_rate": 9.733238361780812e-06, "loss": 3.247, "step": 307 }, { "epoch": 0.5872259294566253, "grad_norm": 6.335846506912687, "learning_rate": 9.729654246998536e-06, "loss": 3.3115, "step": 308 }, { "epoch": 0.5891325071496664, "grad_norm": 5.777414663916508, "learning_rate": 9.726046884020142e-06, "loss": 3.3546, "step": 309 }, { "epoch": 0.5910390848427073, "grad_norm": 5.62470857081294, "learning_rate": 9.722416290577306e-06, "loss": 3.4056, "step": 310 }, { "epoch": 0.5929456625357483, "grad_norm": 7.8718153877412975, "learning_rate": 9.718762484515889e-06, "loss": 3.3654, "step": 311 }, { "epoch": 0.5948522402287894, "grad_norm": 4.9368482988988776, "learning_rate": 9.715085483795846e-06, "loss": 3.2349, "step": 312 }, { "epoch": 0.5967588179218303, "grad_norm": 5.542466194184823, "learning_rate": 9.711385306491153e-06, "loss": 3.2852, "step": 313 }, { "epoch": 0.5986653956148713, "grad_norm": 5.404229607724972, "learning_rate": 9.707661970789699e-06, "loss": 3.4764, "step": 314 }, { "epoch": 0.6005719733079123, "grad_norm": 5.38578963621338, "learning_rate": 9.703915494993215e-06, "loss": 3.3414, "step": 315 }, { "epoch": 0.6024785510009533, "grad_norm": 5.458494754350619, "learning_rate": 9.700145897517168e-06, "loss": 3.2302, "step": 316 }, { "epoch": 0.6043851286939943, "grad_norm": 5.253329944907259, "learning_rate": 9.696353196890681e-06, "loss": 3.2932, "step": 317 }, { "epoch": 0.6062917063870352, "grad_norm": 5.333070074307578, "learning_rate": 9.692537411756438e-06, "loss": 3.3441, "step": 318 }, { "epoch": 0.6081982840800763, "grad_norm": 5.820232698433408, "learning_rate": 9.688698560870591e-06, "loss": 3.3999, "step": 319 }, { "epoch": 0.6101048617731173, "grad_norm": 6.781135440792507, "learning_rate": 9.684836663102671e-06, "loss": 3.4896, "step": 320 }, { "epoch": 0.6120114394661582, "grad_norm": 6.018671174060485, "learning_rate": 9.680951737435497e-06, "loss": 3.4919, "step": 321 }, { "epoch": 0.6139180171591992, "grad_norm": 7.801777519897861, "learning_rate": 9.677043802965074e-06, "loss": 3.5088, "step": 322 }, { "epoch": 0.6158245948522403, "grad_norm": 5.557315777720947, "learning_rate": 9.673112878900509e-06, "loss": 3.2836, "step": 323 }, { "epoch": 0.6177311725452812, "grad_norm": 5.962727675241679, "learning_rate": 9.669158984563909e-06, "loss": 3.5013, "step": 324 }, { "epoch": 0.6196377502383222, "grad_norm": 5.133721558617345, "learning_rate": 9.665182139390292e-06, "loss": 3.2659, "step": 325 }, { "epoch": 0.6215443279313632, "grad_norm": 5.8508053196235315, "learning_rate": 9.66118236292749e-06, "loss": 3.2974, "step": 326 }, { "epoch": 0.6234509056244042, "grad_norm": 5.996511587940156, "learning_rate": 9.657159674836046e-06, "loss": 3.3082, "step": 327 }, { "epoch": 0.6253574833174452, "grad_norm": 5.095770579539413, "learning_rate": 9.653114094889128e-06, "loss": 3.4269, "step": 328 }, { "epoch": 0.6272640610104862, "grad_norm": 5.3020305448977245, "learning_rate": 9.649045642972426e-06, "loss": 3.2415, "step": 329 }, { "epoch": 0.6291706387035272, "grad_norm": 5.615905552018715, "learning_rate": 9.644954339084054e-06, "loss": 3.3159, "step": 330 }, { "epoch": 0.6310772163965681, "grad_norm": 5.464339283444214, "learning_rate": 9.640840203334454e-06, "loss": 3.3668, "step": 331 }, { "epoch": 0.6329837940896091, "grad_norm": 5.658590569638645, "learning_rate": 9.636703255946296e-06, "loss": 3.384, "step": 332 }, { "epoch": 0.6348903717826502, "grad_norm": 6.5113488344199615, "learning_rate": 9.632543517254377e-06, "loss": 3.3032, "step": 333 }, { "epoch": 0.6367969494756911, "grad_norm": 5.670977172234597, "learning_rate": 9.628361007705524e-06, "loss": 3.2674, "step": 334 }, { "epoch": 0.6387035271687321, "grad_norm": 5.249909753917849, "learning_rate": 9.624155747858493e-06, "loss": 3.3072, "step": 335 }, { "epoch": 0.6406101048617732, "grad_norm": 5.9120778544159265, "learning_rate": 9.619927758383863e-06, "loss": 3.37, "step": 336 }, { "epoch": 0.6425166825548141, "grad_norm": 5.064346288608819, "learning_rate": 9.615677060063945e-06, "loss": 3.3694, "step": 337 }, { "epoch": 0.6444232602478551, "grad_norm": 6.055888439146127, "learning_rate": 9.611403673792669e-06, "loss": 3.4164, "step": 338 }, { "epoch": 0.646329837940896, "grad_norm": 5.546946392886267, "learning_rate": 9.607107620575484e-06, "loss": 3.4645, "step": 339 }, { "epoch": 0.6482364156339371, "grad_norm": 5.547150565015598, "learning_rate": 9.60278892152926e-06, "loss": 3.3689, "step": 340 }, { "epoch": 0.6501429933269781, "grad_norm": 5.730302974046868, "learning_rate": 9.598447597882181e-06, "loss": 3.3291, "step": 341 }, { "epoch": 0.652049571020019, "grad_norm": 5.263243218929312, "learning_rate": 9.594083670973638e-06, "loss": 3.3908, "step": 342 }, { "epoch": 0.6539561487130601, "grad_norm": 5.437104345794195, "learning_rate": 9.589697162254124e-06, "loss": 3.3957, "step": 343 }, { "epoch": 0.655862726406101, "grad_norm": 6.1349603841465425, "learning_rate": 9.585288093285134e-06, "loss": 3.2846, "step": 344 }, { "epoch": 0.657769304099142, "grad_norm": 5.303912682450067, "learning_rate": 9.580856485739058e-06, "loss": 3.2898, "step": 345 }, { "epoch": 0.6596758817921831, "grad_norm": 5.753155915351638, "learning_rate": 9.576402361399071e-06, "loss": 3.4487, "step": 346 }, { "epoch": 0.661582459485224, "grad_norm": 5.7619499034761885, "learning_rate": 9.571925742159022e-06, "loss": 3.4597, "step": 347 }, { "epoch": 0.663489037178265, "grad_norm": 5.956307605287388, "learning_rate": 9.56742665002334e-06, "loss": 3.292, "step": 348 }, { "epoch": 0.665395614871306, "grad_norm": 4.885472963081056, "learning_rate": 9.56290510710691e-06, "loss": 3.4725, "step": 349 }, { "epoch": 0.667302192564347, "grad_norm": 5.042327970915872, "learning_rate": 9.558361135634979e-06, "loss": 3.2392, "step": 350 }, { "epoch": 0.669208770257388, "grad_norm": 5.8368421750916735, "learning_rate": 9.553794757943033e-06, "loss": 3.4279, "step": 351 }, { "epoch": 0.6711153479504289, "grad_norm": 5.402930826505062, "learning_rate": 9.549205996476698e-06, "loss": 3.3118, "step": 352 }, { "epoch": 0.67302192564347, "grad_norm": 6.324883992657217, "learning_rate": 9.544594873791625e-06, "loss": 3.4437, "step": 353 }, { "epoch": 0.674928503336511, "grad_norm": 5.770004799408654, "learning_rate": 9.539961412553375e-06, "loss": 3.4136, "step": 354 }, { "epoch": 0.6768350810295519, "grad_norm": 5.627608846692173, "learning_rate": 9.53530563553732e-06, "loss": 3.3137, "step": 355 }, { "epoch": 0.678741658722593, "grad_norm": 5.215119293470038, "learning_rate": 9.530627565628518e-06, "loss": 3.377, "step": 356 }, { "epoch": 0.680648236415634, "grad_norm": 5.896500657135818, "learning_rate": 9.525927225821605e-06, "loss": 3.363, "step": 357 }, { "epoch": 0.6825548141086749, "grad_norm": 8.396075471933672, "learning_rate": 9.521204639220687e-06, "loss": 3.2229, "step": 358 }, { "epoch": 0.684461391801716, "grad_norm": 5.58959760626914, "learning_rate": 9.51645982903922e-06, "loss": 3.4144, "step": 359 }, { "epoch": 0.6863679694947569, "grad_norm": 5.115101963713445, "learning_rate": 9.511692818599895e-06, "loss": 3.3144, "step": 360 }, { "epoch": 0.6882745471877979, "grad_norm": 5.767365253325921, "learning_rate": 9.506903631334532e-06, "loss": 3.3176, "step": 361 }, { "epoch": 0.6901811248808389, "grad_norm": 5.627520188494155, "learning_rate": 9.502092290783953e-06, "loss": 3.3937, "step": 362 }, { "epoch": 0.6920877025738799, "grad_norm": 5.220804332988557, "learning_rate": 9.49725882059788e-06, "loss": 3.3693, "step": 363 }, { "epoch": 0.6939942802669209, "grad_norm": 5.651090702264399, "learning_rate": 9.492403244534804e-06, "loss": 3.3422, "step": 364 }, { "epoch": 0.6959008579599618, "grad_norm": 5.384925146074498, "learning_rate": 9.48752558646188e-06, "loss": 3.3161, "step": 365 }, { "epoch": 0.6978074356530028, "grad_norm": 5.95949178941163, "learning_rate": 9.482625870354807e-06, "loss": 3.4701, "step": 366 }, { "epoch": 0.6997140133460439, "grad_norm": 5.492580984047033, "learning_rate": 9.477704120297698e-06, "loss": 3.2815, "step": 367 }, { "epoch": 0.7016205910390848, "grad_norm": 5.923017985909212, "learning_rate": 9.472760360482983e-06, "loss": 3.1413, "step": 368 }, { "epoch": 0.7035271687321258, "grad_norm": 5.278400141653775, "learning_rate": 9.467794615211276e-06, "loss": 3.3433, "step": 369 }, { "epoch": 0.7054337464251669, "grad_norm": 5.408382708643089, "learning_rate": 9.462806908891255e-06, "loss": 3.2544, "step": 370 }, { "epoch": 0.7073403241182078, "grad_norm": 5.612957961389119, "learning_rate": 9.457797266039552e-06, "loss": 3.2156, "step": 371 }, { "epoch": 0.7092469018112488, "grad_norm": 5.862183383144675, "learning_rate": 9.452765711280618e-06, "loss": 3.486, "step": 372 }, { "epoch": 0.7111534795042898, "grad_norm": 5.319369957782196, "learning_rate": 9.447712269346617e-06, "loss": 3.431, "step": 373 }, { "epoch": 0.7130600571973308, "grad_norm": 5.32261127570129, "learning_rate": 9.442636965077294e-06, "loss": 3.4126, "step": 374 }, { "epoch": 0.7149666348903718, "grad_norm": 5.39024324329459, "learning_rate": 9.437539823419856e-06, "loss": 3.3683, "step": 375 }, { "epoch": 0.7168732125834127, "grad_norm": 5.239180312187728, "learning_rate": 9.432420869428855e-06, "loss": 3.3056, "step": 376 }, { "epoch": 0.7187797902764538, "grad_norm": 5.569636199226671, "learning_rate": 9.427280128266049e-06, "loss": 3.3392, "step": 377 }, { "epoch": 0.7206863679694948, "grad_norm": 5.118369582214001, "learning_rate": 9.422117625200303e-06, "loss": 3.3787, "step": 378 }, { "epoch": 0.7225929456625357, "grad_norm": 5.560075356977722, "learning_rate": 9.416933385607436e-06, "loss": 3.4023, "step": 379 }, { "epoch": 0.7244995233555768, "grad_norm": 5.297079997690299, "learning_rate": 9.411727434970121e-06, "loss": 3.3366, "step": 380 }, { "epoch": 0.7264061010486177, "grad_norm": 5.232612135355929, "learning_rate": 9.406499798877748e-06, "loss": 3.3472, "step": 381 }, { "epoch": 0.7283126787416587, "grad_norm": 5.364495185658796, "learning_rate": 9.401250503026298e-06, "loss": 3.2441, "step": 382 }, { "epoch": 0.7302192564346998, "grad_norm": 5.566287751150542, "learning_rate": 9.39597957321822e-06, "loss": 3.2679, "step": 383 }, { "epoch": 0.7321258341277407, "grad_norm": 5.476073362234039, "learning_rate": 9.3906870353623e-06, "loss": 3.4441, "step": 384 }, { "epoch": 0.7340324118207817, "grad_norm": 5.996316982986525, "learning_rate": 9.38537291547354e-06, "loss": 3.4668, "step": 385 }, { "epoch": 0.7359389895138226, "grad_norm": 5.658835424390824, "learning_rate": 9.380037239673024e-06, "loss": 3.3712, "step": 386 }, { "epoch": 0.7378455672068637, "grad_norm": 5.385726542594866, "learning_rate": 9.374680034187795e-06, "loss": 3.563, "step": 387 }, { "epoch": 0.7397521448999047, "grad_norm": 5.163000476660958, "learning_rate": 9.369301325350721e-06, "loss": 3.3972, "step": 388 }, { "epoch": 0.7416587225929456, "grad_norm": 5.9650289410704165, "learning_rate": 9.363901139600369e-06, "loss": 3.3767, "step": 389 }, { "epoch": 0.7435653002859867, "grad_norm": 5.295281034396351, "learning_rate": 9.358479503480871e-06, "loss": 3.2638, "step": 390 }, { "epoch": 0.7454718779790277, "grad_norm": 5.268286883221677, "learning_rate": 9.3530364436418e-06, "loss": 3.3836, "step": 391 }, { "epoch": 0.7473784556720686, "grad_norm": 5.4325458079645985, "learning_rate": 9.347571986838037e-06, "loss": 3.4342, "step": 392 }, { "epoch": 0.7492850333651097, "grad_norm": 5.734789379994256, "learning_rate": 9.342086159929629e-06, "loss": 3.2957, "step": 393 }, { "epoch": 0.7511916110581506, "grad_norm": 5.740110573285962, "learning_rate": 9.336578989881677e-06, "loss": 3.5213, "step": 394 }, { "epoch": 0.7530981887511916, "grad_norm": 6.082697688471696, "learning_rate": 9.331050503764185e-06, "loss": 3.5067, "step": 395 }, { "epoch": 0.7550047664442326, "grad_norm": 6.016530525354293, "learning_rate": 9.325500728751936e-06, "loss": 3.3227, "step": 396 }, { "epoch": 0.7569113441372736, "grad_norm": 4.923660981474448, "learning_rate": 9.319929692124356e-06, "loss": 3.273, "step": 397 }, { "epoch": 0.7588179218303146, "grad_norm": 5.188922064582844, "learning_rate": 9.314337421265385e-06, "loss": 3.3077, "step": 398 }, { "epoch": 0.7607244995233555, "grad_norm": 5.775565567705561, "learning_rate": 9.30872394366333e-06, "loss": 3.3208, "step": 399 }, { "epoch": 0.7626310772163966, "grad_norm": 5.398594846102756, "learning_rate": 9.30308928691075e-06, "loss": 3.3158, "step": 400 }, { "epoch": 0.7645376549094376, "grad_norm": 5.712641348165823, "learning_rate": 9.297433478704292e-06, "loss": 3.3945, "step": 401 }, { "epoch": 0.7664442326024785, "grad_norm": 5.97707026122799, "learning_rate": 9.291756546844585e-06, "loss": 3.421, "step": 402 }, { "epoch": 0.7683508102955195, "grad_norm": 5.490813653313517, "learning_rate": 9.286058519236083e-06, "loss": 3.3862, "step": 403 }, { "epoch": 0.7702573879885606, "grad_norm": 5.400128108706967, "learning_rate": 9.280339423886935e-06, "loss": 3.3279, "step": 404 }, { "epoch": 0.7721639656816015, "grad_norm": 5.91951580715798, "learning_rate": 9.274599288908847e-06, "loss": 3.3866, "step": 405 }, { "epoch": 0.7740705433746425, "grad_norm": 5.897771188755772, "learning_rate": 9.268838142516943e-06, "loss": 3.4264, "step": 406 }, { "epoch": 0.7759771210676835, "grad_norm": 5.330737948036094, "learning_rate": 9.263056013029629e-06, "loss": 3.4103, "step": 407 }, { "epoch": 0.7778836987607245, "grad_norm": 5.370542401656543, "learning_rate": 9.257252928868452e-06, "loss": 3.3211, "step": 408 }, { "epoch": 0.7797902764537655, "grad_norm": 5.927357765535199, "learning_rate": 9.251428918557952e-06, "loss": 3.2292, "step": 409 }, { "epoch": 0.7816968541468065, "grad_norm": 5.5081494077099675, "learning_rate": 9.245584010725539e-06, "loss": 3.457, "step": 410 }, { "epoch": 0.7836034318398475, "grad_norm": 5.682512436615551, "learning_rate": 9.239718234101338e-06, "loss": 3.3277, "step": 411 }, { "epoch": 0.7855100095328885, "grad_norm": 5.540855054183156, "learning_rate": 9.233831617518055e-06, "loss": 3.3418, "step": 412 }, { "epoch": 0.7874165872259294, "grad_norm": 5.570395873958789, "learning_rate": 9.22792418991083e-06, "loss": 3.3428, "step": 413 }, { "epoch": 0.7893231649189705, "grad_norm": 5.541504880803922, "learning_rate": 9.221995980317102e-06, "loss": 3.3745, "step": 414 }, { "epoch": 0.7912297426120114, "grad_norm": 5.521263416952097, "learning_rate": 9.216047017876459e-06, "loss": 3.4422, "step": 415 }, { "epoch": 0.7931363203050524, "grad_norm": 5.358766265852832, "learning_rate": 9.210077331830497e-06, "loss": 3.4864, "step": 416 }, { "epoch": 0.7950428979980935, "grad_norm": 5.919313946618497, "learning_rate": 9.20408695152268e-06, "loss": 3.4159, "step": 417 }, { "epoch": 0.7969494756911344, "grad_norm": 5.271684351396428, "learning_rate": 9.19807590639819e-06, "loss": 3.2893, "step": 418 }, { "epoch": 0.7988560533841754, "grad_norm": 5.241005035593973, "learning_rate": 9.19204422600379e-06, "loss": 3.2813, "step": 419 }, { "epoch": 0.8007626310772163, "grad_norm": 5.681287590668343, "learning_rate": 9.185991939987664e-06, "loss": 3.2897, "step": 420 }, { "epoch": 0.8026692087702574, "grad_norm": 5.588903194624002, "learning_rate": 9.17991907809929e-06, "loss": 3.3323, "step": 421 }, { "epoch": 0.8045757864632984, "grad_norm": 5.982095515150058, "learning_rate": 9.173825670189287e-06, "loss": 3.1768, "step": 422 }, { "epoch": 0.8064823641563393, "grad_norm": 5.37323459732384, "learning_rate": 9.167711746209254e-06, "loss": 3.4008, "step": 423 }, { "epoch": 0.8083889418493804, "grad_norm": 5.2832417506850735, "learning_rate": 9.161577336211647e-06, "loss": 3.3331, "step": 424 }, { "epoch": 0.8102955195424214, "grad_norm": 6.44342940976069, "learning_rate": 9.15542247034961e-06, "loss": 3.269, "step": 425 }, { "epoch": 0.8122020972354623, "grad_norm": 5.482275355183911, "learning_rate": 9.149247178876847e-06, "loss": 3.2313, "step": 426 }, { "epoch": 0.8141086749285034, "grad_norm": 5.649232278636173, "learning_rate": 9.14305149214745e-06, "loss": 3.3483, "step": 427 }, { "epoch": 0.8160152526215443, "grad_norm": 5.823182202063583, "learning_rate": 9.13683544061577e-06, "loss": 3.3823, "step": 428 }, { "epoch": 0.8179218303145853, "grad_norm": 5.690064233269813, "learning_rate": 9.130599054836258e-06, "loss": 3.3786, "step": 429 }, { "epoch": 0.8198284080076264, "grad_norm": 4.894727358692659, "learning_rate": 9.124342365463314e-06, "loss": 3.2749, "step": 430 }, { "epoch": 0.8217349857006673, "grad_norm": 5.216554071756833, "learning_rate": 9.11806540325114e-06, "loss": 3.361, "step": 431 }, { "epoch": 0.8236415633937083, "grad_norm": 6.152912462366338, "learning_rate": 9.111768199053588e-06, "loss": 3.3663, "step": 432 }, { "epoch": 0.8255481410867492, "grad_norm": 5.237564192426968, "learning_rate": 9.105450783824006e-06, "loss": 3.453, "step": 433 }, { "epoch": 0.8274547187797903, "grad_norm": 5.312845540988894, "learning_rate": 9.099113188615089e-06, "loss": 3.252, "step": 434 }, { "epoch": 0.8293612964728313, "grad_norm": 6.232246323283339, "learning_rate": 9.092755444578726e-06, "loss": 3.3143, "step": 435 }, { "epoch": 0.8312678741658722, "grad_norm": 5.587289631196907, "learning_rate": 9.08637758296584e-06, "loss": 3.2135, "step": 436 }, { "epoch": 0.8331744518589133, "grad_norm": 5.29005227541765, "learning_rate": 9.079979635126248e-06, "loss": 3.1694, "step": 437 }, { "epoch": 0.8350810295519543, "grad_norm": 5.955695839110954, "learning_rate": 9.073561632508495e-06, "loss": 3.4022, "step": 438 }, { "epoch": 0.8369876072449952, "grad_norm": 5.202904421518815, "learning_rate": 9.067123606659701e-06, "loss": 3.3644, "step": 439 }, { "epoch": 0.8388941849380362, "grad_norm": 6.073327472999283, "learning_rate": 9.060665589225412e-06, "loss": 3.3701, "step": 440 }, { "epoch": 0.8408007626310772, "grad_norm": 5.339433668856727, "learning_rate": 9.054187611949444e-06, "loss": 3.1688, "step": 441 }, { "epoch": 0.8427073403241182, "grad_norm": 5.366861743231914, "learning_rate": 9.047689706673718e-06, "loss": 3.3257, "step": 442 }, { "epoch": 0.8446139180171592, "grad_norm": 5.132171987863734, "learning_rate": 9.041171905338108e-06, "loss": 3.1571, "step": 443 }, { "epoch": 0.8465204957102002, "grad_norm": 5.878603280611086, "learning_rate": 9.034634239980292e-06, "loss": 3.3344, "step": 444 }, { "epoch": 0.8484270734032412, "grad_norm": 5.449109180392489, "learning_rate": 9.028076742735583e-06, "loss": 3.2953, "step": 445 }, { "epoch": 0.8503336510962822, "grad_norm": 5.384126322690404, "learning_rate": 9.021499445836778e-06, "loss": 3.2933, "step": 446 }, { "epoch": 0.8522402287893232, "grad_norm": 5.960829709349916, "learning_rate": 9.014902381613997e-06, "loss": 3.5078, "step": 447 }, { "epoch": 0.8541468064823642, "grad_norm": 5.544207010964806, "learning_rate": 9.008285582494524e-06, "loss": 3.2143, "step": 448 }, { "epoch": 0.8560533841754051, "grad_norm": 5.57013938879804, "learning_rate": 9.001649081002648e-06, "loss": 3.3755, "step": 449 }, { "epoch": 0.8579599618684461, "grad_norm": 5.476941052086102, "learning_rate": 8.994992909759504e-06, "loss": 3.2984, "step": 450 }, { "epoch": 0.8598665395614872, "grad_norm": 5.329589403297242, "learning_rate": 8.988317101482912e-06, "loss": 3.3235, "step": 451 }, { "epoch": 0.8617731172545281, "grad_norm": 5.754738809569338, "learning_rate": 8.981621688987214e-06, "loss": 3.3156, "step": 452 }, { "epoch": 0.8636796949475691, "grad_norm": 5.493520792460005, "learning_rate": 8.97490670518312e-06, "loss": 3.4574, "step": 453 }, { "epoch": 0.86558627264061, "grad_norm": 5.816878761464463, "learning_rate": 8.968172183077534e-06, "loss": 3.3362, "step": 454 }, { "epoch": 0.8674928503336511, "grad_norm": 5.670737452383582, "learning_rate": 8.961418155773404e-06, "loss": 3.3575, "step": 455 }, { "epoch": 0.8693994280266921, "grad_norm": 5.127605379413027, "learning_rate": 8.954644656469551e-06, "loss": 3.3312, "step": 456 }, { "epoch": 0.871306005719733, "grad_norm": 5.342249563169675, "learning_rate": 8.947851718460512e-06, "loss": 3.2971, "step": 457 }, { "epoch": 0.8732125834127741, "grad_norm": 5.775875030503828, "learning_rate": 8.94103937513637e-06, "loss": 3.3179, "step": 458 }, { "epoch": 0.8751191611058151, "grad_norm": 5.136652003231521, "learning_rate": 8.9342076599826e-06, "loss": 3.2062, "step": 459 }, { "epoch": 0.877025738798856, "grad_norm": 5.17255625743417, "learning_rate": 8.927356606579882e-06, "loss": 3.212, "step": 460 }, { "epoch": 0.8789323164918971, "grad_norm": 6.100646692815173, "learning_rate": 8.92048624860397e-06, "loss": 3.4395, "step": 461 }, { "epoch": 0.880838894184938, "grad_norm": 5.443087738428869, "learning_rate": 8.913596619825495e-06, "loss": 3.3969, "step": 462 }, { "epoch": 0.882745471877979, "grad_norm": 5.081347400777752, "learning_rate": 8.90668775410982e-06, "loss": 3.0937, "step": 463 }, { "epoch": 0.8846520495710201, "grad_norm": 5.16939260160194, "learning_rate": 8.89975968541686e-06, "loss": 3.3841, "step": 464 }, { "epoch": 0.886558627264061, "grad_norm": 5.930459589682077, "learning_rate": 8.89281244780092e-06, "loss": 3.3846, "step": 465 }, { "epoch": 0.888465204957102, "grad_norm": 5.5936440714079465, "learning_rate": 8.885846075410537e-06, "loss": 3.4034, "step": 466 }, { "epoch": 0.8903717826501429, "grad_norm": 5.705452339398703, "learning_rate": 8.878860602488289e-06, "loss": 3.4061, "step": 467 }, { "epoch": 0.892278360343184, "grad_norm": 5.6792831956691705, "learning_rate": 8.871856063370653e-06, "loss": 3.4244, "step": 468 }, { "epoch": 0.894184938036225, "grad_norm": 5.306306739821467, "learning_rate": 8.864832492487821e-06, "loss": 3.3025, "step": 469 }, { "epoch": 0.8960915157292659, "grad_norm": 5.897754211353829, "learning_rate": 8.857789924363532e-06, "loss": 3.4333, "step": 470 }, { "epoch": 0.897998093422307, "grad_norm": 6.214720556932189, "learning_rate": 8.850728393614903e-06, "loss": 3.3623, "step": 471 }, { "epoch": 0.899904671115348, "grad_norm": 5.416111330979127, "learning_rate": 8.843647934952265e-06, "loss": 3.2495, "step": 472 }, { "epoch": 0.9018112488083889, "grad_norm": 5.673262698263259, "learning_rate": 8.836548583178983e-06, "loss": 3.3984, "step": 473 }, { "epoch": 0.90371782650143, "grad_norm": 5.725248844571677, "learning_rate": 8.829430373191294e-06, "loss": 3.3188, "step": 474 }, { "epoch": 0.9056244041944709, "grad_norm": 5.428302676919765, "learning_rate": 8.822293339978125e-06, "loss": 3.4078, "step": 475 }, { "epoch": 0.9075309818875119, "grad_norm": 5.098566141420619, "learning_rate": 8.81513751862093e-06, "loss": 3.2358, "step": 476 }, { "epoch": 0.909437559580553, "grad_norm": 5.2699418594095615, "learning_rate": 8.807962944293518e-06, "loss": 3.1889, "step": 477 }, { "epoch": 0.9113441372735939, "grad_norm": 5.800561399357849, "learning_rate": 8.800769652261869e-06, "loss": 3.3876, "step": 478 }, { "epoch": 0.9132507149666349, "grad_norm": 5.545329922092334, "learning_rate": 8.793557677883975e-06, "loss": 3.3932, "step": 479 }, { "epoch": 0.9151572926596759, "grad_norm": 5.192506838184484, "learning_rate": 8.786327056609654e-06, "loss": 3.3411, "step": 480 }, { "epoch": 0.9170638703527169, "grad_norm": 5.760510862793341, "learning_rate": 8.779077823980386e-06, "loss": 3.4127, "step": 481 }, { "epoch": 0.9189704480457579, "grad_norm": 5.137820217746843, "learning_rate": 8.771810015629132e-06, "loss": 3.3321, "step": 482 }, { "epoch": 0.9208770257387988, "grad_norm": 6.254250417358799, "learning_rate": 8.764523667280155e-06, "loss": 3.4462, "step": 483 }, { "epoch": 0.9227836034318398, "grad_norm": 5.73700904574578, "learning_rate": 8.75721881474886e-06, "loss": 3.2953, "step": 484 }, { "epoch": 0.9246901811248809, "grad_norm": 6.650650392096274, "learning_rate": 8.749895493941599e-06, "loss": 3.3779, "step": 485 }, { "epoch": 0.9265967588179218, "grad_norm": 5.107510912603686, "learning_rate": 8.742553740855507e-06, "loss": 3.2464, "step": 486 }, { "epoch": 0.9285033365109628, "grad_norm": 6.0722644098402165, "learning_rate": 8.735193591578318e-06, "loss": 3.197, "step": 487 }, { "epoch": 0.9304099142040038, "grad_norm": 5.294184901164172, "learning_rate": 8.727815082288196e-06, "loss": 3.2794, "step": 488 }, { "epoch": 0.9323164918970448, "grad_norm": 5.8316741293905565, "learning_rate": 8.720418249253547e-06, "loss": 3.4345, "step": 489 }, { "epoch": 0.9342230695900858, "grad_norm": 4.714729902383537, "learning_rate": 8.713003128832852e-06, "loss": 3.2429, "step": 490 }, { "epoch": 0.9361296472831268, "grad_norm": 5.426639078990197, "learning_rate": 8.705569757474475e-06, "loss": 3.2806, "step": 491 }, { "epoch": 0.9380362249761678, "grad_norm": 7.397836892681346, "learning_rate": 8.698118171716492e-06, "loss": 3.3623, "step": 492 }, { "epoch": 0.9399428026692088, "grad_norm": 5.546102497879822, "learning_rate": 8.690648408186515e-06, "loss": 3.3919, "step": 493 }, { "epoch": 0.9418493803622497, "grad_norm": 6.52403066752528, "learning_rate": 8.683160503601505e-06, "loss": 3.2953, "step": 494 }, { "epoch": 0.9437559580552908, "grad_norm": 5.465278950461086, "learning_rate": 8.67565449476759e-06, "loss": 3.4699, "step": 495 }, { "epoch": 0.9456625357483317, "grad_norm": 5.477693398495511, "learning_rate": 8.668130418579894e-06, "loss": 3.3881, "step": 496 }, { "epoch": 0.9475691134413727, "grad_norm": 5.6356148395126615, "learning_rate": 8.660588312022345e-06, "loss": 3.3949, "step": 497 }, { "epoch": 0.9494756911344138, "grad_norm": 5.88058201193319, "learning_rate": 8.653028212167498e-06, "loss": 3.3327, "step": 498 }, { "epoch": 0.9513822688274547, "grad_norm": 5.583866963678312, "learning_rate": 8.645450156176355e-06, "loss": 3.2175, "step": 499 }, { "epoch": 0.9532888465204957, "grad_norm": 5.272887242997353, "learning_rate": 8.637854181298178e-06, "loss": 3.5091, "step": 500 }, { "epoch": 0.9551954242135366, "grad_norm": 5.720112581048187, "learning_rate": 8.630240324870309e-06, "loss": 3.3305, "step": 501 }, { "epoch": 0.9571020019065777, "grad_norm": 5.54518272662921, "learning_rate": 8.622608624317983e-06, "loss": 3.4695, "step": 502 }, { "epoch": 0.9590085795996187, "grad_norm": 5.492559117815331, "learning_rate": 8.614959117154148e-06, "loss": 3.385, "step": 503 }, { "epoch": 0.9609151572926596, "grad_norm": 5.224263871461648, "learning_rate": 8.607291840979277e-06, "loss": 3.3479, "step": 504 }, { "epoch": 0.9628217349857007, "grad_norm": 5.72225498440259, "learning_rate": 8.599606833481189e-06, "loss": 3.4739, "step": 505 }, { "epoch": 0.9647283126787417, "grad_norm": 4.868614081237791, "learning_rate": 8.591904132434854e-06, "loss": 3.2296, "step": 506 }, { "epoch": 0.9666348903717826, "grad_norm": 5.764525850002545, "learning_rate": 8.584183775702217e-06, "loss": 3.3456, "step": 507 }, { "epoch": 0.9685414680648237, "grad_norm": 5.173936468469535, "learning_rate": 8.576445801232008e-06, "loss": 3.4444, "step": 508 }, { "epoch": 0.9704480457578646, "grad_norm": 5.680074487595007, "learning_rate": 8.568690247059554e-06, "loss": 3.3075, "step": 509 }, { "epoch": 0.9723546234509056, "grad_norm": 5.228575886547368, "learning_rate": 8.560917151306594e-06, "loss": 3.3565, "step": 510 }, { "epoch": 0.9742612011439467, "grad_norm": 6.0012127130035475, "learning_rate": 8.55312655218109e-06, "loss": 3.2518, "step": 511 }, { "epoch": 0.9761677788369876, "grad_norm": 5.145099092471222, "learning_rate": 8.545318487977043e-06, "loss": 3.5065, "step": 512 }, { "epoch": 0.9780743565300286, "grad_norm": 5.526722705878006, "learning_rate": 8.537492997074299e-06, "loss": 3.3203, "step": 513 }, { "epoch": 0.9799809342230696, "grad_norm": 5.595580826060759, "learning_rate": 8.529650117938364e-06, "loss": 3.5628, "step": 514 }, { "epoch": 0.9818875119161106, "grad_norm": 5.663756895532592, "learning_rate": 8.521789889120217e-06, "loss": 3.3447, "step": 515 }, { "epoch": 0.9837940896091516, "grad_norm": 5.277573737713141, "learning_rate": 8.513912349256115e-06, "loss": 3.194, "step": 516 }, { "epoch": 0.9857006673021925, "grad_norm": 5.896400624795743, "learning_rate": 8.506017537067408e-06, "loss": 3.4024, "step": 517 }, { "epoch": 0.9876072449952336, "grad_norm": 5.416697533837259, "learning_rate": 8.498105491360341e-06, "loss": 3.4154, "step": 518 }, { "epoch": 0.9895138226882746, "grad_norm": 4.720186585989478, "learning_rate": 8.49017625102588e-06, "loss": 3.0074, "step": 519 }, { "epoch": 0.9914204003813155, "grad_norm": 5.74578114843489, "learning_rate": 8.482229855039495e-06, "loss": 3.336, "step": 520 }, { "epoch": 0.9933269780743565, "grad_norm": 5.9526225822764465, "learning_rate": 8.474266342460997e-06, "loss": 3.3025, "step": 521 }, { "epoch": 0.9952335557673975, "grad_norm": 5.153931216808058, "learning_rate": 8.466285752434325e-06, "loss": 3.2762, "step": 522 }, { "epoch": 0.9971401334604385, "grad_norm": 5.271174094233703, "learning_rate": 8.45828812418736e-06, "loss": 3.3836, "step": 523 }, { "epoch": 0.9990467111534795, "grad_norm": 5.0815175045031165, "learning_rate": 8.450273497031736e-06, "loss": 3.3512, "step": 524 }, { "epoch": 1.0, "grad_norm": 5.0815175045031165, "learning_rate": 8.44224191036264e-06, "loss": 3.422, "step": 525 }, { "epoch": 1.001906577693041, "grad_norm": 8.313904962185626, "learning_rate": 8.434193403658629e-06, "loss": 2.7285, "step": 526 }, { "epoch": 1.003813155386082, "grad_norm": 5.267401763556151, "learning_rate": 8.426128016481417e-06, "loss": 2.8895, "step": 527 }, { "epoch": 1.005719733079123, "grad_norm": 5.624243351000887, "learning_rate": 8.418045788475706e-06, "loss": 2.8624, "step": 528 }, { "epoch": 1.007626310772164, "grad_norm": 5.14242689374461, "learning_rate": 8.409946759368967e-06, "loss": 3.0591, "step": 529 }, { "epoch": 1.009532888465205, "grad_norm": 5.043280949829558, "learning_rate": 8.401830968971262e-06, "loss": 2.838, "step": 530 }, { "epoch": 1.011439466158246, "grad_norm": 5.6632717337351774, "learning_rate": 8.393698457175037e-06, "loss": 2.8242, "step": 531 }, { "epoch": 1.013346043851287, "grad_norm": 5.22499946391844, "learning_rate": 8.385549263954935e-06, "loss": 2.7828, "step": 532 }, { "epoch": 1.0152526215443278, "grad_norm": 5.466201553774316, "learning_rate": 8.377383429367593e-06, "loss": 2.8905, "step": 533 }, { "epoch": 1.017159199237369, "grad_norm": 5.417461867698845, "learning_rate": 8.369200993551446e-06, "loss": 2.9233, "step": 534 }, { "epoch": 1.01906577693041, "grad_norm": 5.468075015957901, "learning_rate": 8.361001996726528e-06, "loss": 2.7922, "step": 535 }, { "epoch": 1.0209723546234508, "grad_norm": 5.7136512736836975, "learning_rate": 8.352786479194288e-06, "loss": 2.6634, "step": 536 }, { "epoch": 1.022878932316492, "grad_norm": 5.9295982720049665, "learning_rate": 8.344554481337367e-06, "loss": 2.8365, "step": 537 }, { "epoch": 1.0247855100095329, "grad_norm": 5.278201825857168, "learning_rate": 8.336306043619427e-06, "loss": 2.6958, "step": 538 }, { "epoch": 1.0266920877025738, "grad_norm": 5.353980260744125, "learning_rate": 8.32804120658493e-06, "loss": 2.5503, "step": 539 }, { "epoch": 1.028598665395615, "grad_norm": 5.346351924574419, "learning_rate": 8.319760010858946e-06, "loss": 2.7283, "step": 540 }, { "epoch": 1.0305052430886559, "grad_norm": 5.698600085362061, "learning_rate": 8.311462497146963e-06, "loss": 2.6576, "step": 541 }, { "epoch": 1.0324118207816968, "grad_norm": 6.074515341294887, "learning_rate": 8.303148706234673e-06, "loss": 2.6073, "step": 542 }, { "epoch": 1.034318398474738, "grad_norm": 5.414105736300453, "learning_rate": 8.294818678987777e-06, "loss": 2.7868, "step": 543 }, { "epoch": 1.0362249761677789, "grad_norm": 5.688050617956679, "learning_rate": 8.286472456351785e-06, "loss": 2.8346, "step": 544 }, { "epoch": 1.0381315538608198, "grad_norm": 5.808579778509937, "learning_rate": 8.278110079351817e-06, "loss": 2.7639, "step": 545 }, { "epoch": 1.0400381315538607, "grad_norm": 5.632080973000955, "learning_rate": 8.269731589092392e-06, "loss": 2.6563, "step": 546 }, { "epoch": 1.0419447092469019, "grad_norm": 6.2172675720990656, "learning_rate": 8.261337026757238e-06, "loss": 2.7833, "step": 547 }, { "epoch": 1.0438512869399428, "grad_norm": 5.381846820923297, "learning_rate": 8.252926433609084e-06, "loss": 2.6738, "step": 548 }, { "epoch": 1.0457578646329837, "grad_norm": 4.975576672089522, "learning_rate": 8.244499850989453e-06, "loss": 2.5899, "step": 549 }, { "epoch": 1.0476644423260248, "grad_norm": 5.431064664708909, "learning_rate": 8.236057320318463e-06, "loss": 2.6954, "step": 550 }, { "epoch": 1.0495710200190658, "grad_norm": 6.0489673038553375, "learning_rate": 8.227598883094628e-06, "loss": 2.6358, "step": 551 }, { "epoch": 1.0514775977121067, "grad_norm": 5.086653734565837, "learning_rate": 8.219124580894643e-06, "loss": 2.7283, "step": 552 }, { "epoch": 1.0533841754051478, "grad_norm": 5.603338224567255, "learning_rate": 8.210634455373192e-06, "loss": 2.6679, "step": 553 }, { "epoch": 1.0552907530981888, "grad_norm": 5.976857549636278, "learning_rate": 8.202128548262733e-06, "loss": 3.0186, "step": 554 }, { "epoch": 1.0571973307912297, "grad_norm": 7.158848035245011, "learning_rate": 8.1936069013733e-06, "loss": 2.8846, "step": 555 }, { "epoch": 1.0591039084842708, "grad_norm": 6.490364024155269, "learning_rate": 8.18506955659229e-06, "loss": 2.8039, "step": 556 }, { "epoch": 1.0610104861773118, "grad_norm": 5.701051399491982, "learning_rate": 8.176516555884265e-06, "loss": 2.6856, "step": 557 }, { "epoch": 1.0629170638703527, "grad_norm": 5.644928013947472, "learning_rate": 8.167947941290742e-06, "loss": 2.7312, "step": 558 }, { "epoch": 1.0648236415633936, "grad_norm": 5.904198269369168, "learning_rate": 8.159363754929987e-06, "loss": 2.7511, "step": 559 }, { "epoch": 1.0667302192564347, "grad_norm": 5.350170070999543, "learning_rate": 8.150764038996807e-06, "loss": 2.7899, "step": 560 }, { "epoch": 1.0686367969494757, "grad_norm": 5.734948255291437, "learning_rate": 8.142148835762343e-06, "loss": 2.579, "step": 561 }, { "epoch": 1.0705433746425166, "grad_norm": 5.837827426249328, "learning_rate": 8.133518187573864e-06, "loss": 2.6729, "step": 562 }, { "epoch": 1.0724499523355577, "grad_norm": 5.9196333774692595, "learning_rate": 8.124872136854552e-06, "loss": 2.7376, "step": 563 }, { "epoch": 1.0743565300285987, "grad_norm": 5.542272874925238, "learning_rate": 8.116210726103305e-06, "loss": 2.7684, "step": 564 }, { "epoch": 1.0762631077216396, "grad_norm": 5.750776042459241, "learning_rate": 8.107533997894521e-06, "loss": 2.8168, "step": 565 }, { "epoch": 1.0781696854146807, "grad_norm": 5.984815184293111, "learning_rate": 8.098841994877886e-06, "loss": 2.8719, "step": 566 }, { "epoch": 1.0800762631077216, "grad_norm": 5.583026960995908, "learning_rate": 8.090134759778173e-06, "loss": 2.6957, "step": 567 }, { "epoch": 1.0819828408007626, "grad_norm": 6.211707051297446, "learning_rate": 8.08141233539502e-06, "loss": 2.7425, "step": 568 }, { "epoch": 1.0838894184938037, "grad_norm": 6.01869262267896, "learning_rate": 8.072674764602735e-06, "loss": 2.7704, "step": 569 }, { "epoch": 1.0857959961868446, "grad_norm": 5.35398325164237, "learning_rate": 8.06392209035007e-06, "loss": 2.7679, "step": 570 }, { "epoch": 1.0877025738798856, "grad_norm": 6.061027915255409, "learning_rate": 8.055154355660018e-06, "loss": 2.699, "step": 571 }, { "epoch": 1.0896091515729265, "grad_norm": 6.258633789283423, "learning_rate": 8.046371603629606e-06, "loss": 2.6836, "step": 572 }, { "epoch": 1.0915157292659676, "grad_norm": 5.808559080533671, "learning_rate": 8.03757387742967e-06, "loss": 2.8648, "step": 573 }, { "epoch": 1.0934223069590085, "grad_norm": 6.015937420481165, "learning_rate": 8.028761220304652e-06, "loss": 2.8593, "step": 574 }, { "epoch": 1.0953288846520495, "grad_norm": 6.605955996698069, "learning_rate": 8.019933675572389e-06, "loss": 2.7941, "step": 575 }, { "epoch": 1.0972354623450906, "grad_norm": 5.719750851465378, "learning_rate": 8.011091286623893e-06, "loss": 2.8094, "step": 576 }, { "epoch": 1.0991420400381315, "grad_norm": 5.519055335947484, "learning_rate": 8.002234096923143e-06, "loss": 2.707, "step": 577 }, { "epoch": 1.1010486177311725, "grad_norm": 5.337710212961053, "learning_rate": 7.993362150006873e-06, "loss": 2.5446, "step": 578 }, { "epoch": 1.1029551954242136, "grad_norm": 5.776307162382591, "learning_rate": 7.984475489484346e-06, "loss": 2.5317, "step": 579 }, { "epoch": 1.1048617731172545, "grad_norm": 5.7724909817793035, "learning_rate": 7.975574159037159e-06, "loss": 2.8149, "step": 580 }, { "epoch": 1.1067683508102955, "grad_norm": 5.883126513314996, "learning_rate": 7.966658202419009e-06, "loss": 2.6801, "step": 581 }, { "epoch": 1.1086749285033366, "grad_norm": 5.244303349225299, "learning_rate": 7.95772766345549e-06, "loss": 2.7926, "step": 582 }, { "epoch": 1.1105815061963775, "grad_norm": 5.629273761589271, "learning_rate": 7.948782586043876e-06, "loss": 2.8348, "step": 583 }, { "epoch": 1.1124880838894184, "grad_norm": 6.15669866863214, "learning_rate": 7.939823014152905e-06, "loss": 2.9592, "step": 584 }, { "epoch": 1.1143946615824596, "grad_norm": 5.679424945150879, "learning_rate": 7.930848991822554e-06, "loss": 2.7468, "step": 585 }, { "epoch": 1.1163012392755005, "grad_norm": 5.757419038525037, "learning_rate": 7.921860563163834e-06, "loss": 2.8642, "step": 586 }, { "epoch": 1.1182078169685414, "grad_norm": 5.6078060019676865, "learning_rate": 7.912857772358571e-06, "loss": 2.6571, "step": 587 }, { "epoch": 1.1201143946615824, "grad_norm": 5.686898768166849, "learning_rate": 7.903840663659186e-06, "loss": 2.6883, "step": 588 }, { "epoch": 1.1220209723546235, "grad_norm": 5.324658608698989, "learning_rate": 7.894809281388474e-06, "loss": 2.6385, "step": 589 }, { "epoch": 1.1239275500476644, "grad_norm": 5.878207475139738, "learning_rate": 7.885763669939395e-06, "loss": 2.8502, "step": 590 }, { "epoch": 1.1258341277407053, "grad_norm": 5.791130447103623, "learning_rate": 7.876703873774852e-06, "loss": 2.7475, "step": 591 }, { "epoch": 1.1277407054337465, "grad_norm": 5.568633808193119, "learning_rate": 7.867629937427468e-06, "loss": 2.808, "step": 592 }, { "epoch": 1.1296472831267874, "grad_norm": 5.283052134577616, "learning_rate": 7.858541905499374e-06, "loss": 2.8181, "step": 593 }, { "epoch": 1.1315538608198283, "grad_norm": 6.544521770922512, "learning_rate": 7.849439822661985e-06, "loss": 2.8076, "step": 594 }, { "epoch": 1.1334604385128695, "grad_norm": 5.452280491085309, "learning_rate": 7.84032373365578e-06, "loss": 2.687, "step": 595 }, { "epoch": 1.1353670162059104, "grad_norm": 5.654803693914413, "learning_rate": 7.831193683290087e-06, "loss": 2.7131, "step": 596 }, { "epoch": 1.1372735938989513, "grad_norm": 5.217235236860667, "learning_rate": 7.82204971644286e-06, "loss": 2.6906, "step": 597 }, { "epoch": 1.1391801715919923, "grad_norm": 5.704949393951046, "learning_rate": 7.81289187806046e-06, "loss": 2.5717, "step": 598 }, { "epoch": 1.1410867492850334, "grad_norm": 5.429249915024454, "learning_rate": 7.803720213157424e-06, "loss": 2.6903, "step": 599 }, { "epoch": 1.1429933269780743, "grad_norm": 5.94307717922075, "learning_rate": 7.794534766816262e-06, "loss": 2.793, "step": 600 }, { "epoch": 1.1448999046711155, "grad_norm": 6.145155055900271, "learning_rate": 7.78533558418722e-06, "loss": 2.9386, "step": 601 }, { "epoch": 1.1468064823641564, "grad_norm": 5.66324236789927, "learning_rate": 7.776122710488063e-06, "loss": 2.8813, "step": 602 }, { "epoch": 1.1487130600571973, "grad_norm": 6.00005453323277, "learning_rate": 7.766896191003855e-06, "loss": 2.7039, "step": 603 }, { "epoch": 1.1506196377502382, "grad_norm": 5.236686261619667, "learning_rate": 7.757656071086737e-06, "loss": 2.7838, "step": 604 }, { "epoch": 1.1525262154432794, "grad_norm": 5.473465907122794, "learning_rate": 7.748402396155699e-06, "loss": 2.6237, "step": 605 }, { "epoch": 1.1544327931363203, "grad_norm": 5.773641674565471, "learning_rate": 7.739135211696355e-06, "loss": 2.8374, "step": 606 }, { "epoch": 1.1563393708293612, "grad_norm": 6.088237325933673, "learning_rate": 7.729854563260734e-06, "loss": 2.4613, "step": 607 }, { "epoch": 1.1582459485224024, "grad_norm": 5.795992573021472, "learning_rate": 7.72056049646704e-06, "loss": 2.7535, "step": 608 }, { "epoch": 1.1601525262154433, "grad_norm": 5.793412137547177, "learning_rate": 7.71125305699943e-06, "loss": 2.6037, "step": 609 }, { "epoch": 1.1620591039084842, "grad_norm": 5.261364102214614, "learning_rate": 7.701932290607803e-06, "loss": 2.6065, "step": 610 }, { "epoch": 1.1639656816015251, "grad_norm": 6.200195371865063, "learning_rate": 7.692598243107556e-06, "loss": 2.8671, "step": 611 }, { "epoch": 1.1658722592945663, "grad_norm": 6.071331592347983, "learning_rate": 7.683250960379372e-06, "loss": 2.7337, "step": 612 }, { "epoch": 1.1677788369876072, "grad_norm": 5.472299467202838, "learning_rate": 7.673890488368994e-06, "loss": 2.6414, "step": 613 }, { "epoch": 1.1696854146806483, "grad_norm": 6.212160234406006, "learning_rate": 7.664516873086987e-06, "loss": 2.7982, "step": 614 }, { "epoch": 1.1715919923736893, "grad_norm": 6.02346575754263, "learning_rate": 7.65513016060853e-06, "loss": 2.6512, "step": 615 }, { "epoch": 1.1734985700667302, "grad_norm": 5.24088550278419, "learning_rate": 7.645730397073175e-06, "loss": 2.7875, "step": 616 }, { "epoch": 1.1754051477597711, "grad_norm": 5.758888558024453, "learning_rate": 7.636317628684626e-06, "loss": 2.7315, "step": 617 }, { "epoch": 1.1773117254528123, "grad_norm": 5.619275590924718, "learning_rate": 7.6268919017105115e-06, "loss": 2.786, "step": 618 }, { "epoch": 1.1792183031458532, "grad_norm": 6.149402941233468, "learning_rate": 7.617453262482157e-06, "loss": 2.7742, "step": 619 }, { "epoch": 1.181124880838894, "grad_norm": 5.8158055788723555, "learning_rate": 7.608001757394356e-06, "loss": 2.6158, "step": 620 }, { "epoch": 1.1830314585319353, "grad_norm": 5.812576837651012, "learning_rate": 7.598537432905143e-06, "loss": 2.6403, "step": 621 }, { "epoch": 1.1849380362249762, "grad_norm": 5.722283530770768, "learning_rate": 7.5890603355355685e-06, "loss": 2.7749, "step": 622 }, { "epoch": 1.186844613918017, "grad_norm": 5.40529448278363, "learning_rate": 7.579570511869463e-06, "loss": 2.7531, "step": 623 }, { "epoch": 1.1887511916110582, "grad_norm": 5.667249458342123, "learning_rate": 7.570068008553213e-06, "loss": 2.7179, "step": 624 }, { "epoch": 1.1906577693040992, "grad_norm": 7.865689996083414, "learning_rate": 7.56055287229553e-06, "loss": 2.9119, "step": 625 }, { "epoch": 1.19256434699714, "grad_norm": 5.698056883782881, "learning_rate": 7.551025149867226e-06, "loss": 2.6851, "step": 626 }, { "epoch": 1.1944709246901812, "grad_norm": 5.736805821811366, "learning_rate": 7.541484888100974e-06, "loss": 2.7773, "step": 627 }, { "epoch": 1.1963775023832222, "grad_norm": 5.653188788464322, "learning_rate": 7.531932133891085e-06, "loss": 2.6724, "step": 628 }, { "epoch": 1.198284080076263, "grad_norm": 5.649655413334166, "learning_rate": 7.5223669341932725e-06, "loss": 2.6973, "step": 629 }, { "epoch": 1.200190657769304, "grad_norm": 6.161295047792969, "learning_rate": 7.512789336024433e-06, "loss": 2.8509, "step": 630 }, { "epoch": 1.2020972354623451, "grad_norm": 5.682406502594368, "learning_rate": 7.503199386462397e-06, "loss": 2.6831, "step": 631 }, { "epoch": 1.204003813155386, "grad_norm": 5.9629151156593885, "learning_rate": 7.4935971326457134e-06, "loss": 2.7272, "step": 632 }, { "epoch": 1.205910390848427, "grad_norm": 5.823822407203151, "learning_rate": 7.483982621773408e-06, "loss": 2.6268, "step": 633 }, { "epoch": 1.2078169685414681, "grad_norm": 7.232882893979798, "learning_rate": 7.474355901104758e-06, "loss": 2.6451, "step": 634 }, { "epoch": 1.209723546234509, "grad_norm": 5.420147100603969, "learning_rate": 7.464717017959055e-06, "loss": 2.7763, "step": 635 }, { "epoch": 1.21163012392755, "grad_norm": 6.715798614597261, "learning_rate": 7.455066019715372e-06, "loss": 2.8744, "step": 636 }, { "epoch": 1.2135367016205911, "grad_norm": 5.07703906297626, "learning_rate": 7.445402953812339e-06, "loss": 2.4712, "step": 637 }, { "epoch": 1.215443279313632, "grad_norm": 5.475551974800493, "learning_rate": 7.435727867747898e-06, "loss": 2.7868, "step": 638 }, { "epoch": 1.217349857006673, "grad_norm": 5.655655669494292, "learning_rate": 7.426040809079077e-06, "loss": 2.7114, "step": 639 }, { "epoch": 1.2192564346997141, "grad_norm": 5.575833108727368, "learning_rate": 7.416341825421755e-06, "loss": 2.6622, "step": 640 }, { "epoch": 1.221163012392755, "grad_norm": 5.358621852833905, "learning_rate": 7.406630964450423e-06, "loss": 2.7141, "step": 641 }, { "epoch": 1.223069590085796, "grad_norm": 8.62205993996843, "learning_rate": 7.39690827389796e-06, "loss": 2.6623, "step": 642 }, { "epoch": 1.2249761677788369, "grad_norm": 5.4665583045126045, "learning_rate": 7.387173801555389e-06, "loss": 2.6538, "step": 643 }, { "epoch": 1.226882745471878, "grad_norm": 6.020974591131672, "learning_rate": 7.3774275952716444e-06, "loss": 2.6755, "step": 644 }, { "epoch": 1.228789323164919, "grad_norm": 5.240315047794263, "learning_rate": 7.367669702953341e-06, "loss": 2.5823, "step": 645 }, { "epoch": 1.2306959008579599, "grad_norm": 5.829855192130291, "learning_rate": 7.357900172564532e-06, "loss": 2.688, "step": 646 }, { "epoch": 1.232602478551001, "grad_norm": 6.091074351003968, "learning_rate": 7.348119052126478e-06, "loss": 2.8687, "step": 647 }, { "epoch": 1.234509056244042, "grad_norm": 5.906777458128272, "learning_rate": 7.338326389717409e-06, "loss": 2.7026, "step": 648 }, { "epoch": 1.2364156339370829, "grad_norm": 5.9686616349680035, "learning_rate": 7.32852223347229e-06, "loss": 2.9312, "step": 649 }, { "epoch": 1.238322211630124, "grad_norm": 6.691607879080555, "learning_rate": 7.3187066315825816e-06, "loss": 2.6361, "step": 650 }, { "epoch": 1.240228789323165, "grad_norm": 6.477111046870976, "learning_rate": 7.308879632296004e-06, "loss": 2.696, "step": 651 }, { "epoch": 1.2421353670162059, "grad_norm": 6.131348080952952, "learning_rate": 7.2990412839163015e-06, "loss": 2.6547, "step": 652 }, { "epoch": 1.244041944709247, "grad_norm": 5.836459802552913, "learning_rate": 7.289191634803002e-06, "loss": 2.6195, "step": 653 }, { "epoch": 1.245948522402288, "grad_norm": 6.071388772138896, "learning_rate": 7.279330733371184e-06, "loss": 2.6741, "step": 654 }, { "epoch": 1.2478551000953289, "grad_norm": 5.335164786866538, "learning_rate": 7.269458628091233e-06, "loss": 2.7096, "step": 655 }, { "epoch": 1.2497616777883698, "grad_norm": 5.816059917469385, "learning_rate": 7.2595753674886084e-06, "loss": 2.7745, "step": 656 }, { "epoch": 1.251668255481411, "grad_norm": 5.787163199976156, "learning_rate": 7.2496810001436e-06, "loss": 2.7338, "step": 657 }, { "epoch": 1.2535748331744518, "grad_norm": 5.595985920211398, "learning_rate": 7.2397755746910944e-06, "loss": 2.7167, "step": 658 }, { "epoch": 1.255481410867493, "grad_norm": 6.333044708553903, "learning_rate": 7.229859139820332e-06, "loss": 2.6198, "step": 659 }, { "epoch": 1.257387988560534, "grad_norm": 5.613241393726865, "learning_rate": 7.21993174427467e-06, "loss": 2.8795, "step": 660 }, { "epoch": 1.2592945662535748, "grad_norm": 6.603445772663904, "learning_rate": 7.209993436851338e-06, "loss": 2.8418, "step": 661 }, { "epoch": 1.2612011439466158, "grad_norm": 6.232517453567855, "learning_rate": 7.200044266401209e-06, "loss": 2.802, "step": 662 }, { "epoch": 1.2631077216396567, "grad_norm": 5.752709009736414, "learning_rate": 7.190084281828544e-06, "loss": 2.7493, "step": 663 }, { "epoch": 1.2650142993326978, "grad_norm": 5.747867682821619, "learning_rate": 7.1801135320907685e-06, "loss": 2.7056, "step": 664 }, { "epoch": 1.2669208770257387, "grad_norm": 5.541804996767364, "learning_rate": 7.170132066198217e-06, "loss": 2.5869, "step": 665 }, { "epoch": 1.26882745471878, "grad_norm": 5.5385346036655125, "learning_rate": 7.160139933213899e-06, "loss": 2.6381, "step": 666 }, { "epoch": 1.2707340324118208, "grad_norm": 5.598519754849501, "learning_rate": 7.1501371822532586e-06, "loss": 2.7407, "step": 667 }, { "epoch": 1.2726406101048617, "grad_norm": 5.8731047785454615, "learning_rate": 7.140123862483933e-06, "loss": 2.7575, "step": 668 }, { "epoch": 1.2745471877979027, "grad_norm": 5.9414633004521455, "learning_rate": 7.130100023125506e-06, "loss": 2.7521, "step": 669 }, { "epoch": 1.2764537654909438, "grad_norm": 5.925247021634454, "learning_rate": 7.1200657134492725e-06, "loss": 2.7357, "step": 670 }, { "epoch": 1.2783603431839847, "grad_norm": 6.109853187981374, "learning_rate": 7.110020982777989e-06, "loss": 2.749, "step": 671 }, { "epoch": 1.2802669208770259, "grad_norm": 6.724761964308327, "learning_rate": 7.099965880485644e-06, "loss": 2.8374, "step": 672 }, { "epoch": 1.2821734985700668, "grad_norm": 5.982929638526128, "learning_rate": 7.089900455997198e-06, "loss": 2.753, "step": 673 }, { "epoch": 1.2840800762631077, "grad_norm": 5.976827236767735, "learning_rate": 7.07982475878835e-06, "loss": 2.6963, "step": 674 }, { "epoch": 1.2859866539561486, "grad_norm": 5.565753135581338, "learning_rate": 7.0697388383853025e-06, "loss": 2.7532, "step": 675 }, { "epoch": 1.2878932316491898, "grad_norm": 5.707169720656655, "learning_rate": 7.059642744364499e-06, "loss": 2.6781, "step": 676 }, { "epoch": 1.2897998093422307, "grad_norm": 5.661040233034286, "learning_rate": 7.049536526352394e-06, "loss": 2.6796, "step": 677 }, { "epoch": 1.2917063870352716, "grad_norm": 6.131452206195069, "learning_rate": 7.039420234025208e-06, "loss": 2.834, "step": 678 }, { "epoch": 1.2936129647283128, "grad_norm": 5.898288769260731, "learning_rate": 7.029293917108678e-06, "loss": 2.7761, "step": 679 }, { "epoch": 1.2955195424213537, "grad_norm": 5.827830488606018, "learning_rate": 7.0191576253778195e-06, "loss": 2.7813, "step": 680 }, { "epoch": 1.2974261201143946, "grad_norm": 6.115806339731799, "learning_rate": 7.009011408656672e-06, "loss": 2.7162, "step": 681 }, { "epoch": 1.2993326978074355, "grad_norm": 5.811707977945406, "learning_rate": 6.998855316818065e-06, "loss": 2.7391, "step": 682 }, { "epoch": 1.3012392755004767, "grad_norm": 6.128791713268644, "learning_rate": 6.988689399783371e-06, "loss": 2.8552, "step": 683 }, { "epoch": 1.3031458531935176, "grad_norm": 6.888992132957313, "learning_rate": 6.978513707522248e-06, "loss": 2.5641, "step": 684 }, { "epoch": 1.3050524308865588, "grad_norm": 5.917068635448027, "learning_rate": 6.968328290052415e-06, "loss": 2.9157, "step": 685 }, { "epoch": 1.3069590085795997, "grad_norm": 6.234309312805226, "learning_rate": 6.958133197439385e-06, "loss": 2.7104, "step": 686 }, { "epoch": 1.3088655862726406, "grad_norm": 6.188597273527159, "learning_rate": 6.9479284797962335e-06, "loss": 2.6959, "step": 687 }, { "epoch": 1.3107721639656815, "grad_norm": 5.996330857992042, "learning_rate": 6.937714187283344e-06, "loss": 2.5877, "step": 688 }, { "epoch": 1.3126787416587227, "grad_norm": 6.099948935145451, "learning_rate": 6.927490370108167e-06, "loss": 2.7434, "step": 689 }, { "epoch": 1.3145853193517636, "grad_norm": 5.7661898732463595, "learning_rate": 6.917257078524968e-06, "loss": 2.6538, "step": 690 }, { "epoch": 1.3164918970448045, "grad_norm": 6.132192116126521, "learning_rate": 6.9070143628345844e-06, "loss": 2.7165, "step": 691 }, { "epoch": 1.3183984747378457, "grad_norm": 5.518600552543245, "learning_rate": 6.896762273384179e-06, "loss": 2.8013, "step": 692 }, { "epoch": 1.3203050524308866, "grad_norm": 5.88529938718483, "learning_rate": 6.886500860566987e-06, "loss": 2.6531, "step": 693 }, { "epoch": 1.3222116301239275, "grad_norm": 5.525829513537348, "learning_rate": 6.876230174822074e-06, "loss": 2.7538, "step": 694 }, { "epoch": 1.3241182078169684, "grad_norm": 5.969816899810715, "learning_rate": 6.865950266634085e-06, "loss": 2.8045, "step": 695 }, { "epoch": 1.3260247855100096, "grad_norm": 6.000853140860954, "learning_rate": 6.855661186532996e-06, "loss": 2.8866, "step": 696 }, { "epoch": 1.3279313632030505, "grad_norm": 6.499346548179471, "learning_rate": 6.8453629850938695e-06, "loss": 2.7597, "step": 697 }, { "epoch": 1.3298379408960916, "grad_norm": 6.14903679417467, "learning_rate": 6.835055712936603e-06, "loss": 2.6877, "step": 698 }, { "epoch": 1.3317445185891326, "grad_norm": 5.913721731570846, "learning_rate": 6.824739420725675e-06, "loss": 2.7025, "step": 699 }, { "epoch": 1.3336510962821735, "grad_norm": 5.774945084967542, "learning_rate": 6.814414159169909e-06, "loss": 2.6921, "step": 700 }, { "epoch": 1.3355576739752144, "grad_norm": 6.352278486174908, "learning_rate": 6.804079979022209e-06, "loss": 2.9025, "step": 701 }, { "epoch": 1.3374642516682556, "grad_norm": 6.11877910322647, "learning_rate": 6.793736931079325e-06, "loss": 2.7275, "step": 702 }, { "epoch": 1.3393708293612965, "grad_norm": 6.10567321187883, "learning_rate": 6.78338506618159e-06, "loss": 2.7909, "step": 703 }, { "epoch": 1.3412774070543374, "grad_norm": 5.550442609407546, "learning_rate": 6.773024435212678e-06, "loss": 2.7622, "step": 704 }, { "epoch": 1.3431839847473785, "grad_norm": 6.427472573349018, "learning_rate": 6.762655089099353e-06, "loss": 2.7783, "step": 705 }, { "epoch": 1.3450905624404195, "grad_norm": 6.108866738064591, "learning_rate": 6.752277078811218e-06, "loss": 2.8244, "step": 706 }, { "epoch": 1.3469971401334604, "grad_norm": 5.891144073254848, "learning_rate": 6.741890455360458e-06, "loss": 2.691, "step": 707 }, { "epoch": 1.3489037178265013, "grad_norm": 5.578341428227693, "learning_rate": 6.731495269801604e-06, "loss": 2.7446, "step": 708 }, { "epoch": 1.3508102955195425, "grad_norm": 5.760190278028284, "learning_rate": 6.721091573231267e-06, "loss": 2.7257, "step": 709 }, { "epoch": 1.3527168732125834, "grad_norm": 5.348090589560754, "learning_rate": 6.710679416787893e-06, "loss": 2.6399, "step": 710 }, { "epoch": 1.3546234509056245, "grad_norm": 5.44950320559314, "learning_rate": 6.700258851651519e-06, "loss": 2.8274, "step": 711 }, { "epoch": 1.3565300285986654, "grad_norm": 5.450538184243979, "learning_rate": 6.689829929043504e-06, "loss": 2.8418, "step": 712 }, { "epoch": 1.3584366062917064, "grad_norm": 6.528337693965505, "learning_rate": 6.679392700226296e-06, "loss": 2.9393, "step": 713 }, { "epoch": 1.3603431839847473, "grad_norm": 5.974826163707531, "learning_rate": 6.6689472165031655e-06, "loss": 2.7924, "step": 714 }, { "epoch": 1.3622497616777884, "grad_norm": 5.735485014860257, "learning_rate": 6.658493529217963e-06, "loss": 2.6824, "step": 715 }, { "epoch": 1.3641563393708294, "grad_norm": 5.974209771039, "learning_rate": 6.648031689754861e-06, "loss": 2.6339, "step": 716 }, { "epoch": 1.3660629170638703, "grad_norm": 5.147691716037623, "learning_rate": 6.637561749538104e-06, "loss": 2.7848, "step": 717 }, { "epoch": 1.3679694947569114, "grad_norm": 5.787581948424895, "learning_rate": 6.627083760031755e-06, "loss": 2.6699, "step": 718 }, { "epoch": 1.3698760724499524, "grad_norm": 5.6196013564311675, "learning_rate": 6.616597772739441e-06, "loss": 2.6435, "step": 719 }, { "epoch": 1.3717826501429933, "grad_norm": 5.788795780739395, "learning_rate": 6.6061038392041055e-06, "loss": 2.697, "step": 720 }, { "epoch": 1.3736892278360342, "grad_norm": 6.124953849257112, "learning_rate": 6.595602011007747e-06, "loss": 2.7406, "step": 721 }, { "epoch": 1.3755958055290753, "grad_norm": 5.397284298610729, "learning_rate": 6.585092339771169e-06, "loss": 2.6139, "step": 722 }, { "epoch": 1.3775023832221163, "grad_norm": 5.811045016094093, "learning_rate": 6.574574877153733e-06, "loss": 2.5686, "step": 723 }, { "epoch": 1.3794089609151574, "grad_norm": 5.93728532276096, "learning_rate": 6.56404967485309e-06, "loss": 2.6968, "step": 724 }, { "epoch": 1.3813155386081983, "grad_norm": 5.480881296298576, "learning_rate": 6.553516784604941e-06, "loss": 2.6459, "step": 725 }, { "epoch": 1.3832221163012393, "grad_norm": 5.024213093480167, "learning_rate": 6.5429762581827735e-06, "loss": 2.6859, "step": 726 }, { "epoch": 1.3851286939942802, "grad_norm": 6.363483885488669, "learning_rate": 6.532428147397609e-06, "loss": 2.8059, "step": 727 }, { "epoch": 1.3870352716873213, "grad_norm": 5.330615601033536, "learning_rate": 6.521872504097755e-06, "loss": 2.6943, "step": 728 }, { "epoch": 1.3889418493803622, "grad_norm": 5.7258508023594334, "learning_rate": 6.511309380168535e-06, "loss": 2.7997, "step": 729 }, { "epoch": 1.3908484270734032, "grad_norm": 6.030932184741667, "learning_rate": 6.500738827532052e-06, "loss": 2.6631, "step": 730 }, { "epoch": 1.3927550047664443, "grad_norm": 5.9347345665854085, "learning_rate": 6.490160898146919e-06, "loss": 2.5924, "step": 731 }, { "epoch": 1.3946615824594852, "grad_norm": 5.6660392155226145, "learning_rate": 6.4795756440080095e-06, "loss": 2.6495, "step": 732 }, { "epoch": 1.3965681601525262, "grad_norm": 6.132373147316881, "learning_rate": 6.468983117146204e-06, "loss": 2.6882, "step": 733 }, { "epoch": 1.398474737845567, "grad_norm": 5.999644007797458, "learning_rate": 6.458383369628125e-06, "loss": 2.5887, "step": 734 }, { "epoch": 1.4003813155386082, "grad_norm": 5.755500401685643, "learning_rate": 6.447776453555896e-06, "loss": 2.6826, "step": 735 }, { "epoch": 1.4022878932316492, "grad_norm": 5.651125412297895, "learning_rate": 6.437162421066871e-06, "loss": 2.5738, "step": 736 }, { "epoch": 1.4041944709246903, "grad_norm": 5.9511827779499935, "learning_rate": 6.426541324333384e-06, "loss": 2.7265, "step": 737 }, { "epoch": 1.4061010486177312, "grad_norm": 5.723051682843794, "learning_rate": 6.415913215562496e-06, "loss": 2.6717, "step": 738 }, { "epoch": 1.4080076263107721, "grad_norm": 5.154373887822125, "learning_rate": 6.405278146995733e-06, "loss": 2.8204, "step": 739 }, { "epoch": 1.409914204003813, "grad_norm": 7.074667877835955, "learning_rate": 6.394636170908831e-06, "loss": 2.8492, "step": 740 }, { "epoch": 1.4118207816968542, "grad_norm": 5.971906266968128, "learning_rate": 6.383987339611482e-06, "loss": 2.662, "step": 741 }, { "epoch": 1.4137273593898951, "grad_norm": 6.411929991867859, "learning_rate": 6.37333170544707e-06, "loss": 2.6032, "step": 742 }, { "epoch": 1.415633937082936, "grad_norm": 5.655402252403605, "learning_rate": 6.362669320792419e-06, "loss": 2.684, "step": 743 }, { "epoch": 1.4175405147759772, "grad_norm": 6.212968746516409, "learning_rate": 6.3520002380575395e-06, "loss": 2.7098, "step": 744 }, { "epoch": 1.4194470924690181, "grad_norm": 5.181299270341057, "learning_rate": 6.3413245096853585e-06, "loss": 2.7347, "step": 745 }, { "epoch": 1.421353670162059, "grad_norm": 5.755729920235851, "learning_rate": 6.330642188151473e-06, "loss": 2.6627, "step": 746 }, { "epoch": 1.4232602478551, "grad_norm": 5.978046474191513, "learning_rate": 6.3199533259638875e-06, "loss": 2.6977, "step": 747 }, { "epoch": 1.4251668255481411, "grad_norm": 6.228899302178255, "learning_rate": 6.309257975662756e-06, "loss": 2.7831, "step": 748 }, { "epoch": 1.427073403241182, "grad_norm": 6.046848608918224, "learning_rate": 6.2985561898201264e-06, "loss": 2.7359, "step": 749 }, { "epoch": 1.4289799809342232, "grad_norm": 6.276694917282604, "learning_rate": 6.287848021039676e-06, "loss": 2.7874, "step": 750 }, { "epoch": 1.430886558627264, "grad_norm": 6.403841132327003, "learning_rate": 6.27713352195646e-06, "loss": 2.8274, "step": 751 }, { "epoch": 1.432793136320305, "grad_norm": 6.288903897971148, "learning_rate": 6.266412745236649e-06, "loss": 2.6776, "step": 752 }, { "epoch": 1.434699714013346, "grad_norm": 5.223816678746735, "learning_rate": 6.255685743577269e-06, "loss": 2.5998, "step": 753 }, { "epoch": 1.436606291706387, "grad_norm": 5.664215593903397, "learning_rate": 6.2449525697059475e-06, "loss": 2.6117, "step": 754 }, { "epoch": 1.438512869399428, "grad_norm": 5.965168521378787, "learning_rate": 6.234213276380647e-06, "loss": 2.5739, "step": 755 }, { "epoch": 1.440419447092469, "grad_norm": 5.4414963169371395, "learning_rate": 6.223467916389415e-06, "loss": 2.845, "step": 756 }, { "epoch": 1.44232602478551, "grad_norm": 5.896174426320067, "learning_rate": 6.212716542550112e-06, "loss": 2.9534, "step": 757 }, { "epoch": 1.444232602478551, "grad_norm": 6.501933194231186, "learning_rate": 6.201959207710164e-06, "loss": 2.6549, "step": 758 }, { "epoch": 1.446139180171592, "grad_norm": 6.061895627638344, "learning_rate": 6.191195964746299e-06, "loss": 2.786, "step": 759 }, { "epoch": 1.4480457578646329, "grad_norm": 5.3087679169381605, "learning_rate": 6.180426866564281e-06, "loss": 2.6406, "step": 760 }, { "epoch": 1.449952335557674, "grad_norm": 5.709449156984866, "learning_rate": 6.169651966098659e-06, "loss": 2.7635, "step": 761 }, { "epoch": 1.451858913250715, "grad_norm": 6.105138368422822, "learning_rate": 6.158871316312499e-06, "loss": 2.854, "step": 762 }, { "epoch": 1.453765490943756, "grad_norm": 6.127651855848987, "learning_rate": 6.14808497019713e-06, "loss": 2.7344, "step": 763 }, { "epoch": 1.455672068636797, "grad_norm": 7.093132530970147, "learning_rate": 6.137292980771881e-06, "loss": 2.9585, "step": 764 }, { "epoch": 1.457578646329838, "grad_norm": 5.519159712870281, "learning_rate": 6.126495401083818e-06, "loss": 2.7114, "step": 765 }, { "epoch": 1.4594852240228788, "grad_norm": 5.9685773047620865, "learning_rate": 6.115692284207488e-06, "loss": 2.6816, "step": 766 }, { "epoch": 1.46139180171592, "grad_norm": 5.61044144976271, "learning_rate": 6.104883683244651e-06, "loss": 2.7327, "step": 767 }, { "epoch": 1.463298379408961, "grad_norm": 6.3202995378707945, "learning_rate": 6.094069651324029e-06, "loss": 2.837, "step": 768 }, { "epoch": 1.4652049571020018, "grad_norm": 6.103504378575533, "learning_rate": 6.083250241601039e-06, "loss": 2.7335, "step": 769 }, { "epoch": 1.467111534795043, "grad_norm": 5.4112456161614695, "learning_rate": 6.072425507257528e-06, "loss": 2.859, "step": 770 }, { "epoch": 1.469018112488084, "grad_norm": 6.05734520771955, "learning_rate": 6.061595501501518e-06, "loss": 2.845, "step": 771 }, { "epoch": 1.4709246901811248, "grad_norm": 5.980874870973497, "learning_rate": 6.0507602775669415e-06, "loss": 2.7457, "step": 772 }, { "epoch": 1.4728312678741657, "grad_norm": 6.366293784882874, "learning_rate": 6.039919888713383e-06, "loss": 2.8232, "step": 773 }, { "epoch": 1.4747378455672069, "grad_norm": 5.62336556424146, "learning_rate": 6.029074388225811e-06, "loss": 2.608, "step": 774 }, { "epoch": 1.4766444232602478, "grad_norm": 6.631927774151473, "learning_rate": 6.018223829414322e-06, "loss": 2.8441, "step": 775 }, { "epoch": 1.478551000953289, "grad_norm": 5.632442210846927, "learning_rate": 6.007368265613877e-06, "loss": 2.6216, "step": 776 }, { "epoch": 1.4804575786463299, "grad_norm": 5.912222683725123, "learning_rate": 5.996507750184036e-06, "loss": 2.7868, "step": 777 }, { "epoch": 1.4823641563393708, "grad_norm": 6.098847057784483, "learning_rate": 5.9856423365087e-06, "loss": 2.8223, "step": 778 }, { "epoch": 1.4842707340324117, "grad_norm": 5.823875161843027, "learning_rate": 5.974772077995848e-06, "loss": 2.7731, "step": 779 }, { "epoch": 1.4861773117254529, "grad_norm": 5.39279539747426, "learning_rate": 5.96389702807727e-06, "loss": 2.7864, "step": 780 }, { "epoch": 1.4880838894184938, "grad_norm": 6.002256577188399, "learning_rate": 5.953017240208309e-06, "loss": 2.7403, "step": 781 }, { "epoch": 1.489990467111535, "grad_norm": 6.270172177008326, "learning_rate": 5.9421327678676025e-06, "loss": 2.7254, "step": 782 }, { "epoch": 1.4918970448045759, "grad_norm": 5.493437287263334, "learning_rate": 5.931243664556803e-06, "loss": 2.7308, "step": 783 }, { "epoch": 1.4938036224976168, "grad_norm": 5.8277139104845945, "learning_rate": 5.9203499838003365e-06, "loss": 2.6574, "step": 784 }, { "epoch": 1.4957102001906577, "grad_norm": 5.684005537318376, "learning_rate": 5.9094517791451234e-06, "loss": 2.7519, "step": 785 }, { "epoch": 1.4976167778836986, "grad_norm": 5.894008692923786, "learning_rate": 5.898549104160322e-06, "loss": 2.7396, "step": 786 }, { "epoch": 1.4995233555767398, "grad_norm": 6.132872983121301, "learning_rate": 5.887642012437066e-06, "loss": 2.8907, "step": 787 }, { "epoch": 1.5014299332697807, "grad_norm": 5.990744818261809, "learning_rate": 5.876730557588196e-06, "loss": 2.6961, "step": 788 }, { "epoch": 1.5033365109628218, "grad_norm": 5.680663930682165, "learning_rate": 5.865814793248003e-06, "loss": 2.8058, "step": 789 }, { "epoch": 1.5052430886558628, "grad_norm": 5.992177798875706, "learning_rate": 5.854894773071957e-06, "loss": 2.8244, "step": 790 }, { "epoch": 1.5071496663489037, "grad_norm": 6.285187556645649, "learning_rate": 5.84397055073645e-06, "loss": 2.5896, "step": 791 }, { "epoch": 1.5090562440419446, "grad_norm": 5.584910504268757, "learning_rate": 5.833042179938528e-06, "loss": 2.7272, "step": 792 }, { "epoch": 1.5109628217349858, "grad_norm": 6.3206350761098244, "learning_rate": 5.822109714395629e-06, "loss": 2.8347, "step": 793 }, { "epoch": 1.5128693994280267, "grad_norm": 6.439747495404946, "learning_rate": 5.811173207845318e-06, "loss": 2.7115, "step": 794 }, { "epoch": 1.5147759771210678, "grad_norm": 5.25948033639652, "learning_rate": 5.800232714045022e-06, "loss": 2.6703, "step": 795 }, { "epoch": 1.5166825548141087, "grad_norm": 5.622374381777427, "learning_rate": 5.7892882867717705e-06, "loss": 2.6117, "step": 796 }, { "epoch": 1.5185891325071497, "grad_norm": 5.40093342607045, "learning_rate": 5.778339979821924e-06, "loss": 2.6084, "step": 797 }, { "epoch": 1.5204957102001906, "grad_norm": 5.4644991767928, "learning_rate": 5.7673878470109146e-06, "loss": 2.4377, "step": 798 }, { "epoch": 1.5224022878932315, "grad_norm": 5.678471450997284, "learning_rate": 5.756431942172982e-06, "loss": 2.691, "step": 799 }, { "epoch": 1.5243088655862727, "grad_norm": 6.057563305189675, "learning_rate": 5.745472319160902e-06, "loss": 2.7778, "step": 800 }, { "epoch": 1.5262154432793136, "grad_norm": 6.235510239784785, "learning_rate": 5.734509031845731e-06, "loss": 2.8622, "step": 801 }, { "epoch": 1.5281220209723547, "grad_norm": 5.7564305925173676, "learning_rate": 5.723542134116537e-06, "loss": 2.8732, "step": 802 }, { "epoch": 1.5300285986653956, "grad_norm": 5.773565036373122, "learning_rate": 5.712571679880134e-06, "loss": 2.7402, "step": 803 }, { "epoch": 1.5319351763584366, "grad_norm": 5.687511616661107, "learning_rate": 5.7015977230608145e-06, "loss": 2.6716, "step": 804 }, { "epoch": 1.5338417540514775, "grad_norm": 5.818996285546961, "learning_rate": 5.690620317600091e-06, "loss": 2.7043, "step": 805 }, { "epoch": 1.5357483317445186, "grad_norm": 6.075204180653496, "learning_rate": 5.679639517456427e-06, "loss": 2.7648, "step": 806 }, { "epoch": 1.5376549094375596, "grad_norm": 5.327676662738987, "learning_rate": 5.668655376604973e-06, "loss": 2.6771, "step": 807 }, { "epoch": 1.5395614871306007, "grad_norm": 5.864900295558339, "learning_rate": 5.6576679490373e-06, "loss": 2.8991, "step": 808 }, { "epoch": 1.5414680648236416, "grad_norm": 6.152767668199201, "learning_rate": 5.646677288761132e-06, "loss": 2.7629, "step": 809 }, { "epoch": 1.5433746425166825, "grad_norm": 5.923870868625676, "learning_rate": 5.635683449800086e-06, "loss": 2.6482, "step": 810 }, { "epoch": 1.5452812202097235, "grad_norm": 5.744519953556653, "learning_rate": 5.624686486193401e-06, "loss": 2.638, "step": 811 }, { "epoch": 1.5471877979027644, "grad_norm": 5.489372510446527, "learning_rate": 5.6136864519956815e-06, "loss": 2.8122, "step": 812 }, { "epoch": 1.5490943755958055, "grad_norm": 6.395486720706888, "learning_rate": 5.6026834012766155e-06, "loss": 2.6911, "step": 813 }, { "epoch": 1.5510009532888467, "grad_norm": 6.147184360906165, "learning_rate": 5.591677388120724e-06, "loss": 2.7529, "step": 814 }, { "epoch": 1.5529075309818876, "grad_norm": 5.665542260706258, "learning_rate": 5.580668466627093e-06, "loss": 2.6765, "step": 815 }, { "epoch": 1.5548141086749285, "grad_norm": 5.7490599710015, "learning_rate": 5.569656690909095e-06, "loss": 2.5651, "step": 816 }, { "epoch": 1.5567206863679695, "grad_norm": 5.592105516499953, "learning_rate": 5.558642115094138e-06, "loss": 2.7178, "step": 817 }, { "epoch": 1.5586272640610104, "grad_norm": 5.639069784203279, "learning_rate": 5.547624793323395e-06, "loss": 2.8242, "step": 818 }, { "epoch": 1.5605338417540515, "grad_norm": 5.8640866192293055, "learning_rate": 5.5366047797515335e-06, "loss": 2.74, "step": 819 }, { "epoch": 1.5624404194470924, "grad_norm": 5.768830178908581, "learning_rate": 5.525582128546453e-06, "loss": 2.8348, "step": 820 }, { "epoch": 1.5643469971401336, "grad_norm": 5.727661096570827, "learning_rate": 5.514556893889017e-06, "loss": 2.5892, "step": 821 }, { "epoch": 1.5662535748331745, "grad_norm": 5.476463613639716, "learning_rate": 5.503529129972792e-06, "loss": 2.6615, "step": 822 }, { "epoch": 1.5681601525262154, "grad_norm": 5.3899840353457655, "learning_rate": 5.492498891003772e-06, "loss": 2.6448, "step": 823 }, { "epoch": 1.5700667302192564, "grad_norm": 5.721184299081477, "learning_rate": 5.4814662312001166e-06, "loss": 2.6235, "step": 824 }, { "epoch": 1.5719733079122973, "grad_norm": 5.785231216266448, "learning_rate": 5.4704312047918905e-06, "loss": 2.6968, "step": 825 }, { "epoch": 1.5738798856053384, "grad_norm": 6.010246205555324, "learning_rate": 5.459393866020785e-06, "loss": 2.7008, "step": 826 }, { "epoch": 1.5757864632983796, "grad_norm": 5.751909493386484, "learning_rate": 5.448354269139863e-06, "loss": 2.6018, "step": 827 }, { "epoch": 1.5776930409914205, "grad_norm": 5.746938002579965, "learning_rate": 5.4373124684132804e-06, "loss": 2.7153, "step": 828 }, { "epoch": 1.5795996186844614, "grad_norm": 6.065226028475929, "learning_rate": 5.426268518116032e-06, "loss": 2.8545, "step": 829 }, { "epoch": 1.5815061963775023, "grad_norm": 5.500925572589964, "learning_rate": 5.415222472533679e-06, "loss": 2.767, "step": 830 }, { "epoch": 1.5834127740705433, "grad_norm": 5.319832620131874, "learning_rate": 5.404174385962073e-06, "loss": 2.749, "step": 831 }, { "epoch": 1.5853193517635844, "grad_norm": 5.1710065803615715, "learning_rate": 5.39312431270711e-06, "loss": 2.6481, "step": 832 }, { "epoch": 1.5872259294566253, "grad_norm": 5.303006827799691, "learning_rate": 5.382072307084441e-06, "loss": 2.8924, "step": 833 }, { "epoch": 1.5891325071496665, "grad_norm": 5.7717364196055865, "learning_rate": 5.37101842341922e-06, "loss": 2.6917, "step": 834 }, { "epoch": 1.5910390848427074, "grad_norm": 5.584864535859132, "learning_rate": 5.359962716045836e-06, "loss": 2.6375, "step": 835 }, { "epoch": 1.5929456625357483, "grad_norm": 5.737697087815125, "learning_rate": 5.3489052393076345e-06, "loss": 2.6626, "step": 836 }, { "epoch": 1.5948522402287892, "grad_norm": 5.760919274540062, "learning_rate": 5.337846047556664e-06, "loss": 2.8719, "step": 837 }, { "epoch": 1.5967588179218302, "grad_norm": 5.569233396278203, "learning_rate": 5.3267851951534e-06, "loss": 2.6789, "step": 838 }, { "epoch": 1.5986653956148713, "grad_norm": 5.604334059981682, "learning_rate": 5.315722736466482e-06, "loss": 2.8562, "step": 839 }, { "epoch": 1.6005719733079125, "grad_norm": 6.059301329910365, "learning_rate": 5.304658725872447e-06, "loss": 2.7033, "step": 840 }, { "epoch": 1.6024785510009534, "grad_norm": 6.715292356612858, "learning_rate": 5.293593217755455e-06, "loss": 2.7619, "step": 841 }, { "epoch": 1.6043851286939943, "grad_norm": 5.659678507711031, "learning_rate": 5.282526266507032e-06, "loss": 2.661, "step": 842 }, { "epoch": 1.6062917063870352, "grad_norm": 5.496907326476709, "learning_rate": 5.2714579265257935e-06, "loss": 2.7248, "step": 843 }, { "epoch": 1.6081982840800761, "grad_norm": 5.8435533108390345, "learning_rate": 5.260388252217186e-06, "loss": 2.7522, "step": 844 }, { "epoch": 1.6101048617731173, "grad_norm": 5.911539029352856, "learning_rate": 5.249317297993211e-06, "loss": 2.7759, "step": 845 }, { "epoch": 1.6120114394661582, "grad_norm": 5.935089879169347, "learning_rate": 5.2382451182721615e-06, "loss": 2.7333, "step": 846 }, { "epoch": 1.6139180171591994, "grad_norm": 5.793044803102689, "learning_rate": 5.227171767478357e-06, "loss": 2.7285, "step": 847 }, { "epoch": 1.6158245948522403, "grad_norm": 5.808184309140038, "learning_rate": 5.21609730004187e-06, "loss": 2.6984, "step": 848 }, { "epoch": 1.6177311725452812, "grad_norm": 5.894564511691245, "learning_rate": 5.205021770398264e-06, "loss": 2.7979, "step": 849 }, { "epoch": 1.6196377502383221, "grad_norm": 6.1587665401137155, "learning_rate": 5.193945232988324e-06, "loss": 2.9181, "step": 850 }, { "epoch": 1.621544327931363, "grad_norm": 5.858991331651028, "learning_rate": 5.182867742257787e-06, "loss": 2.7021, "step": 851 }, { "epoch": 1.6234509056244042, "grad_norm": 5.310160406699071, "learning_rate": 5.171789352657074e-06, "loss": 2.7262, "step": 852 }, { "epoch": 1.6253574833174453, "grad_norm": 6.408595868954976, "learning_rate": 5.160710118641031e-06, "loss": 2.6642, "step": 853 }, { "epoch": 1.6272640610104863, "grad_norm": 6.538112563551994, "learning_rate": 5.149630094668647e-06, "loss": 2.8788, "step": 854 }, { "epoch": 1.6291706387035272, "grad_norm": 5.6910910516655, "learning_rate": 5.138549335202801e-06, "loss": 2.7394, "step": 855 }, { "epoch": 1.631077216396568, "grad_norm": 5.3642748932646365, "learning_rate": 5.12746789470998e-06, "loss": 2.7684, "step": 856 }, { "epoch": 1.632983794089609, "grad_norm": 5.35811077969871, "learning_rate": 5.116385827660024e-06, "loss": 2.6408, "step": 857 }, { "epoch": 1.6348903717826502, "grad_norm": 5.84290151911092, "learning_rate": 5.105303188525852e-06, "loss": 2.773, "step": 858 }, { "epoch": 1.636796949475691, "grad_norm": 6.423076162990959, "learning_rate": 5.094220031783191e-06, "loss": 2.7431, "step": 859 }, { "epoch": 1.6387035271687322, "grad_norm": 5.453818518752896, "learning_rate": 5.0831364119103196e-06, "loss": 2.5913, "step": 860 }, { "epoch": 1.6406101048617732, "grad_norm": 5.2404653400681855, "learning_rate": 5.072052383387787e-06, "loss": 2.5809, "step": 861 }, { "epoch": 1.642516682554814, "grad_norm": 5.736347822799904, "learning_rate": 5.06096800069815e-06, "loss": 2.809, "step": 862 }, { "epoch": 1.644423260247855, "grad_norm": 6.037579365161111, "learning_rate": 5.0498833183257125e-06, "loss": 3.0127, "step": 863 }, { "epoch": 1.646329837940896, "grad_norm": 6.136059015966307, "learning_rate": 5.038798390756246e-06, "loss": 2.8152, "step": 864 }, { "epoch": 1.648236415633937, "grad_norm": 6.180685424395929, "learning_rate": 5.027713272476729e-06, "loss": 2.9138, "step": 865 }, { "epoch": 1.6501429933269782, "grad_norm": 6.959161105423599, "learning_rate": 5.016628017975078e-06, "loss": 2.6766, "step": 866 }, { "epoch": 1.6520495710200191, "grad_norm": 5.6371183600159105, "learning_rate": 5.005542681739879e-06, "loss": 2.8009, "step": 867 }, { "epoch": 1.65395614871306, "grad_norm": 5.8242939714679505, "learning_rate": 4.994457318260122e-06, "loss": 2.6397, "step": 868 }, { "epoch": 1.655862726406101, "grad_norm": 6.026996193503577, "learning_rate": 4.983371982024922e-06, "loss": 2.9047, "step": 869 }, { "epoch": 1.657769304099142, "grad_norm": 5.733481238299455, "learning_rate": 4.972286727523272e-06, "loss": 2.8499, "step": 870 }, { "epoch": 1.659675881792183, "grad_norm": 6.32051037384349, "learning_rate": 4.961201609243755e-06, "loss": 2.853, "step": 871 }, { "epoch": 1.661582459485224, "grad_norm": 5.911190752531471, "learning_rate": 4.95011668167429e-06, "loss": 2.5472, "step": 872 }, { "epoch": 1.6634890371782651, "grad_norm": 5.630039323820939, "learning_rate": 4.9390319993018506e-06, "loss": 2.7837, "step": 873 }, { "epoch": 1.665395614871306, "grad_norm": 6.498252665592895, "learning_rate": 4.927947616612216e-06, "loss": 2.6638, "step": 874 }, { "epoch": 1.667302192564347, "grad_norm": 5.713391572042626, "learning_rate": 4.916863588089681e-06, "loss": 2.7325, "step": 875 }, { "epoch": 1.669208770257388, "grad_norm": 6.5181404575612945, "learning_rate": 4.90577996821681e-06, "loss": 2.708, "step": 876 }, { "epoch": 1.6711153479504288, "grad_norm": 5.6450399620929455, "learning_rate": 4.894696811474148e-06, "loss": 2.8138, "step": 877 }, { "epoch": 1.67302192564347, "grad_norm": 5.877989186251913, "learning_rate": 4.883614172339977e-06, "loss": 2.8075, "step": 878 }, { "epoch": 1.674928503336511, "grad_norm": 5.955954179627267, "learning_rate": 4.87253210529002e-06, "loss": 2.9097, "step": 879 }, { "epoch": 1.676835081029552, "grad_norm": 5.619732087501845, "learning_rate": 4.861450664797201e-06, "loss": 2.8289, "step": 880 }, { "epoch": 1.678741658722593, "grad_norm": 6.181660479156348, "learning_rate": 4.850369905331353e-06, "loss": 2.7614, "step": 881 }, { "epoch": 1.6806482364156339, "grad_norm": 5.9171498190259495, "learning_rate": 4.8392898813589705e-06, "loss": 2.5694, "step": 882 }, { "epoch": 1.6825548141086748, "grad_norm": 6.495241760939766, "learning_rate": 4.828210647342926e-06, "loss": 2.7078, "step": 883 }, { "epoch": 1.684461391801716, "grad_norm": 5.443899501316038, "learning_rate": 4.817132257742216e-06, "loss": 2.6713, "step": 884 }, { "epoch": 1.6863679694947569, "grad_norm": 5.816896741853167, "learning_rate": 4.806054767011677e-06, "loss": 2.7651, "step": 885 }, { "epoch": 1.688274547187798, "grad_norm": 6.23430329314366, "learning_rate": 4.794978229601737e-06, "loss": 2.7186, "step": 886 }, { "epoch": 1.690181124880839, "grad_norm": 5.867082974519474, "learning_rate": 4.78390269995813e-06, "loss": 2.6918, "step": 887 }, { "epoch": 1.6920877025738799, "grad_norm": 5.680550336488418, "learning_rate": 4.772828232521644e-06, "loss": 2.6137, "step": 888 }, { "epoch": 1.6939942802669208, "grad_norm": 5.6834350173970485, "learning_rate": 4.761754881727839e-06, "loss": 2.833, "step": 889 }, { "epoch": 1.6959008579599617, "grad_norm": 5.896621473614477, "learning_rate": 4.750682702006791e-06, "loss": 2.8249, "step": 890 }, { "epoch": 1.6978074356530028, "grad_norm": 5.65318489853198, "learning_rate": 4.739611747782815e-06, "loss": 2.9427, "step": 891 }, { "epoch": 1.699714013346044, "grad_norm": 6.135607119463372, "learning_rate": 4.728542073474208e-06, "loss": 2.7646, "step": 892 }, { "epoch": 1.701620591039085, "grad_norm": 6.565582703094557, "learning_rate": 4.71747373349297e-06, "loss": 2.9197, "step": 893 }, { "epoch": 1.7035271687321258, "grad_norm": 6.3466033232272965, "learning_rate": 4.706406782244547e-06, "loss": 2.6395, "step": 894 }, { "epoch": 1.7054337464251668, "grad_norm": 5.803936288422345, "learning_rate": 4.695341274127553e-06, "loss": 2.7804, "step": 895 }, { "epoch": 1.7073403241182077, "grad_norm": 6.33729078266259, "learning_rate": 4.6842772635335185e-06, "loss": 2.6453, "step": 896 }, { "epoch": 1.7092469018112488, "grad_norm": 6.228485725407679, "learning_rate": 4.6732148048466e-06, "loss": 2.6868, "step": 897 }, { "epoch": 1.7111534795042898, "grad_norm": 5.582329671869577, "learning_rate": 4.662153952443337e-06, "loss": 2.7435, "step": 898 }, { "epoch": 1.713060057197331, "grad_norm": 6.196119257856374, "learning_rate": 4.6510947606923654e-06, "loss": 2.8192, "step": 899 }, { "epoch": 1.7149666348903718, "grad_norm": 6.438003303727222, "learning_rate": 4.640037283954165e-06, "loss": 2.7597, "step": 900 }, { "epoch": 1.7168732125834127, "grad_norm": 6.371282584013763, "learning_rate": 4.62898157658078e-06, "loss": 2.7625, "step": 901 }, { "epoch": 1.7187797902764537, "grad_norm": 6.4779540896728465, "learning_rate": 4.617927692915562e-06, "loss": 2.7562, "step": 902 }, { "epoch": 1.7206863679694948, "grad_norm": 5.638262150878674, "learning_rate": 4.6068756872928924e-06, "loss": 2.7734, "step": 903 }, { "epoch": 1.7225929456625357, "grad_norm": 6.0738484527130785, "learning_rate": 4.5958256140379285e-06, "loss": 2.7153, "step": 904 }, { "epoch": 1.7244995233555769, "grad_norm": 5.952768325888439, "learning_rate": 4.584777527466322e-06, "loss": 2.6835, "step": 905 }, { "epoch": 1.7264061010486178, "grad_norm": 6.42063574173481, "learning_rate": 4.573731481883969e-06, "loss": 2.8235, "step": 906 }, { "epoch": 1.7283126787416587, "grad_norm": 5.892234989103079, "learning_rate": 4.5626875315867195e-06, "loss": 2.7347, "step": 907 }, { "epoch": 1.7302192564346996, "grad_norm": 5.3891518327973476, "learning_rate": 4.5516457308601385e-06, "loss": 2.7912, "step": 908 }, { "epoch": 1.7321258341277406, "grad_norm": 6.209230283040101, "learning_rate": 4.540606133979215e-06, "loss": 2.8207, "step": 909 }, { "epoch": 1.7340324118207817, "grad_norm": 6.164437268835737, "learning_rate": 4.529568795208111e-06, "loss": 2.6417, "step": 910 }, { "epoch": 1.7359389895138226, "grad_norm": 5.264035490247147, "learning_rate": 4.518533768799884e-06, "loss": 2.6562, "step": 911 }, { "epoch": 1.7378455672068638, "grad_norm": 5.67887462789405, "learning_rate": 4.507501108996231e-06, "loss": 2.6888, "step": 912 }, { "epoch": 1.7397521448999047, "grad_norm": 5.413388193897916, "learning_rate": 4.496470870027209e-06, "loss": 2.4744, "step": 913 }, { "epoch": 1.7416587225929456, "grad_norm": 6.118534963692724, "learning_rate": 4.485443106110984e-06, "loss": 2.7412, "step": 914 }, { "epoch": 1.7435653002859866, "grad_norm": 6.175313098238824, "learning_rate": 4.474417871453547e-06, "loss": 2.6447, "step": 915 }, { "epoch": 1.7454718779790277, "grad_norm": 6.1003348039699326, "learning_rate": 4.463395220248468e-06, "loss": 2.7521, "step": 916 }, { "epoch": 1.7473784556720686, "grad_norm": 5.693270761880058, "learning_rate": 4.452375206676605e-06, "loss": 2.7203, "step": 917 }, { "epoch": 1.7492850333651098, "grad_norm": 5.504591203621616, "learning_rate": 4.441357884905862e-06, "loss": 2.721, "step": 918 }, { "epoch": 1.7511916110581507, "grad_norm": 5.55493845852455, "learning_rate": 4.430343309090906e-06, "loss": 2.6451, "step": 919 }, { "epoch": 1.7530981887511916, "grad_norm": 6.656574118496605, "learning_rate": 4.419331533372909e-06, "loss": 2.6199, "step": 920 }, { "epoch": 1.7550047664442325, "grad_norm": 6.08395107661724, "learning_rate": 4.408322611879276e-06, "loss": 2.5586, "step": 921 }, { "epoch": 1.7569113441372735, "grad_norm": 6.969971085412462, "learning_rate": 4.397316598723385e-06, "loss": 2.8778, "step": 922 }, { "epoch": 1.7588179218303146, "grad_norm": 6.000252057863175, "learning_rate": 4.386313548004319e-06, "loss": 2.6826, "step": 923 }, { "epoch": 1.7607244995233555, "grad_norm": 5.473286476736256, "learning_rate": 4.3753135138065995e-06, "loss": 2.7517, "step": 924 }, { "epoch": 1.7626310772163967, "grad_norm": 6.173735772303119, "learning_rate": 4.364316550199915e-06, "loss": 2.8189, "step": 925 }, { "epoch": 1.7645376549094376, "grad_norm": 5.426954332118352, "learning_rate": 4.3533227112388694e-06, "loss": 2.6888, "step": 926 }, { "epoch": 1.7664442326024785, "grad_norm": 6.519238945457629, "learning_rate": 4.342332050962701e-06, "loss": 2.7854, "step": 927 }, { "epoch": 1.7683508102955194, "grad_norm": 5.607534480106626, "learning_rate": 4.3313446233950275e-06, "loss": 2.7972, "step": 928 }, { "epoch": 1.7702573879885606, "grad_norm": 5.910010799081079, "learning_rate": 4.320360482543573e-06, "loss": 2.8217, "step": 929 }, { "epoch": 1.7721639656816015, "grad_norm": 6.065758147446333, "learning_rate": 4.309379682399911e-06, "loss": 2.7451, "step": 930 }, { "epoch": 1.7740705433746427, "grad_norm": 5.817341375017382, "learning_rate": 4.298402276939187e-06, "loss": 2.7077, "step": 931 }, { "epoch": 1.7759771210676836, "grad_norm": 5.687693809686674, "learning_rate": 4.287428320119869e-06, "loss": 2.7196, "step": 932 }, { "epoch": 1.7778836987607245, "grad_norm": 5.64490344607854, "learning_rate": 4.276457865883462e-06, "loss": 2.7989, "step": 933 }, { "epoch": 1.7797902764537654, "grad_norm": 5.892419195178762, "learning_rate": 4.26549096815427e-06, "loss": 2.7043, "step": 934 }, { "epoch": 1.7816968541468063, "grad_norm": 5.699924783188236, "learning_rate": 4.254527680839098e-06, "loss": 2.6613, "step": 935 }, { "epoch": 1.7836034318398475, "grad_norm": 5.746746003468564, "learning_rate": 4.243568057827019e-06, "loss": 2.7629, "step": 936 }, { "epoch": 1.7855100095328886, "grad_norm": 5.985378756375549, "learning_rate": 4.232612152989085e-06, "loss": 2.7474, "step": 937 }, { "epoch": 1.7874165872259296, "grad_norm": 6.136916165061805, "learning_rate": 4.221660020178077e-06, "loss": 2.7454, "step": 938 }, { "epoch": 1.7893231649189705, "grad_norm": 5.608912605254785, "learning_rate": 4.21071171322823e-06, "loss": 2.5537, "step": 939 }, { "epoch": 1.7912297426120114, "grad_norm": 5.932331256887867, "learning_rate": 4.199767285954979e-06, "loss": 2.7312, "step": 940 }, { "epoch": 1.7931363203050523, "grad_norm": 5.637570707287579, "learning_rate": 4.188826792154683e-06, "loss": 2.7509, "step": 941 }, { "epoch": 1.7950428979980935, "grad_norm": 6.2007986760505975, "learning_rate": 4.177890285604372e-06, "loss": 2.73, "step": 942 }, { "epoch": 1.7969494756911344, "grad_norm": 5.767001803692692, "learning_rate": 4.166957820061472e-06, "loss": 2.691, "step": 943 }, { "epoch": 1.7988560533841755, "grad_norm": 6.110354651846191, "learning_rate": 4.156029449263552e-06, "loss": 2.842, "step": 944 }, { "epoch": 1.8007626310772165, "grad_norm": 5.754823171205444, "learning_rate": 4.145105226928043e-06, "loss": 2.5161, "step": 945 }, { "epoch": 1.8026692087702574, "grad_norm": 5.978498971473416, "learning_rate": 4.134185206751998e-06, "loss": 2.734, "step": 946 }, { "epoch": 1.8045757864632983, "grad_norm": 5.960602448944853, "learning_rate": 4.1232694424118044e-06, "loss": 2.6968, "step": 947 }, { "epoch": 1.8064823641563392, "grad_norm": 6.193766366678732, "learning_rate": 4.1123579875629355e-06, "loss": 2.8224, "step": 948 }, { "epoch": 1.8083889418493804, "grad_norm": 6.118326574971906, "learning_rate": 4.101450895839679e-06, "loss": 2.6899, "step": 949 }, { "epoch": 1.8102955195424215, "grad_norm": 6.358599571965752, "learning_rate": 4.090548220854878e-06, "loss": 2.7708, "step": 950 }, { "epoch": 1.8122020972354624, "grad_norm": 6.502962827298543, "learning_rate": 4.079650016199665e-06, "loss": 2.7148, "step": 951 }, { "epoch": 1.8141086749285034, "grad_norm": 6.951162047592987, "learning_rate": 4.0687563354431986e-06, "loss": 2.7162, "step": 952 }, { "epoch": 1.8160152526215443, "grad_norm": 6.041286872542385, "learning_rate": 4.057867232132398e-06, "loss": 2.7212, "step": 953 }, { "epoch": 1.8179218303145852, "grad_norm": 6.085246174308792, "learning_rate": 4.046982759791691e-06, "loss": 2.8252, "step": 954 }, { "epoch": 1.8198284080076264, "grad_norm": 6.226178608171546, "learning_rate": 4.036102971922731e-06, "loss": 2.8065, "step": 955 }, { "epoch": 1.8217349857006673, "grad_norm": 6.998711581891694, "learning_rate": 4.025227922004153e-06, "loss": 2.5846, "step": 956 }, { "epoch": 1.8236415633937084, "grad_norm": 5.798574468994434, "learning_rate": 4.0143576634913e-06, "loss": 2.6381, "step": 957 }, { "epoch": 1.8255481410867493, "grad_norm": 5.936128462348273, "learning_rate": 4.003492249815965e-06, "loss": 2.5787, "step": 958 }, { "epoch": 1.8274547187797903, "grad_norm": 5.898822680713807, "learning_rate": 3.992631734386124e-06, "loss": 2.916, "step": 959 }, { "epoch": 1.8293612964728312, "grad_norm": 6.237214287608162, "learning_rate": 3.9817761705856795e-06, "loss": 2.8031, "step": 960 }, { "epoch": 1.8312678741658721, "grad_norm": 6.344306668206571, "learning_rate": 3.97092561177419e-06, "loss": 2.8377, "step": 961 }, { "epoch": 1.8331744518589133, "grad_norm": 6.34299376707076, "learning_rate": 3.96008011128662e-06, "loss": 2.6898, "step": 962 }, { "epoch": 1.8350810295519544, "grad_norm": 6.406817594081465, "learning_rate": 3.949239722433058e-06, "loss": 2.6378, "step": 963 }, { "epoch": 1.8369876072449953, "grad_norm": 5.530513493317788, "learning_rate": 3.938404498498483e-06, "loss": 2.7566, "step": 964 }, { "epoch": 1.8388941849380362, "grad_norm": 6.424229277087239, "learning_rate": 3.927574492742473e-06, "loss": 2.8311, "step": 965 }, { "epoch": 1.8408007626310772, "grad_norm": 6.066800444017284, "learning_rate": 3.916749758398962e-06, "loss": 2.7685, "step": 966 }, { "epoch": 1.842707340324118, "grad_norm": 5.96422428774266, "learning_rate": 3.905930348675971e-06, "loss": 2.5928, "step": 967 }, { "epoch": 1.8446139180171592, "grad_norm": 5.9998466562060555, "learning_rate": 3.895116316755351e-06, "loss": 2.8858, "step": 968 }, { "epoch": 1.8465204957102002, "grad_norm": 5.429695435696091, "learning_rate": 3.884307715792515e-06, "loss": 2.6721, "step": 969 }, { "epoch": 1.8484270734032413, "grad_norm": 6.213908954902925, "learning_rate": 3.873504598916184e-06, "loss": 2.5311, "step": 970 }, { "epoch": 1.8503336510962822, "grad_norm": 6.137527235742344, "learning_rate": 3.86270701922812e-06, "loss": 2.8261, "step": 971 }, { "epoch": 1.8522402287893232, "grad_norm": 5.876390213484988, "learning_rate": 3.851915029802872e-06, "loss": 2.7642, "step": 972 }, { "epoch": 1.854146806482364, "grad_norm": 5.243805785719302, "learning_rate": 3.841128683687501e-06, "loss": 2.5912, "step": 973 }, { "epoch": 1.856053384175405, "grad_norm": 5.988520397014252, "learning_rate": 3.830348033901342e-06, "loss": 2.7839, "step": 974 }, { "epoch": 1.8579599618684461, "grad_norm": 6.10086522462344, "learning_rate": 3.8195731334357185e-06, "loss": 2.5464, "step": 975 }, { "epoch": 1.8598665395614873, "grad_norm": 5.859547095098912, "learning_rate": 3.808804035253702e-06, "loss": 2.8473, "step": 976 }, { "epoch": 1.8617731172545282, "grad_norm": 5.9175597470132875, "learning_rate": 3.7980407922898357e-06, "loss": 2.9039, "step": 977 }, { "epoch": 1.8636796949475691, "grad_norm": 5.771924100393367, "learning_rate": 3.7872834574498894e-06, "loss": 2.6477, "step": 978 }, { "epoch": 1.86558627264061, "grad_norm": 5.901784484913301, "learning_rate": 3.776532083610587e-06, "loss": 2.7661, "step": 979 }, { "epoch": 1.867492850333651, "grad_norm": 5.988585516704924, "learning_rate": 3.765786723619354e-06, "loss": 2.748, "step": 980 }, { "epoch": 1.8693994280266921, "grad_norm": 5.960444601875205, "learning_rate": 3.7550474302940525e-06, "loss": 2.7509, "step": 981 }, { "epoch": 1.871306005719733, "grad_norm": 5.54717552955932, "learning_rate": 3.744314256422733e-06, "loss": 2.8593, "step": 982 }, { "epoch": 1.8732125834127742, "grad_norm": 5.90465567088123, "learning_rate": 3.733587254763352e-06, "loss": 2.7171, "step": 983 }, { "epoch": 1.8751191611058151, "grad_norm": 5.76061459411467, "learning_rate": 3.722866478043541e-06, "loss": 2.8207, "step": 984 }, { "epoch": 1.877025738798856, "grad_norm": 6.276057490595418, "learning_rate": 3.7121519789603243e-06, "loss": 2.7891, "step": 985 }, { "epoch": 1.878932316491897, "grad_norm": 5.602297092512342, "learning_rate": 3.701443810179875e-06, "loss": 2.7193, "step": 986 }, { "epoch": 1.8808388941849379, "grad_norm": 6.271119446049779, "learning_rate": 3.6907420243372437e-06, "loss": 2.8507, "step": 987 }, { "epoch": 1.882745471877979, "grad_norm": 5.533840393485677, "learning_rate": 3.680046674036114e-06, "loss": 2.619, "step": 988 }, { "epoch": 1.8846520495710202, "grad_norm": 5.662560891479389, "learning_rate": 3.669357811848528e-06, "loss": 2.789, "step": 989 }, { "epoch": 1.886558627264061, "grad_norm": 6.39012561596541, "learning_rate": 3.658675490314643e-06, "loss": 2.8811, "step": 990 }, { "epoch": 1.888465204957102, "grad_norm": 6.055817768384256, "learning_rate": 3.6479997619424605e-06, "loss": 2.8639, "step": 991 }, { "epoch": 1.890371782650143, "grad_norm": 5.875008867588468, "learning_rate": 3.637330679207582e-06, "loss": 2.7633, "step": 992 }, { "epoch": 1.8922783603431839, "grad_norm": 6.009278121312029, "learning_rate": 3.6266682945529307e-06, "loss": 2.6369, "step": 993 }, { "epoch": 1.894184938036225, "grad_norm": 6.229598471965302, "learning_rate": 3.616012660388519e-06, "loss": 2.728, "step": 994 }, { "epoch": 1.896091515729266, "grad_norm": 5.777353012922398, "learning_rate": 3.605363829091169e-06, "loss": 2.7368, "step": 995 }, { "epoch": 1.897998093422307, "grad_norm": 5.50102065107249, "learning_rate": 3.5947218530042682e-06, "loss": 2.733, "step": 996 }, { "epoch": 1.899904671115348, "grad_norm": 6.550339968959376, "learning_rate": 3.584086784437505e-06, "loss": 2.8203, "step": 997 }, { "epoch": 1.901811248808389, "grad_norm": 5.43336051167204, "learning_rate": 3.5734586756666178e-06, "loss": 2.7219, "step": 998 }, { "epoch": 1.9037178265014298, "grad_norm": 5.680489587176046, "learning_rate": 3.562837578933131e-06, "loss": 2.7472, "step": 999 }, { "epoch": 1.9056244041944708, "grad_norm": 5.821848400646796, "learning_rate": 3.552223546444106e-06, "loss": 2.6902, "step": 1000 }, { "epoch": 1.907530981887512, "grad_norm": 5.805326554082773, "learning_rate": 3.5416166303718747e-06, "loss": 2.6343, "step": 1001 }, { "epoch": 1.909437559580553, "grad_norm": 6.098229444663467, "learning_rate": 3.5310168828537976e-06, "loss": 2.7775, "step": 1002 }, { "epoch": 1.911344137273594, "grad_norm": 6.665446556678186, "learning_rate": 3.5204243559919905e-06, "loss": 2.8791, "step": 1003 }, { "epoch": 1.913250714966635, "grad_norm": 6.387577928222472, "learning_rate": 3.509839101853082e-06, "loss": 2.6699, "step": 1004 }, { "epoch": 1.9151572926596758, "grad_norm": 5.825378969538368, "learning_rate": 3.4992611724679487e-06, "loss": 2.7853, "step": 1005 }, { "epoch": 1.9170638703527167, "grad_norm": 6.079647824067391, "learning_rate": 3.4886906198314663e-06, "loss": 2.722, "step": 1006 }, { "epoch": 1.918970448045758, "grad_norm": 6.068129609371163, "learning_rate": 3.478127495902247e-06, "loss": 2.6984, "step": 1007 }, { "epoch": 1.9208770257387988, "grad_norm": 5.6391019752473195, "learning_rate": 3.4675718526023916e-06, "loss": 2.7067, "step": 1008 }, { "epoch": 1.92278360343184, "grad_norm": 5.972608051477931, "learning_rate": 3.457023741817227e-06, "loss": 2.8052, "step": 1009 }, { "epoch": 1.9246901811248809, "grad_norm": 5.335608267744448, "learning_rate": 3.4464832153950606e-06, "loss": 2.6513, "step": 1010 }, { "epoch": 1.9265967588179218, "grad_norm": 6.770115720455367, "learning_rate": 3.4359503251469093e-06, "loss": 2.8166, "step": 1011 }, { "epoch": 1.9285033365109627, "grad_norm": 6.35390203469948, "learning_rate": 3.4254251228462678e-06, "loss": 2.5902, "step": 1012 }, { "epoch": 1.9304099142040037, "grad_norm": 5.694296188841719, "learning_rate": 3.41490766022883e-06, "loss": 2.8456, "step": 1013 }, { "epoch": 1.9323164918970448, "grad_norm": 6.352701259409514, "learning_rate": 3.404397988992254e-06, "loss": 2.6744, "step": 1014 }, { "epoch": 1.934223069590086, "grad_norm": 5.507397346501469, "learning_rate": 3.393896160795895e-06, "loss": 2.8176, "step": 1015 }, { "epoch": 1.9361296472831269, "grad_norm": 6.082647516390172, "learning_rate": 3.3834022272605594e-06, "loss": 2.7155, "step": 1016 }, { "epoch": 1.9380362249761678, "grad_norm": 5.835338805930488, "learning_rate": 3.372916239968246e-06, "loss": 2.7291, "step": 1017 }, { "epoch": 1.9399428026692087, "grad_norm": 5.486918572024454, "learning_rate": 3.3624382504618975e-06, "loss": 2.6706, "step": 1018 }, { "epoch": 1.9418493803622496, "grad_norm": 5.4643949299490835, "learning_rate": 3.351968310245139e-06, "loss": 2.7084, "step": 1019 }, { "epoch": 1.9437559580552908, "grad_norm": 6.105830685052399, "learning_rate": 3.341506470782039e-06, "loss": 2.904, "step": 1020 }, { "epoch": 1.9456625357483317, "grad_norm": 6.296308985828208, "learning_rate": 3.331052783496835e-06, "loss": 2.7753, "step": 1021 }, { "epoch": 1.9475691134413728, "grad_norm": 6.361913875018344, "learning_rate": 3.320607299773705e-06, "loss": 2.8035, "step": 1022 }, { "epoch": 1.9494756911344138, "grad_norm": 5.909264659768186, "learning_rate": 3.3101700709564964e-06, "loss": 2.7226, "step": 1023 }, { "epoch": 1.9513822688274547, "grad_norm": 5.990267800769071, "learning_rate": 3.2997411483484826e-06, "loss": 2.8013, "step": 1024 }, { "epoch": 1.9532888465204956, "grad_norm": 5.771308927652161, "learning_rate": 3.2893205832121066e-06, "loss": 2.8699, "step": 1025 }, { "epoch": 1.9551954242135365, "grad_norm": 5.638459093135103, "learning_rate": 3.2789084267687353e-06, "loss": 2.7555, "step": 1026 }, { "epoch": 1.9571020019065777, "grad_norm": 5.4453041965177915, "learning_rate": 3.2685047301983973e-06, "loss": 2.7047, "step": 1027 }, { "epoch": 1.9590085795996188, "grad_norm": 10.174126847931365, "learning_rate": 3.2581095446395426e-06, "loss": 2.5869, "step": 1028 }, { "epoch": 1.9609151572926597, "grad_norm": 5.511317075293709, "learning_rate": 3.2477229211887828e-06, "loss": 2.6767, "step": 1029 }, { "epoch": 1.9628217349857007, "grad_norm": 5.59150532845545, "learning_rate": 3.2373449109006476e-06, "loss": 2.6888, "step": 1030 }, { "epoch": 1.9647283126787416, "grad_norm": 5.854275685987852, "learning_rate": 3.226975564787322e-06, "loss": 2.6846, "step": 1031 }, { "epoch": 1.9666348903717825, "grad_norm": 6.160347070734732, "learning_rate": 3.216614933818412e-06, "loss": 2.7313, "step": 1032 }, { "epoch": 1.9685414680648237, "grad_norm": 5.68285114316413, "learning_rate": 3.2062630689206765e-06, "loss": 2.589, "step": 1033 }, { "epoch": 1.9704480457578646, "grad_norm": 5.810046976863802, "learning_rate": 3.1959200209777923e-06, "loss": 2.8109, "step": 1034 }, { "epoch": 1.9723546234509057, "grad_norm": 5.2163762650450565, "learning_rate": 3.185585840830093e-06, "loss": 2.6658, "step": 1035 }, { "epoch": 1.9742612011439467, "grad_norm": 5.911815378692056, "learning_rate": 3.1752605792743275e-06, "loss": 2.9143, "step": 1036 }, { "epoch": 1.9761677788369876, "grad_norm": 6.3015964516452545, "learning_rate": 3.1649442870633985e-06, "loss": 2.6848, "step": 1037 }, { "epoch": 1.9780743565300285, "grad_norm": 5.795957808229444, "learning_rate": 3.154637014906132e-06, "loss": 2.7939, "step": 1038 }, { "epoch": 1.9799809342230696, "grad_norm": 5.804447087927288, "learning_rate": 3.1443388134670046e-06, "loss": 2.6518, "step": 1039 }, { "epoch": 1.9818875119161106, "grad_norm": 5.8521891727371935, "learning_rate": 3.134049733365917e-06, "loss": 2.6954, "step": 1040 }, { "epoch": 1.9837940896091517, "grad_norm": 5.6328663949713, "learning_rate": 3.123769825177927e-06, "loss": 2.6374, "step": 1041 }, { "epoch": 1.9857006673021926, "grad_norm": 5.987963057901918, "learning_rate": 3.1134991394330145e-06, "loss": 2.9657, "step": 1042 }, { "epoch": 1.9876072449952336, "grad_norm": 6.212161489804199, "learning_rate": 3.103237726615822e-06, "loss": 2.6511, "step": 1043 }, { "epoch": 1.9895138226882745, "grad_norm": 5.810861354717922, "learning_rate": 3.0929856371654164e-06, "loss": 2.7839, "step": 1044 }, { "epoch": 1.9914204003813154, "grad_norm": 6.200858585476431, "learning_rate": 3.0827429214750337e-06, "loss": 2.7073, "step": 1045 }, { "epoch": 1.9933269780743565, "grad_norm": 5.362192985503895, "learning_rate": 3.0725096298918354e-06, "loss": 2.66, "step": 1046 }, { "epoch": 1.9952335557673975, "grad_norm": 5.941087592782926, "learning_rate": 3.0622858127166565e-06, "loss": 2.7165, "step": 1047 }, { "epoch": 1.9971401334604386, "grad_norm": 5.5397939074719105, "learning_rate": 3.0520715202037686e-06, "loss": 2.6038, "step": 1048 }, { "epoch": 1.9990467111534795, "grad_norm": 5.82383280422543, "learning_rate": 3.041866802560615e-06, "loss": 2.818, "step": 1049 }, { "epoch": 2.0, "grad_norm": 10.10871621071303, "learning_rate": 3.031671709947586e-06, "loss": 2.5609, "step": 1050 }, { "epoch": 2.001906577693041, "grad_norm": 6.200470685583575, "learning_rate": 3.0214862924777516e-06, "loss": 2.2541, "step": 1051 }, { "epoch": 2.003813155386082, "grad_norm": 6.7762263175065165, "learning_rate": 3.0113106002166314e-06, "loss": 2.2831, "step": 1052 }, { "epoch": 2.0057197330791228, "grad_norm": 6.7353607140385146, "learning_rate": 3.001144683181935e-06, "loss": 2.3172, "step": 1053 }, { "epoch": 2.007626310772164, "grad_norm": 6.094617484099409, "learning_rate": 2.99098859134333e-06, "loss": 2.2081, "step": 1054 }, { "epoch": 2.009532888465205, "grad_norm": 6.243062227653791, "learning_rate": 2.9808423746221826e-06, "loss": 2.1321, "step": 1055 }, { "epoch": 2.011439466158246, "grad_norm": 6.825356597138233, "learning_rate": 2.9707060828913226e-06, "loss": 2.277, "step": 1056 }, { "epoch": 2.013346043851287, "grad_norm": 6.335521290296946, "learning_rate": 2.960579765974792e-06, "loss": 2.2886, "step": 1057 }, { "epoch": 2.015252621544328, "grad_norm": 6.49501290206848, "learning_rate": 2.950463473647608e-06, "loss": 2.3481, "step": 1058 }, { "epoch": 2.0171591992373687, "grad_norm": 6.451146272203546, "learning_rate": 2.940357255635502e-06, "loss": 2.3443, "step": 1059 }, { "epoch": 2.01906577693041, "grad_norm": 5.661587843413756, "learning_rate": 2.9302611616146988e-06, "loss": 2.1517, "step": 1060 }, { "epoch": 2.020972354623451, "grad_norm": 5.973940010216026, "learning_rate": 2.9201752412116493e-06, "loss": 2.2029, "step": 1061 }, { "epoch": 2.022878932316492, "grad_norm": 6.085462273532068, "learning_rate": 2.9100995440028053e-06, "loss": 2.3194, "step": 1062 }, { "epoch": 2.024785510009533, "grad_norm": 6.323058467989772, "learning_rate": 2.9000341195143556e-06, "loss": 2.1596, "step": 1063 }, { "epoch": 2.026692087702574, "grad_norm": 6.020784411056396, "learning_rate": 2.889979017222011e-06, "loss": 2.0284, "step": 1064 }, { "epoch": 2.0285986653956147, "grad_norm": 6.805615492134133, "learning_rate": 2.879934286550729e-06, "loss": 2.2277, "step": 1065 }, { "epoch": 2.0305052430886557, "grad_norm": 6.374457326356524, "learning_rate": 2.869899976874495e-06, "loss": 2.1372, "step": 1066 }, { "epoch": 2.032411820781697, "grad_norm": 6.4902487828990525, "learning_rate": 2.8598761375160676e-06, "loss": 2.1176, "step": 1067 }, { "epoch": 2.034318398474738, "grad_norm": 7.675287263647378, "learning_rate": 2.849862817746743e-06, "loss": 2.1715, "step": 1068 }, { "epoch": 2.036224976167779, "grad_norm": 6.530702069795095, "learning_rate": 2.8398600667861032e-06, "loss": 2.0009, "step": 1069 }, { "epoch": 2.03813155386082, "grad_norm": 7.048162839061161, "learning_rate": 2.8298679338017843e-06, "loss": 2.0817, "step": 1070 }, { "epoch": 2.0400381315538607, "grad_norm": 7.92612210571457, "learning_rate": 2.819886467909232e-06, "loss": 2.1786, "step": 1071 }, { "epoch": 2.0419447092469016, "grad_norm": 7.201616466606453, "learning_rate": 2.8099157181714577e-06, "loss": 2.1835, "step": 1072 }, { "epoch": 2.043851286939943, "grad_norm": 7.4020689460588285, "learning_rate": 2.7999557335987914e-06, "loss": 2.1905, "step": 1073 }, { "epoch": 2.045757864632984, "grad_norm": 7.856880726660519, "learning_rate": 2.7900065631486633e-06, "loss": 2.2415, "step": 1074 }, { "epoch": 2.047664442326025, "grad_norm": 7.000869516558582, "learning_rate": 2.780068255725332e-06, "loss": 2.1455, "step": 1075 }, { "epoch": 2.0495710200190658, "grad_norm": 7.915530932344932, "learning_rate": 2.7701408601796688e-06, "loss": 2.2476, "step": 1076 }, { "epoch": 2.0514775977121067, "grad_norm": 6.575585466293001, "learning_rate": 2.760224425308906e-06, "loss": 2.0819, "step": 1077 }, { "epoch": 2.0533841754051476, "grad_norm": 6.544559436726865, "learning_rate": 2.7503189998564017e-06, "loss": 1.9868, "step": 1078 }, { "epoch": 2.0552907530981885, "grad_norm": 6.322472411205264, "learning_rate": 2.7404246325113936e-06, "loss": 1.9309, "step": 1079 }, { "epoch": 2.05719733079123, "grad_norm": 6.742482414594474, "learning_rate": 2.7305413719087683e-06, "loss": 2.096, "step": 1080 }, { "epoch": 2.059103908484271, "grad_norm": 6.356474908694823, "learning_rate": 2.720669266628817e-06, "loss": 2.0717, "step": 1081 }, { "epoch": 2.0610104861773118, "grad_norm": 7.037162510855397, "learning_rate": 2.710808365197e-06, "loss": 1.9894, "step": 1082 }, { "epoch": 2.0629170638703527, "grad_norm": 6.722233912475012, "learning_rate": 2.7009587160836994e-06, "loss": 2.0864, "step": 1083 }, { "epoch": 2.0648236415633936, "grad_norm": 6.468042510837373, "learning_rate": 2.6911203677039977e-06, "loss": 2.1852, "step": 1084 }, { "epoch": 2.0667302192564345, "grad_norm": 5.559595392600497, "learning_rate": 2.6812933684174193e-06, "loss": 1.9658, "step": 1085 }, { "epoch": 2.068636796949476, "grad_norm": 6.778265075175668, "learning_rate": 2.67147776652771e-06, "loss": 2.1409, "step": 1086 }, { "epoch": 2.070543374642517, "grad_norm": 6.494073607660245, "learning_rate": 2.6616736102825906e-06, "loss": 2.1031, "step": 1087 }, { "epoch": 2.0724499523355577, "grad_norm": 7.00697860791799, "learning_rate": 2.6518809478735243e-06, "loss": 2.1576, "step": 1088 }, { "epoch": 2.0743565300285987, "grad_norm": 7.2887361009448695, "learning_rate": 2.64209982743547e-06, "loss": 2.3742, "step": 1089 }, { "epoch": 2.0762631077216396, "grad_norm": 5.844110437773209, "learning_rate": 2.6323302970466604e-06, "loss": 2.0, "step": 1090 }, { "epoch": 2.0781696854146805, "grad_norm": 6.457533432470302, "learning_rate": 2.6225724047283564e-06, "loss": 1.9467, "step": 1091 }, { "epoch": 2.0800762631077214, "grad_norm": 5.965468281071334, "learning_rate": 2.612826198444612e-06, "loss": 2.0146, "step": 1092 }, { "epoch": 2.081982840800763, "grad_norm": 7.238944162528596, "learning_rate": 2.60309172610204e-06, "loss": 2.0133, "step": 1093 }, { "epoch": 2.0838894184938037, "grad_norm": 7.22550938573297, "learning_rate": 2.5933690355495787e-06, "loss": 2.1563, "step": 1094 }, { "epoch": 2.0857959961868446, "grad_norm": 6.862271287550726, "learning_rate": 2.5836581745782474e-06, "loss": 2.1809, "step": 1095 }, { "epoch": 2.0877025738798856, "grad_norm": 6.386471187710955, "learning_rate": 2.5739591909209233e-06, "loss": 2.1726, "step": 1096 }, { "epoch": 2.0896091515729265, "grad_norm": 7.121667090535104, "learning_rate": 2.5642721322521014e-06, "loss": 2.1415, "step": 1097 }, { "epoch": 2.0915157292659674, "grad_norm": 6.88556953686062, "learning_rate": 2.5545970461876624e-06, "loss": 2.1835, "step": 1098 }, { "epoch": 2.0934223069590088, "grad_norm": 6.551682895974867, "learning_rate": 2.5449339802846268e-06, "loss": 2.0357, "step": 1099 }, { "epoch": 2.0953288846520497, "grad_norm": 6.624709053796721, "learning_rate": 2.535282982040947e-06, "loss": 2.1501, "step": 1100 }, { "epoch": 2.0972354623450906, "grad_norm": 6.941651679249668, "learning_rate": 2.5256440988952428e-06, "loss": 2.3528, "step": 1101 }, { "epoch": 2.0991420400381315, "grad_norm": 6.191447693938032, "learning_rate": 2.516017378226592e-06, "loss": 1.9159, "step": 1102 }, { "epoch": 2.1010486177311725, "grad_norm": 6.382035113406856, "learning_rate": 2.5064028673542874e-06, "loss": 2.1402, "step": 1103 }, { "epoch": 2.1029551954242134, "grad_norm": 6.9601748975894875, "learning_rate": 2.4968006135376046e-06, "loss": 2.1575, "step": 1104 }, { "epoch": 2.1048617731172543, "grad_norm": 6.637419858087578, "learning_rate": 2.4872106639755684e-06, "loss": 1.9971, "step": 1105 }, { "epoch": 2.1067683508102957, "grad_norm": 7.494909727243204, "learning_rate": 2.4776330658067275e-06, "loss": 2.2777, "step": 1106 }, { "epoch": 2.1086749285033366, "grad_norm": 7.118062193358437, "learning_rate": 2.468067866108917e-06, "loss": 2.3136, "step": 1107 }, { "epoch": 2.1105815061963775, "grad_norm": 7.044145103925187, "learning_rate": 2.4585151118990286e-06, "loss": 2.1614, "step": 1108 }, { "epoch": 2.1124880838894184, "grad_norm": 6.647677981137968, "learning_rate": 2.448974850132774e-06, "loss": 2.1347, "step": 1109 }, { "epoch": 2.1143946615824594, "grad_norm": 6.682490965292744, "learning_rate": 2.439447127704471e-06, "loss": 2.244, "step": 1110 }, { "epoch": 2.1163012392755003, "grad_norm": 7.0263710687304055, "learning_rate": 2.429931991446788e-06, "loss": 2.1876, "step": 1111 }, { "epoch": 2.1182078169685417, "grad_norm": 6.980167351038054, "learning_rate": 2.420429488130538e-06, "loss": 2.2918, "step": 1112 }, { "epoch": 2.1201143946615826, "grad_norm": 6.794071803987173, "learning_rate": 2.410939664464432e-06, "loss": 2.0854, "step": 1113 }, { "epoch": 2.1220209723546235, "grad_norm": 6.611452673925243, "learning_rate": 2.4014625670948584e-06, "loss": 2.0538, "step": 1114 }, { "epoch": 2.1239275500476644, "grad_norm": 6.997139203362343, "learning_rate": 2.391998242605646e-06, "loss": 2.262, "step": 1115 }, { "epoch": 2.1258341277407053, "grad_norm": 6.668765801177306, "learning_rate": 2.3825467375178447e-06, "loss": 2.1444, "step": 1116 }, { "epoch": 2.1277407054337463, "grad_norm": 6.74875695055662, "learning_rate": 2.373108098289489e-06, "loss": 2.0999, "step": 1117 }, { "epoch": 2.129647283126787, "grad_norm": 6.574865787784748, "learning_rate": 2.3636823713153754e-06, "loss": 2.093, "step": 1118 }, { "epoch": 2.1315538608198286, "grad_norm": 6.584890353804752, "learning_rate": 2.354269602926824e-06, "loss": 2.0723, "step": 1119 }, { "epoch": 2.1334604385128695, "grad_norm": 6.589555601991108, "learning_rate": 2.3448698393914704e-06, "loss": 2.14, "step": 1120 }, { "epoch": 2.1353670162059104, "grad_norm": 7.005166831650654, "learning_rate": 2.3354831269130133e-06, "loss": 2.1686, "step": 1121 }, { "epoch": 2.1372735938989513, "grad_norm": 6.840729894027851, "learning_rate": 2.3261095116310073e-06, "loss": 2.2843, "step": 1122 }, { "epoch": 2.1391801715919923, "grad_norm": 6.530313313707763, "learning_rate": 2.3167490396206277e-06, "loss": 2.1182, "step": 1123 }, { "epoch": 2.141086749285033, "grad_norm": 7.034791374927135, "learning_rate": 2.3074017568924463e-06, "loss": 2.0712, "step": 1124 }, { "epoch": 2.1429933269780745, "grad_norm": 6.336057866686705, "learning_rate": 2.298067709392199e-06, "loss": 1.957, "step": 1125 }, { "epoch": 2.1448999046711155, "grad_norm": 6.407330137487336, "learning_rate": 2.2887469430005704e-06, "loss": 2.1541, "step": 1126 }, { "epoch": 2.1468064823641564, "grad_norm": 6.441524756488472, "learning_rate": 2.2794395035329616e-06, "loss": 2.1456, "step": 1127 }, { "epoch": 2.1487130600571973, "grad_norm": 6.210806304184773, "learning_rate": 2.2701454367392672e-06, "loss": 2.175, "step": 1128 }, { "epoch": 2.1506196377502382, "grad_norm": 6.9066038081234655, "learning_rate": 2.260864788303644e-06, "loss": 2.2381, "step": 1129 }, { "epoch": 2.152526215443279, "grad_norm": 6.2019825723273385, "learning_rate": 2.2515976038443034e-06, "loss": 1.9499, "step": 1130 }, { "epoch": 2.15443279313632, "grad_norm": 6.960172819263362, "learning_rate": 2.2423439289132635e-06, "loss": 2.2123, "step": 1131 }, { "epoch": 2.1563393708293614, "grad_norm": 7.068749106391073, "learning_rate": 2.2331038089961447e-06, "loss": 2.2444, "step": 1132 }, { "epoch": 2.1582459485224024, "grad_norm": 6.21377409264196, "learning_rate": 2.2238772895119375e-06, "loss": 2.1638, "step": 1133 }, { "epoch": 2.1601525262154433, "grad_norm": 6.814631256840516, "learning_rate": 2.2146644158127827e-06, "loss": 2.1809, "step": 1134 }, { "epoch": 2.162059103908484, "grad_norm": 6.227724182769759, "learning_rate": 2.205465233183739e-06, "loss": 2.1254, "step": 1135 }, { "epoch": 2.163965681601525, "grad_norm": 6.711157772156754, "learning_rate": 2.196279786842576e-06, "loss": 2.2184, "step": 1136 }, { "epoch": 2.165872259294566, "grad_norm": 8.127422174715782, "learning_rate": 2.1871081219395407e-06, "loss": 2.2868, "step": 1137 }, { "epoch": 2.1677788369876074, "grad_norm": 7.018759122715207, "learning_rate": 2.1779502835571405e-06, "loss": 2.2478, "step": 1138 }, { "epoch": 2.1696854146806483, "grad_norm": 6.495984929150095, "learning_rate": 2.1688063167099122e-06, "loss": 2.1806, "step": 1139 }, { "epoch": 2.1715919923736893, "grad_norm": 6.614666485825436, "learning_rate": 2.159676266344222e-06, "loss": 2.1948, "step": 1140 }, { "epoch": 2.17349857006673, "grad_norm": 7.132094092708455, "learning_rate": 2.150560177338017e-06, "loss": 2.0285, "step": 1141 }, { "epoch": 2.175405147759771, "grad_norm": 6.600187483494467, "learning_rate": 2.1414580945006265e-06, "loss": 2.1051, "step": 1142 }, { "epoch": 2.177311725452812, "grad_norm": 6.954814420989218, "learning_rate": 2.1323700625725313e-06, "loss": 2.1372, "step": 1143 }, { "epoch": 2.179218303145853, "grad_norm": 6.267465935439692, "learning_rate": 2.1232961262251488e-06, "loss": 2.0443, "step": 1144 }, { "epoch": 2.1811248808388943, "grad_norm": 7.066789379817989, "learning_rate": 2.1142363300606055e-06, "loss": 2.1672, "step": 1145 }, { "epoch": 2.1830314585319353, "grad_norm": 6.481334655032877, "learning_rate": 2.105190718611527e-06, "loss": 1.9934, "step": 1146 }, { "epoch": 2.184938036224976, "grad_norm": 6.785838646570115, "learning_rate": 2.0961593363408154e-06, "loss": 2.0629, "step": 1147 }, { "epoch": 2.186844613918017, "grad_norm": 6.743699306237948, "learning_rate": 2.0871422276414304e-06, "loss": 2.0003, "step": 1148 }, { "epoch": 2.188751191611058, "grad_norm": 6.672850807467303, "learning_rate": 2.078139436836166e-06, "loss": 2.1701, "step": 1149 }, { "epoch": 2.190657769304099, "grad_norm": 7.021403630993581, "learning_rate": 2.069151008177448e-06, "loss": 2.1683, "step": 1150 }, { "epoch": 2.1925643469971403, "grad_norm": 6.008987578154338, "learning_rate": 2.060176985847096e-06, "loss": 1.9669, "step": 1151 }, { "epoch": 2.1944709246901812, "grad_norm": 6.622657303360847, "learning_rate": 2.051217413956123e-06, "loss": 2.0314, "step": 1152 }, { "epoch": 2.196377502383222, "grad_norm": 6.170240286971733, "learning_rate": 2.04227233654451e-06, "loss": 2.0353, "step": 1153 }, { "epoch": 2.198284080076263, "grad_norm": 6.875576550508573, "learning_rate": 2.033341797580994e-06, "loss": 2.0646, "step": 1154 }, { "epoch": 2.200190657769304, "grad_norm": 6.72200816505001, "learning_rate": 2.0244258409628433e-06, "loss": 2.0665, "step": 1155 }, { "epoch": 2.202097235462345, "grad_norm": 6.934170626206378, "learning_rate": 2.0155245105156544e-06, "loss": 2.185, "step": 1156 }, { "epoch": 2.204003813155386, "grad_norm": 7.489035298008023, "learning_rate": 2.0066378499931283e-06, "loss": 2.2476, "step": 1157 }, { "epoch": 2.205910390848427, "grad_norm": 6.750034715886643, "learning_rate": 1.9977659030768583e-06, "loss": 2.1175, "step": 1158 }, { "epoch": 2.207816968541468, "grad_norm": 6.9123995379920995, "learning_rate": 1.988908713376107e-06, "loss": 2.1254, "step": 1159 }, { "epoch": 2.209723546234509, "grad_norm": 6.6121509960258384, "learning_rate": 1.980066324427613e-06, "loss": 2.0149, "step": 1160 }, { "epoch": 2.21163012392755, "grad_norm": 6.162726750750801, "learning_rate": 1.9712387796953497e-06, "loss": 2.0531, "step": 1161 }, { "epoch": 2.213536701620591, "grad_norm": 6.598279268117259, "learning_rate": 1.9624261225703316e-06, "loss": 2.1367, "step": 1162 }, { "epoch": 2.215443279313632, "grad_norm": 6.554492314390867, "learning_rate": 1.953628396370395e-06, "loss": 2.071, "step": 1163 }, { "epoch": 2.217349857006673, "grad_norm": 7.075422472411845, "learning_rate": 1.9448456443399833e-06, "loss": 2.003, "step": 1164 }, { "epoch": 2.219256434699714, "grad_norm": 6.725597847564434, "learning_rate": 1.936077909649933e-06, "loss": 2.0355, "step": 1165 }, { "epoch": 2.221163012392755, "grad_norm": 6.591525365997407, "learning_rate": 1.9273252353972673e-06, "loss": 2.0425, "step": 1166 }, { "epoch": 2.223069590085796, "grad_norm": 6.668194930825103, "learning_rate": 1.9185876646049804e-06, "loss": 2.2086, "step": 1167 }, { "epoch": 2.224976167778837, "grad_norm": 6.200290865923421, "learning_rate": 1.909865240221828e-06, "loss": 2.1635, "step": 1168 }, { "epoch": 2.226882745471878, "grad_norm": 6.253309066272892, "learning_rate": 1.901158005122114e-06, "loss": 2.0988, "step": 1169 }, { "epoch": 2.228789323164919, "grad_norm": 6.61028272768087, "learning_rate": 1.8924660021054808e-06, "loss": 2.0547, "step": 1170 }, { "epoch": 2.23069590085796, "grad_norm": 7.154737338312215, "learning_rate": 1.8837892738966962e-06, "loss": 2.048, "step": 1171 }, { "epoch": 2.232602478551001, "grad_norm": 6.530743882558572, "learning_rate": 1.8751278631454495e-06, "loss": 2.2539, "step": 1172 }, { "epoch": 2.234509056244042, "grad_norm": 6.433190622630085, "learning_rate": 1.8664818124261375e-06, "loss": 2.0128, "step": 1173 }, { "epoch": 2.236415633937083, "grad_norm": 7.2711025562003595, "learning_rate": 1.8578511642376585e-06, "loss": 2.3928, "step": 1174 }, { "epoch": 2.238322211630124, "grad_norm": 6.341506657600693, "learning_rate": 1.8492359610031917e-06, "loss": 2.1391, "step": 1175 }, { "epoch": 2.2402287893231647, "grad_norm": 6.35684673228532, "learning_rate": 1.8406362450700133e-06, "loss": 2.0091, "step": 1176 }, { "epoch": 2.242135367016206, "grad_norm": 6.1554309465951755, "learning_rate": 1.8320520587092583e-06, "loss": 2.1681, "step": 1177 }, { "epoch": 2.244041944709247, "grad_norm": 6.852877918279779, "learning_rate": 1.8234834441157357e-06, "loss": 2.1598, "step": 1178 }, { "epoch": 2.245948522402288, "grad_norm": 6.78252846624434, "learning_rate": 1.8149304434077115e-06, "loss": 2.1506, "step": 1179 }, { "epoch": 2.247855100095329, "grad_norm": 7.430061037032722, "learning_rate": 1.8063930986267025e-06, "loss": 2.1413, "step": 1180 }, { "epoch": 2.2497616777883698, "grad_norm": 6.880102822279806, "learning_rate": 1.7978714517372681e-06, "loss": 2.1741, "step": 1181 }, { "epoch": 2.2516682554814107, "grad_norm": 6.723286062558048, "learning_rate": 1.7893655446268087e-06, "loss": 2.1717, "step": 1182 }, { "epoch": 2.2535748331744516, "grad_norm": 6.179480351605467, "learning_rate": 1.7808754191053574e-06, "loss": 2.1518, "step": 1183 }, { "epoch": 2.255481410867493, "grad_norm": 6.110416811689817, "learning_rate": 1.772401116905375e-06, "loss": 2.2213, "step": 1184 }, { "epoch": 2.257387988560534, "grad_norm": 7.72490668235385, "learning_rate": 1.7639426796815368e-06, "loss": 2.1659, "step": 1185 }, { "epoch": 2.259294566253575, "grad_norm": 6.645760541378897, "learning_rate": 1.755500149010549e-06, "loss": 2.015, "step": 1186 }, { "epoch": 2.2612011439466158, "grad_norm": 6.771503916017915, "learning_rate": 1.7470735663909167e-06, "loss": 2.2732, "step": 1187 }, { "epoch": 2.2631077216396567, "grad_norm": 5.655718552814828, "learning_rate": 1.7386629732427611e-06, "loss": 2.0391, "step": 1188 }, { "epoch": 2.265014299332698, "grad_norm": 6.16277882539588, "learning_rate": 1.7302684109076078e-06, "loss": 2.0468, "step": 1189 }, { "epoch": 2.266920877025739, "grad_norm": 6.651863546065097, "learning_rate": 1.7218899206481854e-06, "loss": 2.0167, "step": 1190 }, { "epoch": 2.26882745471878, "grad_norm": 6.414251184358879, "learning_rate": 1.7135275436482164e-06, "loss": 2.2559, "step": 1191 }, { "epoch": 2.270734032411821, "grad_norm": 6.426063730290265, "learning_rate": 1.7051813210122243e-06, "loss": 2.2638, "step": 1192 }, { "epoch": 2.2726406101048617, "grad_norm": 6.354746937493074, "learning_rate": 1.6968512937653282e-06, "loss": 2.2765, "step": 1193 }, { "epoch": 2.2745471877979027, "grad_norm": 6.844979884172814, "learning_rate": 1.6885375028530387e-06, "loss": 2.0631, "step": 1194 }, { "epoch": 2.2764537654909436, "grad_norm": 6.768840899877247, "learning_rate": 1.6802399891410538e-06, "loss": 2.2001, "step": 1195 }, { "epoch": 2.2783603431839845, "grad_norm": 6.055758270692035, "learning_rate": 1.6719587934150722e-06, "loss": 1.9958, "step": 1196 }, { "epoch": 2.280266920877026, "grad_norm": 6.846163493333718, "learning_rate": 1.6636939563805738e-06, "loss": 2.3861, "step": 1197 }, { "epoch": 2.282173498570067, "grad_norm": 6.973180261413589, "learning_rate": 1.6554455186626322e-06, "loss": 2.2081, "step": 1198 }, { "epoch": 2.2840800762631077, "grad_norm": 6.719777559522845, "learning_rate": 1.6472135208057128e-06, "loss": 2.133, "step": 1199 }, { "epoch": 2.2859866539561486, "grad_norm": 6.618255547090349, "learning_rate": 1.6389980032734726e-06, "loss": 2.2055, "step": 1200 }, { "epoch": 2.2878932316491896, "grad_norm": 6.144753757503895, "learning_rate": 1.6307990064485568e-06, "loss": 2.0054, "step": 1201 }, { "epoch": 2.289799809342231, "grad_norm": 7.002379705271234, "learning_rate": 1.6226165706324076e-06, "loss": 2.1705, "step": 1202 }, { "epoch": 2.291706387035272, "grad_norm": 6.210196699246412, "learning_rate": 1.6144507360450645e-06, "loss": 2.0606, "step": 1203 }, { "epoch": 2.2936129647283128, "grad_norm": 6.805962124672768, "learning_rate": 1.606301542824964e-06, "loss": 2.2215, "step": 1204 }, { "epoch": 2.2955195424213537, "grad_norm": 6.512796159116104, "learning_rate": 1.5981690310287383e-06, "loss": 1.9861, "step": 1205 }, { "epoch": 2.2974261201143946, "grad_norm": 6.478147535139437, "learning_rate": 1.5900532406310343e-06, "loss": 1.9981, "step": 1206 }, { "epoch": 2.2993326978074355, "grad_norm": 6.64117574866836, "learning_rate": 1.581954211524296e-06, "loss": 2.0532, "step": 1207 }, { "epoch": 2.3012392755004765, "grad_norm": 7.169265484099766, "learning_rate": 1.5738719835185838e-06, "loss": 2.175, "step": 1208 }, { "epoch": 2.3031458531935174, "grad_norm": 6.7716881785724725, "learning_rate": 1.5658065963413732e-06, "loss": 2.0836, "step": 1209 }, { "epoch": 2.3050524308865588, "grad_norm": 6.380682903156265, "learning_rate": 1.5577580896373612e-06, "loss": 2.2525, "step": 1210 }, { "epoch": 2.3069590085795997, "grad_norm": 6.909911488880871, "learning_rate": 1.5497265029682656e-06, "loss": 2.209, "step": 1211 }, { "epoch": 2.3088655862726406, "grad_norm": 7.462301589638039, "learning_rate": 1.5417118758126408e-06, "loss": 2.1842, "step": 1212 }, { "epoch": 2.3107721639656815, "grad_norm": 6.5040504945236775, "learning_rate": 1.5337142475656759e-06, "loss": 2.0198, "step": 1213 }, { "epoch": 2.3126787416587224, "grad_norm": 6.732534641586817, "learning_rate": 1.5257336575390042e-06, "loss": 2.1765, "step": 1214 }, { "epoch": 2.314585319351764, "grad_norm": 6.916939842109315, "learning_rate": 1.5177701449605042e-06, "loss": 2.1962, "step": 1215 }, { "epoch": 2.3164918970448047, "grad_norm": 7.111130773706916, "learning_rate": 1.509823748974123e-06, "loss": 2.2941, "step": 1216 }, { "epoch": 2.3183984747378457, "grad_norm": 6.349404733925756, "learning_rate": 1.5018945086396598e-06, "loss": 2.0538, "step": 1217 }, { "epoch": 2.3203050524308866, "grad_norm": 6.613311347695292, "learning_rate": 1.4939824629325938e-06, "loss": 2.3381, "step": 1218 }, { "epoch": 2.3222116301239275, "grad_norm": 6.674747949769982, "learning_rate": 1.4860876507438854e-06, "loss": 2.0582, "step": 1219 }, { "epoch": 2.3241182078169684, "grad_norm": 6.698973788095343, "learning_rate": 1.4782101108797842e-06, "loss": 2.1537, "step": 1220 }, { "epoch": 2.3260247855100094, "grad_norm": 6.625935529153224, "learning_rate": 1.470349882061637e-06, "loss": 2.1211, "step": 1221 }, { "epoch": 2.3279313632030503, "grad_norm": 6.750573028317114, "learning_rate": 1.4625070029257028e-06, "loss": 2.0121, "step": 1222 }, { "epoch": 2.3298379408960916, "grad_norm": 6.194659847124417, "learning_rate": 1.4546815120229579e-06, "loss": 2.0827, "step": 1223 }, { "epoch": 2.3317445185891326, "grad_norm": 6.715784442533597, "learning_rate": 1.4468734478189117e-06, "loss": 2.1473, "step": 1224 }, { "epoch": 2.3336510962821735, "grad_norm": 6.18129540731257, "learning_rate": 1.439082848693406e-06, "loss": 2.1149, "step": 1225 }, { "epoch": 2.3355576739752144, "grad_norm": 7.55991135730825, "learning_rate": 1.4313097529404469e-06, "loss": 1.9871, "step": 1226 }, { "epoch": 2.3374642516682553, "grad_norm": 6.52936180784428, "learning_rate": 1.4235541987679924e-06, "loss": 2.1069, "step": 1227 }, { "epoch": 2.3393708293612967, "grad_norm": 7.360564968964438, "learning_rate": 1.415816224297783e-06, "loss": 2.1546, "step": 1228 }, { "epoch": 2.3412774070543376, "grad_norm": 6.502157832449864, "learning_rate": 1.4080958675651464e-06, "loss": 2.0309, "step": 1229 }, { "epoch": 2.3431839847473785, "grad_norm": 7.03260475996859, "learning_rate": 1.4003931665188131e-06, "loss": 2.1419, "step": 1230 }, { "epoch": 2.3450905624404195, "grad_norm": 6.204071962099826, "learning_rate": 1.3927081590207236e-06, "loss": 2.1368, "step": 1231 }, { "epoch": 2.3469971401334604, "grad_norm": 6.4198432909748, "learning_rate": 1.3850408828458534e-06, "loss": 2.0779, "step": 1232 }, { "epoch": 2.3489037178265013, "grad_norm": 5.958256736907576, "learning_rate": 1.3773913756820179e-06, "loss": 2.13, "step": 1233 }, { "epoch": 2.3508102955195422, "grad_norm": 7.4582862348793935, "learning_rate": 1.3697596751296928e-06, "loss": 2.0673, "step": 1234 }, { "epoch": 2.3527168732125836, "grad_norm": 6.9534209094651205, "learning_rate": 1.3621458187018217e-06, "loss": 2.1603, "step": 1235 }, { "epoch": 2.3546234509056245, "grad_norm": 6.591725870918129, "learning_rate": 1.3545498438236466e-06, "loss": 2.0909, "step": 1236 }, { "epoch": 2.3565300285986654, "grad_norm": 6.929976685682677, "learning_rate": 1.3469717878325034e-06, "loss": 2.2084, "step": 1237 }, { "epoch": 2.3584366062917064, "grad_norm": 7.007999673514834, "learning_rate": 1.339411687977657e-06, "loss": 2.0712, "step": 1238 }, { "epoch": 2.3603431839847473, "grad_norm": 7.310191075835195, "learning_rate": 1.3318695814201072e-06, "loss": 2.171, "step": 1239 }, { "epoch": 2.362249761677788, "grad_norm": 6.745311860460879, "learning_rate": 1.3243455052324116e-06, "loss": 1.9971, "step": 1240 }, { "epoch": 2.3641563393708296, "grad_norm": 6.400192410676587, "learning_rate": 1.316839496398497e-06, "loss": 2.1304, "step": 1241 }, { "epoch": 2.3660629170638705, "grad_norm": 6.881800082419432, "learning_rate": 1.3093515918134857e-06, "loss": 2.1718, "step": 1242 }, { "epoch": 2.3679694947569114, "grad_norm": 7.665791716553441, "learning_rate": 1.3018818282835088e-06, "loss": 2.1422, "step": 1243 }, { "epoch": 2.3698760724499524, "grad_norm": 7.240478594013899, "learning_rate": 1.2944302425255268e-06, "loss": 2.1785, "step": 1244 }, { "epoch": 2.3717826501429933, "grad_norm": 6.491461393186075, "learning_rate": 1.2869968711671483e-06, "loss": 2.2381, "step": 1245 }, { "epoch": 2.373689227836034, "grad_norm": 6.185730029393945, "learning_rate": 1.2795817507464532e-06, "loss": 2.17, "step": 1246 }, { "epoch": 2.375595805529075, "grad_norm": 6.625268689793941, "learning_rate": 1.272184917711805e-06, "loss": 2.0845, "step": 1247 }, { "epoch": 2.3775023832221165, "grad_norm": 6.563203040223173, "learning_rate": 1.2648064084216828e-06, "loss": 2.0433, "step": 1248 }, { "epoch": 2.3794089609151574, "grad_norm": 6.217106425657633, "learning_rate": 1.257446259144494e-06, "loss": 2.1978, "step": 1249 }, { "epoch": 2.3813155386081983, "grad_norm": 7.388457676296042, "learning_rate": 1.2501045060584026e-06, "loss": 2.2787, "step": 1250 }, { "epoch": 2.3832221163012393, "grad_norm": 7.348827848160648, "learning_rate": 1.2427811852511396e-06, "loss": 2.1944, "step": 1251 }, { "epoch": 2.38512869399428, "grad_norm": 7.327226668497229, "learning_rate": 1.235476332719845e-06, "loss": 2.1701, "step": 1252 }, { "epoch": 2.387035271687321, "grad_norm": 6.84548446071144, "learning_rate": 1.228189984370869e-06, "loss": 2.2504, "step": 1253 }, { "epoch": 2.3889418493803625, "grad_norm": 7.089255276244038, "learning_rate": 1.2209221760196133e-06, "loss": 2.1718, "step": 1254 }, { "epoch": 2.3908484270734034, "grad_norm": 6.9105769589017925, "learning_rate": 1.2136729433903455e-06, "loss": 2.065, "step": 1255 }, { "epoch": 2.3927550047664443, "grad_norm": 6.376770455071304, "learning_rate": 1.2064423221160266e-06, "loss": 2.0957, "step": 1256 }, { "epoch": 2.3946615824594852, "grad_norm": 6.760164937652605, "learning_rate": 1.1992303477381312e-06, "loss": 2.0667, "step": 1257 }, { "epoch": 2.396568160152526, "grad_norm": 6.228034592744551, "learning_rate": 1.1920370557064825e-06, "loss": 2.0927, "step": 1258 }, { "epoch": 2.398474737845567, "grad_norm": 6.4922466531245, "learning_rate": 1.1848624813790693e-06, "loss": 2.0693, "step": 1259 }, { "epoch": 2.400381315538608, "grad_norm": 6.603984113813945, "learning_rate": 1.1777066600218767e-06, "loss": 2.2633, "step": 1260 }, { "epoch": 2.4022878932316494, "grad_norm": 6.884484796379801, "learning_rate": 1.1705696268087064e-06, "loss": 2.2035, "step": 1261 }, { "epoch": 2.4041944709246903, "grad_norm": 6.908333662404297, "learning_rate": 1.163451416821017e-06, "loss": 2.1421, "step": 1262 }, { "epoch": 2.406101048617731, "grad_norm": 6.744882253523864, "learning_rate": 1.1563520650477355e-06, "loss": 2.1975, "step": 1263 }, { "epoch": 2.408007626310772, "grad_norm": 6.708144938653109, "learning_rate": 1.1492716063850973e-06, "loss": 2.1101, "step": 1264 }, { "epoch": 2.409914204003813, "grad_norm": 6.831134201475528, "learning_rate": 1.142210075636469e-06, "loss": 2.1903, "step": 1265 }, { "epoch": 2.411820781696854, "grad_norm": 6.847183279299655, "learning_rate": 1.1351675075121804e-06, "loss": 2.0267, "step": 1266 }, { "epoch": 2.4137273593898954, "grad_norm": 7.081645875087188, "learning_rate": 1.1281439366293474e-06, "loss": 2.0604, "step": 1267 }, { "epoch": 2.4156339370829363, "grad_norm": 6.748589542469892, "learning_rate": 1.1211393975117123e-06, "loss": 2.1535, "step": 1268 }, { "epoch": 2.417540514775977, "grad_norm": 6.596269228242316, "learning_rate": 1.114153924589465e-06, "loss": 2.2093, "step": 1269 }, { "epoch": 2.419447092469018, "grad_norm": 6.740617169191115, "learning_rate": 1.107187552199081e-06, "loss": 2.2439, "step": 1270 }, { "epoch": 2.421353670162059, "grad_norm": 6.714045620265672, "learning_rate": 1.1002403145831409e-06, "loss": 2.0965, "step": 1271 }, { "epoch": 2.4232602478551, "grad_norm": 6.74820405705551, "learning_rate": 1.0933122458901806e-06, "loss": 2.1071, "step": 1272 }, { "epoch": 2.425166825548141, "grad_norm": 7.067640603394402, "learning_rate": 1.086403380174505e-06, "loss": 2.2403, "step": 1273 }, { "epoch": 2.4270734032411823, "grad_norm": 6.8757825870165465, "learning_rate": 1.0795137513960307e-06, "loss": 1.9315, "step": 1274 }, { "epoch": 2.428979980934223, "grad_norm": 7.0374386342001145, "learning_rate": 1.0726433934201174e-06, "loss": 2.1011, "step": 1275 }, { "epoch": 2.430886558627264, "grad_norm": 6.235067419285874, "learning_rate": 1.0657923400174031e-06, "loss": 2.0038, "step": 1276 }, { "epoch": 2.432793136320305, "grad_norm": 6.722151152762094, "learning_rate": 1.0589606248636291e-06, "loss": 2.1387, "step": 1277 }, { "epoch": 2.434699714013346, "grad_norm": 6.87076146789668, "learning_rate": 1.052148281539488e-06, "loss": 2.0585, "step": 1278 }, { "epoch": 2.436606291706387, "grad_norm": 7.0511854528330895, "learning_rate": 1.0453553435304492e-06, "loss": 2.3234, "step": 1279 }, { "epoch": 2.4385128693994282, "grad_norm": 6.758867835960008, "learning_rate": 1.0385818442265981e-06, "loss": 2.1118, "step": 1280 }, { "epoch": 2.440419447092469, "grad_norm": 6.8251892479738085, "learning_rate": 1.0318278169224666e-06, "loss": 2.0752, "step": 1281 }, { "epoch": 2.44232602478551, "grad_norm": 9.25846278364727, "learning_rate": 1.0250932948168817e-06, "loss": 2.2669, "step": 1282 }, { "epoch": 2.444232602478551, "grad_norm": 7.266148771309099, "learning_rate": 1.018378311012786e-06, "loss": 2.1164, "step": 1283 }, { "epoch": 2.446139180171592, "grad_norm": 6.633630858420934, "learning_rate": 1.011682898517089e-06, "loss": 2.0975, "step": 1284 }, { "epoch": 2.448045757864633, "grad_norm": 6.817740799508313, "learning_rate": 1.0050070902404962e-06, "loss": 1.9866, "step": 1285 }, { "epoch": 2.4499523355576738, "grad_norm": 6.9868061254194025, "learning_rate": 9.983509189973534e-07, "loss": 2.1014, "step": 1286 }, { "epoch": 2.451858913250715, "grad_norm": 6.927245615779341, "learning_rate": 9.917144175054778e-07, "loss": 2.1007, "step": 1287 }, { "epoch": 2.453765490943756, "grad_norm": 6.486888026568518, "learning_rate": 9.850976183860039e-07, "loss": 2.0511, "step": 1288 }, { "epoch": 2.455672068636797, "grad_norm": 6.913789453489807, "learning_rate": 9.785005541632225e-07, "loss": 2.0646, "step": 1289 }, { "epoch": 2.457578646329838, "grad_norm": 6.746849666372858, "learning_rate": 9.719232572644189e-07, "loss": 2.2029, "step": 1290 }, { "epoch": 2.459485224022879, "grad_norm": 7.043934910488933, "learning_rate": 9.653657600197092e-07, "loss": 2.1599, "step": 1291 }, { "epoch": 2.4613918017159198, "grad_norm": 6.983571101603111, "learning_rate": 9.588280946618939e-07, "loss": 2.1358, "step": 1292 }, { "epoch": 2.463298379408961, "grad_norm": 6.414505709882128, "learning_rate": 9.523102933262845e-07, "loss": 2.1675, "step": 1293 }, { "epoch": 2.465204957102002, "grad_norm": 6.340347639555769, "learning_rate": 9.458123880505565e-07, "loss": 2.0915, "step": 1294 }, { "epoch": 2.467111534795043, "grad_norm": 7.017840550867588, "learning_rate": 9.393344107745872e-07, "loss": 2.3005, "step": 1295 }, { "epoch": 2.469018112488084, "grad_norm": 7.097559527962726, "learning_rate": 9.328763933403012e-07, "loss": 2.1784, "step": 1296 }, { "epoch": 2.470924690181125, "grad_norm": 6.559213525397916, "learning_rate": 9.264383674915073e-07, "loss": 2.0665, "step": 1297 }, { "epoch": 2.4728312678741657, "grad_norm": 6.432650715219975, "learning_rate": 9.200203648737527e-07, "loss": 1.9866, "step": 1298 }, { "epoch": 2.4747378455672067, "grad_norm": 7.1433722261916, "learning_rate": 9.136224170341601e-07, "loss": 2.0525, "step": 1299 }, { "epoch": 2.476644423260248, "grad_norm": 6.513411490829954, "learning_rate": 9.072445554212761e-07, "loss": 2.1427, "step": 1300 }, { "epoch": 2.478551000953289, "grad_norm": 7.344603144720768, "learning_rate": 9.0088681138491e-07, "loss": 2.373, "step": 1301 }, { "epoch": 2.48045757864633, "grad_norm": 6.867214017234363, "learning_rate": 8.945492161759945e-07, "loss": 2.0474, "step": 1302 }, { "epoch": 2.482364156339371, "grad_norm": 6.7932253613821185, "learning_rate": 8.882318009464124e-07, "loss": 2.0474, "step": 1303 }, { "epoch": 2.4842707340324117, "grad_norm": 6.664362422658804, "learning_rate": 8.819345967488601e-07, "loss": 2.1575, "step": 1304 }, { "epoch": 2.4861773117254526, "grad_norm": 6.145088070237586, "learning_rate": 8.756576345366863e-07, "loss": 2.0794, "step": 1305 }, { "epoch": 2.488083889418494, "grad_norm": 6.345510034310956, "learning_rate": 8.69400945163743e-07, "loss": 2.1243, "step": 1306 }, { "epoch": 2.489990467111535, "grad_norm": 7.068815833364072, "learning_rate": 8.631645593842303e-07, "loss": 2.2017, "step": 1307 }, { "epoch": 2.491897044804576, "grad_norm": 6.83475763432575, "learning_rate": 8.569485078525508e-07, "loss": 1.9851, "step": 1308 }, { "epoch": 2.493803622497617, "grad_norm": 6.154182689096151, "learning_rate": 8.507528211231542e-07, "loss": 1.9117, "step": 1309 }, { "epoch": 2.4957102001906577, "grad_norm": 6.988557470620112, "learning_rate": 8.445775296503905e-07, "loss": 2.1931, "step": 1310 }, { "epoch": 2.4976167778836986, "grad_norm": 6.402144828494648, "learning_rate": 8.384226637883541e-07, "loss": 2.0553, "step": 1311 }, { "epoch": 2.4995233555767395, "grad_norm": 7.002094652836175, "learning_rate": 8.322882537907473e-07, "loss": 2.0343, "step": 1312 }, { "epoch": 2.5014299332697805, "grad_norm": 6.12815352567569, "learning_rate": 8.26174329810715e-07, "loss": 2.1176, "step": 1313 }, { "epoch": 2.503336510962822, "grad_norm": 6.755946556854752, "learning_rate": 8.200809219007095e-07, "loss": 2.1748, "step": 1314 }, { "epoch": 2.5052430886558628, "grad_norm": 7.001709998180338, "learning_rate": 8.140080600123368e-07, "loss": 2.1604, "step": 1315 }, { "epoch": 2.5071496663489037, "grad_norm": 6.788322293679778, "learning_rate": 8.079557739962129e-07, "loss": 2.2213, "step": 1316 }, { "epoch": 2.5090562440419446, "grad_norm": 7.285839098316677, "learning_rate": 8.019240936018108e-07, "loss": 2.2792, "step": 1317 }, { "epoch": 2.510962821734986, "grad_norm": 6.664809948452973, "learning_rate": 7.959130484773208e-07, "loss": 2.1808, "step": 1318 }, { "epoch": 2.512869399428027, "grad_norm": 6.596021540707343, "learning_rate": 7.899226681695038e-07, "loss": 2.1417, "step": 1319 }, { "epoch": 2.514775977121068, "grad_norm": 6.699978197613699, "learning_rate": 7.83952982123542e-07, "loss": 2.1276, "step": 1320 }, { "epoch": 2.5166825548141087, "grad_norm": 6.380873994892273, "learning_rate": 7.78004019682898e-07, "loss": 2.0032, "step": 1321 }, { "epoch": 2.5185891325071497, "grad_norm": 7.130165307813641, "learning_rate": 7.720758100891707e-07, "loss": 2.1752, "step": 1322 }, { "epoch": 2.5204957102001906, "grad_norm": 7.277088465558861, "learning_rate": 7.661683824819465e-07, "loss": 2.3101, "step": 1323 }, { "epoch": 2.5224022878932315, "grad_norm": 6.80704118480283, "learning_rate": 7.602817658986628e-07, "loss": 2.1769, "step": 1324 }, { "epoch": 2.5243088655862724, "grad_norm": 7.6933986156942575, "learning_rate": 7.544159892744618e-07, "loss": 2.0021, "step": 1325 }, { "epoch": 2.5262154432793134, "grad_norm": 6.609165600762268, "learning_rate": 7.485710814420499e-07, "loss": 2.0558, "step": 1326 }, { "epoch": 2.5281220209723547, "grad_norm": 6.037073468877887, "learning_rate": 7.427470711315493e-07, "loss": 1.9775, "step": 1327 }, { "epoch": 2.5300285986653956, "grad_norm": 6.659049465634622, "learning_rate": 7.36943986970371e-07, "loss": 2.2555, "step": 1328 }, { "epoch": 2.5319351763584366, "grad_norm": 7.418789038068912, "learning_rate": 7.31161857483057e-07, "loss": 2.0796, "step": 1329 }, { "epoch": 2.5338417540514775, "grad_norm": 7.416219248550903, "learning_rate": 7.254007110911543e-07, "loss": 2.2449, "step": 1330 }, { "epoch": 2.535748331744519, "grad_norm": 6.747466963427126, "learning_rate": 7.19660576113066e-07, "loss": 2.0304, "step": 1331 }, { "epoch": 2.53765490943756, "grad_norm": 6.857403146594891, "learning_rate": 7.139414807639189e-07, "loss": 2.1781, "step": 1332 }, { "epoch": 2.5395614871306007, "grad_norm": 6.573668207710863, "learning_rate": 7.082434531554161e-07, "loss": 2.0814, "step": 1333 }, { "epoch": 2.5414680648236416, "grad_norm": 6.081326662288895, "learning_rate": 7.025665212957084e-07, "loss": 2.1903, "step": 1334 }, { "epoch": 2.5433746425166825, "grad_norm": 6.527385309617562, "learning_rate": 6.969107130892511e-07, "loss": 2.1482, "step": 1335 }, { "epoch": 2.5452812202097235, "grad_norm": 7.282067827971279, "learning_rate": 6.912760563366694e-07, "loss": 2.1799, "step": 1336 }, { "epoch": 2.5471877979027644, "grad_norm": 6.320401544636059, "learning_rate": 6.85662578734615e-07, "loss": 2.1712, "step": 1337 }, { "epoch": 2.5490943755958053, "grad_norm": 6.6696484311807716, "learning_rate": 6.800703078756443e-07, "loss": 2.1497, "step": 1338 }, { "epoch": 2.5510009532888467, "grad_norm": 7.663694214726258, "learning_rate": 6.744992712480652e-07, "loss": 2.2638, "step": 1339 }, { "epoch": 2.5529075309818876, "grad_norm": 6.843995085770422, "learning_rate": 6.689494962358161e-07, "loss": 2.2201, "step": 1340 }, { "epoch": 2.5548141086749285, "grad_norm": 7.090520285491839, "learning_rate": 6.63421010118323e-07, "loss": 2.1234, "step": 1341 }, { "epoch": 2.5567206863679695, "grad_norm": 6.813911561921021, "learning_rate": 6.579138400703716e-07, "loss": 2.1218, "step": 1342 }, { "epoch": 2.5586272640610104, "grad_norm": 6.027359069631765, "learning_rate": 6.524280131619648e-07, "loss": 2.018, "step": 1343 }, { "epoch": 2.5605338417540517, "grad_norm": 6.218993896356816, "learning_rate": 6.469635563581994e-07, "loss": 1.9326, "step": 1344 }, { "epoch": 2.5624404194470927, "grad_norm": 6.519206771326411, "learning_rate": 6.415204965191291e-07, "loss": 2.1338, "step": 1345 }, { "epoch": 2.5643469971401336, "grad_norm": 6.83769830633941, "learning_rate": 6.360988603996326e-07, "loss": 2.072, "step": 1346 }, { "epoch": 2.5662535748331745, "grad_norm": 6.531143453085927, "learning_rate": 6.306986746492788e-07, "loss": 2.0945, "step": 1347 }, { "epoch": 2.5681601525262154, "grad_norm": 7.05667788902347, "learning_rate": 6.253199658122056e-07, "loss": 2.1788, "step": 1348 }, { "epoch": 2.5700667302192564, "grad_norm": 6.557974823203261, "learning_rate": 6.199627603269764e-07, "loss": 1.9948, "step": 1349 }, { "epoch": 2.5719733079122973, "grad_norm": 6.86687352076149, "learning_rate": 6.146270845264607e-07, "loss": 2.1814, "step": 1350 }, { "epoch": 2.573879885605338, "grad_norm": 6.565053063042379, "learning_rate": 6.093129646377006e-07, "loss": 2.1015, "step": 1351 }, { "epoch": 2.5757864632983796, "grad_norm": 6.33705059013816, "learning_rate": 6.040204267817817e-07, "loss": 2.0469, "step": 1352 }, { "epoch": 2.5776930409914205, "grad_norm": 7.0023473068552, "learning_rate": 5.987494969737023e-07, "loss": 2.0884, "step": 1353 }, { "epoch": 2.5795996186844614, "grad_norm": 7.181024832216259, "learning_rate": 5.935002011222518e-07, "loss": 2.1218, "step": 1354 }, { "epoch": 2.5815061963775023, "grad_norm": 6.711296211292987, "learning_rate": 5.882725650298787e-07, "loss": 2.1799, "step": 1355 }, { "epoch": 2.5834127740705433, "grad_norm": 6.864271237426593, "learning_rate": 5.830666143925657e-07, "loss": 2.0624, "step": 1356 }, { "epoch": 2.5853193517635846, "grad_norm": 6.6131568922778445, "learning_rate": 5.778823747996981e-07, "loss": 2.0571, "step": 1357 }, { "epoch": 2.5872259294566256, "grad_norm": 6.540094191322392, "learning_rate": 5.727198717339511e-07, "loss": 1.9556, "step": 1358 }, { "epoch": 2.5891325071496665, "grad_norm": 7.073514078144344, "learning_rate": 5.675791305711465e-07, "loss": 2.0489, "step": 1359 }, { "epoch": 2.5910390848427074, "grad_norm": 7.089850398444931, "learning_rate": 5.624601765801435e-07, "loss": 2.152, "step": 1360 }, { "epoch": 2.5929456625357483, "grad_norm": 6.780173532952317, "learning_rate": 5.573630349227066e-07, "loss": 2.2812, "step": 1361 }, { "epoch": 2.5948522402287892, "grad_norm": 6.952904011332501, "learning_rate": 5.522877306533841e-07, "loss": 2.3102, "step": 1362 }, { "epoch": 2.59675881792183, "grad_norm": 6.998618919438626, "learning_rate": 5.472342887193832e-07, "loss": 2.2104, "step": 1363 }, { "epoch": 2.598665395614871, "grad_norm": 6.852759975015944, "learning_rate": 5.422027339604496e-07, "loss": 2.1864, "step": 1364 }, { "epoch": 2.6005719733079125, "grad_norm": 6.407368486472391, "learning_rate": 5.371930911087448e-07, "loss": 1.9413, "step": 1365 }, { "epoch": 2.6024785510009534, "grad_norm": 7.189443574385448, "learning_rate": 5.32205384788726e-07, "loss": 2.1101, "step": 1366 }, { "epoch": 2.6043851286939943, "grad_norm": 7.305652738603871, "learning_rate": 5.272396395170171e-07, "loss": 2.1907, "step": 1367 }, { "epoch": 2.6062917063870352, "grad_norm": 6.606792095541143, "learning_rate": 5.222958797023036e-07, "loss": 2.0112, "step": 1368 }, { "epoch": 2.608198284080076, "grad_norm": 6.763393999131886, "learning_rate": 5.173741296451957e-07, "loss": 2.0805, "step": 1369 }, { "epoch": 2.6101048617731175, "grad_norm": 7.030932191990372, "learning_rate": 5.124744135381193e-07, "loss": 2.1793, "step": 1370 }, { "epoch": 2.6120114394661584, "grad_norm": 7.260274077377693, "learning_rate": 5.075967554651956e-07, "loss": 2.1821, "step": 1371 }, { "epoch": 2.6139180171591994, "grad_norm": 6.591183557264943, "learning_rate": 5.027411794021214e-07, "loss": 2.1181, "step": 1372 }, { "epoch": 2.6158245948522403, "grad_norm": 7.412880785453292, "learning_rate": 4.979077092160472e-07, "loss": 2.2198, "step": 1373 }, { "epoch": 2.617731172545281, "grad_norm": 7.87435727596157, "learning_rate": 4.9309636866547e-07, "loss": 2.2801, "step": 1374 }, { "epoch": 2.619637750238322, "grad_norm": 7.1088335645848355, "learning_rate": 4.883071814001056e-07, "loss": 2.1283, "step": 1375 }, { "epoch": 2.621544327931363, "grad_norm": 7.153452100647113, "learning_rate": 4.835401709607823e-07, "loss": 2.0719, "step": 1376 }, { "epoch": 2.623450905624404, "grad_norm": 7.15636041272107, "learning_rate": 4.787953607793127e-07, "loss": 2.1027, "step": 1377 }, { "epoch": 2.6253574833174453, "grad_norm": 6.875074582540537, "learning_rate": 4.740727741783957e-07, "loss": 2.2214, "step": 1378 }, { "epoch": 2.6272640610104863, "grad_norm": 7.156657903891168, "learning_rate": 4.693724343714834e-07, "loss": 2.1071, "step": 1379 }, { "epoch": 2.629170638703527, "grad_norm": 7.305037157331129, "learning_rate": 4.646943644626811e-07, "loss": 2.2041, "step": 1380 }, { "epoch": 2.631077216396568, "grad_norm": 7.228503908554767, "learning_rate": 4.6003858744662564e-07, "loss": 2.2368, "step": 1381 }, { "epoch": 2.632983794089609, "grad_norm": 6.66275605176191, "learning_rate": 4.554051262083775e-07, "loss": 2.0093, "step": 1382 }, { "epoch": 2.6348903717826504, "grad_norm": 6.624784043963348, "learning_rate": 4.5079400352330325e-07, "loss": 2.0502, "step": 1383 }, { "epoch": 2.6367969494756913, "grad_norm": 6.387323358337287, "learning_rate": 4.462052420569679e-07, "loss": 1.9918, "step": 1384 }, { "epoch": 2.6387035271687322, "grad_norm": 6.239683899731278, "learning_rate": 4.416388643650221e-07, "loss": 2.0828, "step": 1385 }, { "epoch": 2.640610104861773, "grad_norm": 6.829234219883221, "learning_rate": 4.370948928930907e-07, "loss": 2.2069, "step": 1386 }, { "epoch": 2.642516682554814, "grad_norm": 6.300705077837421, "learning_rate": 4.32573349976661e-07, "loss": 2.2285, "step": 1387 }, { "epoch": 2.644423260247855, "grad_norm": 6.142289596733273, "learning_rate": 4.2807425784097855e-07, "loss": 2.0056, "step": 1388 }, { "epoch": 2.646329837940896, "grad_norm": 6.908824109923565, "learning_rate": 4.235976386009305e-07, "loss": 2.1104, "step": 1389 }, { "epoch": 2.648236415633937, "grad_norm": 6.653814869916138, "learning_rate": 4.191435142609418e-07, "loss": 2.2022, "step": 1390 }, { "epoch": 2.6501429933269782, "grad_norm": 6.79455882635331, "learning_rate": 4.147119067148664e-07, "loss": 2.0403, "step": 1391 }, { "epoch": 2.652049571020019, "grad_norm": 7.155223616298371, "learning_rate": 4.103028377458784e-07, "loss": 2.1166, "step": 1392 }, { "epoch": 2.65395614871306, "grad_norm": 7.195978984939847, "learning_rate": 4.0591632902636347e-07, "loss": 2.0143, "step": 1393 }, { "epoch": 2.655862726406101, "grad_norm": 6.595101657001995, "learning_rate": 4.0155240211781966e-07, "loss": 1.9564, "step": 1394 }, { "epoch": 2.657769304099142, "grad_norm": 6.835433519685526, "learning_rate": 3.972110784707395e-07, "loss": 2.1537, "step": 1395 }, { "epoch": 2.6596758817921833, "grad_norm": 7.016672208212436, "learning_rate": 3.92892379424516e-07, "loss": 2.1572, "step": 1396 }, { "epoch": 2.661582459485224, "grad_norm": 6.6349657467251735, "learning_rate": 3.885963262073317e-07, "loss": 2.1894, "step": 1397 }, { "epoch": 2.663489037178265, "grad_norm": 6.762461033292943, "learning_rate": 3.8432293993605543e-07, "loss": 2.0113, "step": 1398 }, { "epoch": 2.665395614871306, "grad_norm": 6.493239553737912, "learning_rate": 3.8007224161613734e-07, "loss": 2.0479, "step": 1399 }, { "epoch": 2.667302192564347, "grad_norm": 7.2911931457699355, "learning_rate": 3.758442521415084e-07, "loss": 2.1314, "step": 1400 }, { "epoch": 2.669208770257388, "grad_norm": 6.550418330960005, "learning_rate": 3.716389922944769e-07, "loss": 2.2146, "step": 1401 }, { "epoch": 2.671115347950429, "grad_norm": 7.107629824648081, "learning_rate": 3.674564827456251e-07, "loss": 2.1134, "step": 1402 }, { "epoch": 2.6730219256434697, "grad_norm": 6.265814132466134, "learning_rate": 3.63296744053705e-07, "loss": 2.1288, "step": 1403 }, { "epoch": 2.674928503336511, "grad_norm": 7.006039937177485, "learning_rate": 3.591597966655469e-07, "loss": 2.177, "step": 1404 }, { "epoch": 2.676835081029552, "grad_norm": 6.66112855437115, "learning_rate": 3.5504566091594694e-07, "loss": 2.1732, "step": 1405 }, { "epoch": 2.678741658722593, "grad_norm": 6.914474829346755, "learning_rate": 3.509543570275753e-07, "loss": 2.1085, "step": 1406 }, { "epoch": 2.680648236415634, "grad_norm": 7.229141511227422, "learning_rate": 3.4688590511087304e-07, "loss": 2.2831, "step": 1407 }, { "epoch": 2.682554814108675, "grad_norm": 6.469428354841321, "learning_rate": 3.428403251639556e-07, "loss": 2.0655, "step": 1408 }, { "epoch": 2.684461391801716, "grad_norm": 6.686941494825127, "learning_rate": 3.388176370725116e-07, "loss": 2.1607, "step": 1409 }, { "epoch": 2.686367969494757, "grad_norm": 7.036997310733365, "learning_rate": 3.34817860609708e-07, "loss": 2.194, "step": 1410 }, { "epoch": 2.688274547187798, "grad_norm": 6.321613826875935, "learning_rate": 3.3084101543609136e-07, "loss": 2.1512, "step": 1411 }, { "epoch": 2.690181124880839, "grad_norm": 7.087108310965536, "learning_rate": 3.2688712109949283e-07, "loss": 1.9646, "step": 1412 }, { "epoch": 2.69208770257388, "grad_norm": 6.5729635036999925, "learning_rate": 3.2295619703492597e-07, "loss": 2.1455, "step": 1413 }, { "epoch": 2.693994280266921, "grad_norm": 7.0364161883788245, "learning_rate": 3.1904826256450414e-07, "loss": 2.2224, "step": 1414 }, { "epoch": 2.6959008579599617, "grad_norm": 7.448889794410271, "learning_rate": 3.151633368973295e-07, "loss": 2.2862, "step": 1415 }, { "epoch": 2.6978074356530026, "grad_norm": 6.766437678648138, "learning_rate": 3.1130143912941056e-07, "loss": 2.1583, "step": 1416 }, { "epoch": 2.699714013346044, "grad_norm": 6.525641276466056, "learning_rate": 3.0746258824356336e-07, "loss": 2.0814, "step": 1417 }, { "epoch": 2.701620591039085, "grad_norm": 6.726268157763317, "learning_rate": 3.036468031093204e-07, "loss": 2.0858, "step": 1418 }, { "epoch": 2.703527168732126, "grad_norm": 6.787024212055952, "learning_rate": 2.9985410248283274e-07, "loss": 2.1883, "step": 1419 }, { "epoch": 2.7054337464251668, "grad_norm": 6.89859403494982, "learning_rate": 2.9608450500678566e-07, "loss": 2.3843, "step": 1420 }, { "epoch": 2.7073403241182077, "grad_norm": 6.49775670639588, "learning_rate": 2.923380292103012e-07, "loss": 2.0688, "step": 1421 }, { "epoch": 2.709246901811249, "grad_norm": 6.956049655584201, "learning_rate": 2.886146935088491e-07, "loss": 2.0536, "step": 1422 }, { "epoch": 2.71115347950429, "grad_norm": 7.008357716232977, "learning_rate": 2.849145162041544e-07, "loss": 2.0647, "step": 1423 }, { "epoch": 2.713060057197331, "grad_norm": 6.346321164391841, "learning_rate": 2.8123751548411293e-07, "loss": 2.0095, "step": 1424 }, { "epoch": 2.714966634890372, "grad_norm": 7.124093771215006, "learning_rate": 2.775837094226941e-07, "loss": 2.1807, "step": 1425 }, { "epoch": 2.7168732125834127, "grad_norm": 6.624560937253457, "learning_rate": 2.739531159798581e-07, "loss": 2.162, "step": 1426 }, { "epoch": 2.7187797902764537, "grad_norm": 6.463694155868496, "learning_rate": 2.703457530014647e-07, "loss": 2.0614, "step": 1427 }, { "epoch": 2.7206863679694946, "grad_norm": 6.625499816602903, "learning_rate": 2.667616382191879e-07, "loss": 2.0929, "step": 1428 }, { "epoch": 2.7225929456625355, "grad_norm": 8.01232330176841, "learning_rate": 2.6320078925042516e-07, "loss": 2.3342, "step": 1429 }, { "epoch": 2.724499523355577, "grad_norm": 6.8165047817424, "learning_rate": 2.596632235982138e-07, "loss": 2.0387, "step": 1430 }, { "epoch": 2.726406101048618, "grad_norm": 6.705622223428054, "learning_rate": 2.561489586511445e-07, "loss": 2.1467, "step": 1431 }, { "epoch": 2.7283126787416587, "grad_norm": 6.340524216562229, "learning_rate": 2.526580116832761e-07, "loss": 2.1607, "step": 1432 }, { "epoch": 2.7302192564346996, "grad_norm": 7.46396180415956, "learning_rate": 2.4919039985404626e-07, "loss": 2.2158, "step": 1433 }, { "epoch": 2.7321258341277406, "grad_norm": 6.587404374834363, "learning_rate": 2.457461402081962e-07, "loss": 2.1375, "step": 1434 }, { "epoch": 2.734032411820782, "grad_norm": 7.679307028216608, "learning_rate": 2.423252496756773e-07, "loss": 2.1515, "step": 1435 }, { "epoch": 2.735938989513823, "grad_norm": 6.420025416197147, "learning_rate": 2.389277450715738e-07, "loss": 2.0238, "step": 1436 }, { "epoch": 2.737845567206864, "grad_norm": 7.209075268958359, "learning_rate": 2.355536430960187e-07, "loss": 2.1825, "step": 1437 }, { "epoch": 2.7397521448999047, "grad_norm": 6.405254434325643, "learning_rate": 2.3220296033411137e-07, "loss": 2.1059, "step": 1438 }, { "epoch": 2.7416587225929456, "grad_norm": 6.918646579880431, "learning_rate": 2.2887571325583556e-07, "loss": 2.1008, "step": 1439 }, { "epoch": 2.7435653002859866, "grad_norm": 6.946647243920879, "learning_rate": 2.255719182159788e-07, "loss": 2.0314, "step": 1440 }, { "epoch": 2.7454718779790275, "grad_norm": 6.861477560275952, "learning_rate": 2.2229159145405298e-07, "loss": 2.1339, "step": 1441 }, { "epoch": 2.7473784556720684, "grad_norm": 6.864269297180268, "learning_rate": 2.1903474909421462e-07, "loss": 2.1171, "step": 1442 }, { "epoch": 2.7492850333651098, "grad_norm": 6.465585467135994, "learning_rate": 2.1580140714518138e-07, "loss": 2.0579, "step": 1443 }, { "epoch": 2.7511916110581507, "grad_norm": 6.989881209780569, "learning_rate": 2.1259158150016111e-07, "loss": 2.064, "step": 1444 }, { "epoch": 2.7530981887511916, "grad_norm": 7.324771950824169, "learning_rate": 2.0940528793676529e-07, "loss": 2.3974, "step": 1445 }, { "epoch": 2.7550047664442325, "grad_norm": 6.321813497782112, "learning_rate": 2.0624254211693894e-07, "loss": 1.9029, "step": 1446 }, { "epoch": 2.7569113441372735, "grad_norm": 6.878894428362027, "learning_rate": 2.0310335958687754e-07, "loss": 1.9856, "step": 1447 }, { "epoch": 2.758817921830315, "grad_norm": 6.491185698234335, "learning_rate": 1.999877557769564e-07, "loss": 2.189, "step": 1448 }, { "epoch": 2.7607244995233557, "grad_norm": 6.873147691755883, "learning_rate": 1.968957460016474e-07, "loss": 2.0687, "step": 1449 }, { "epoch": 2.7626310772163967, "grad_norm": 6.795259653976469, "learning_rate": 1.9382734545945136e-07, "loss": 2.0128, "step": 1450 }, { "epoch": 2.7645376549094376, "grad_norm": 6.885795653359006, "learning_rate": 1.9078256923281968e-07, "loss": 2.1391, "step": 1451 }, { "epoch": 2.7664442326024785, "grad_norm": 6.691124293123898, "learning_rate": 1.8776143228807997e-07, "loss": 2.2158, "step": 1452 }, { "epoch": 2.7683508102955194, "grad_norm": 7.111504687582559, "learning_rate": 1.847639494753617e-07, "loss": 2.0828, "step": 1453 }, { "epoch": 2.7702573879885604, "grad_norm": 6.439048758865793, "learning_rate": 1.8179013552852786e-07, "loss": 1.9618, "step": 1454 }, { "epoch": 2.7721639656816013, "grad_norm": 6.559942569922363, "learning_rate": 1.788400050650957e-07, "loss": 2.2388, "step": 1455 }, { "epoch": 2.7740705433746427, "grad_norm": 6.73130902068663, "learning_rate": 1.759135725861699e-07, "loss": 2.0569, "step": 1456 }, { "epoch": 2.7759771210676836, "grad_norm": 6.568240852432341, "learning_rate": 1.7301085247637073e-07, "loss": 2.1202, "step": 1457 }, { "epoch": 2.7778836987607245, "grad_norm": 6.442636736435364, "learning_rate": 1.7013185900376207e-07, "loss": 2.1452, "step": 1458 }, { "epoch": 2.7797902764537654, "grad_norm": 6.565112060849425, "learning_rate": 1.6727660631977894e-07, "loss": 2.0199, "step": 1459 }, { "epoch": 2.7816968541468063, "grad_norm": 6.598915419262629, "learning_rate": 1.6444510845916474e-07, "loss": 2.1544, "step": 1460 }, { "epoch": 2.7836034318398477, "grad_norm": 7.020804999757374, "learning_rate": 1.6163737933989508e-07, "loss": 2.2349, "step": 1461 }, { "epoch": 2.7855100095328886, "grad_norm": 7.773375386066651, "learning_rate": 1.5885343276311417e-07, "loss": 2.1172, "step": 1462 }, { "epoch": 2.7874165872259296, "grad_norm": 6.1757588631821925, "learning_rate": 1.5609328241306344e-07, "loss": 1.981, "step": 1463 }, { "epoch": 2.7893231649189705, "grad_norm": 6.97207252638456, "learning_rate": 1.533569418570191e-07, "loss": 2.1485, "step": 1464 }, { "epoch": 2.7912297426120114, "grad_norm": 6.378157075892338, "learning_rate": 1.5064442454521876e-07, "loss": 2.0753, "step": 1465 }, { "epoch": 2.7931363203050523, "grad_norm": 6.469401640873402, "learning_rate": 1.47955743810802e-07, "loss": 2.3039, "step": 1466 }, { "epoch": 2.7950428979980932, "grad_norm": 6.937717276348364, "learning_rate": 1.4529091286973994e-07, "loss": 2.2159, "step": 1467 }, { "epoch": 2.796949475691134, "grad_norm": 7.180287249363699, "learning_rate": 1.42649944820773e-07, "loss": 2.1354, "step": 1468 }, { "epoch": 2.7988560533841755, "grad_norm": 6.565589752231693, "learning_rate": 1.4003285264534483e-07, "loss": 2.1616, "step": 1469 }, { "epoch": 2.8007626310772165, "grad_norm": 6.546309590617579, "learning_rate": 1.3743964920754138e-07, "loss": 2.1096, "step": 1470 }, { "epoch": 2.8026692087702574, "grad_norm": 6.979728898177081, "learning_rate": 1.348703472540236e-07, "loss": 2.1927, "step": 1471 }, { "epoch": 2.8045757864632983, "grad_norm": 7.009203618925823, "learning_rate": 1.323249594139664e-07, "loss": 2.2221, "step": 1472 }, { "epoch": 2.8064823641563392, "grad_norm": 6.535200032049088, "learning_rate": 1.298034981989993e-07, "loss": 2.1171, "step": 1473 }, { "epoch": 2.8083889418493806, "grad_norm": 7.949485301305099, "learning_rate": 1.2730597600314144e-07, "loss": 2.1154, "step": 1474 }, { "epoch": 2.8102955195424215, "grad_norm": 6.348570472703038, "learning_rate": 1.2483240510274108e-07, "loss": 1.9853, "step": 1475 }, { "epoch": 2.8122020972354624, "grad_norm": 7.40652183558425, "learning_rate": 1.2238279765641625e-07, "loss": 2.2543, "step": 1476 }, { "epoch": 2.8141086749285034, "grad_norm": 6.575648581316472, "learning_rate": 1.1995716570499638e-07, "loss": 2.1391, "step": 1477 }, { "epoch": 2.8160152526215443, "grad_norm": 6.8372730788941904, "learning_rate": 1.1755552117145963e-07, "loss": 2.157, "step": 1478 }, { "epoch": 2.817921830314585, "grad_norm": 6.80829358693753, "learning_rate": 1.1517787586087626e-07, "loss": 2.0837, "step": 1479 }, { "epoch": 2.819828408007626, "grad_norm": 6.11905845289894, "learning_rate": 1.1282424146035254e-07, "loss": 2.0266, "step": 1480 }, { "epoch": 2.821734985700667, "grad_norm": 6.939557087705344, "learning_rate": 1.1049462953896806e-07, "loss": 2.0698, "step": 1481 }, { "epoch": 2.8236415633937084, "grad_norm": 6.413024331458544, "learning_rate": 1.0818905154772408e-07, "loss": 2.0559, "step": 1482 }, { "epoch": 2.8255481410867493, "grad_norm": 6.024771417603635, "learning_rate": 1.0590751881948525e-07, "loss": 2.0851, "step": 1483 }, { "epoch": 2.8274547187797903, "grad_norm": 6.8465122523323005, "learning_rate": 1.0365004256892352e-07, "loss": 2.1451, "step": 1484 }, { "epoch": 2.829361296472831, "grad_norm": 6.850233290958077, "learning_rate": 1.014166338924627e-07, "loss": 2.2087, "step": 1485 }, { "epoch": 2.831267874165872, "grad_norm": 6.279575302559668, "learning_rate": 9.920730376822508e-08, "loss": 2.1443, "step": 1486 }, { "epoch": 2.8331744518589135, "grad_norm": 7.446235071015882, "learning_rate": 9.702206305597761e-08, "loss": 2.0684, "step": 1487 }, { "epoch": 2.8350810295519544, "grad_norm": 6.811462655370548, "learning_rate": 9.486092249707701e-08, "loss": 2.0616, "step": 1488 }, { "epoch": 2.8369876072449953, "grad_norm": 6.269532363331429, "learning_rate": 9.272389271441751e-08, "loss": 2.0586, "step": 1489 }, { "epoch": 2.8388941849380362, "grad_norm": 7.844214879009372, "learning_rate": 9.061098421238035e-08, "loss": 2.2747, "step": 1490 }, { "epoch": 2.840800762631077, "grad_norm": 6.531795140951721, "learning_rate": 8.85222073767794e-08, "loss": 2.247, "step": 1491 }, { "epoch": 2.842707340324118, "grad_norm": 6.687263245144009, "learning_rate": 8.645757247481234e-08, "loss": 2.1569, "step": 1492 }, { "epoch": 2.844613918017159, "grad_norm": 6.933017677573418, "learning_rate": 8.441708965500895e-08, "loss": 2.2105, "step": 1493 }, { "epoch": 2.8465204957102, "grad_norm": 6.754527821972973, "learning_rate": 8.240076894718175e-08, "loss": 2.1698, "step": 1494 }, { "epoch": 2.8484270734032413, "grad_norm": 6.494917804600435, "learning_rate": 8.040862026237605e-08, "loss": 2.1279, "step": 1495 }, { "epoch": 2.8503336510962822, "grad_norm": 6.4259246387177145, "learning_rate": 7.84406533928217e-08, "loss": 2.0504, "step": 1496 }, { "epoch": 2.852240228789323, "grad_norm": 7.1086731833391665, "learning_rate": 7.649687801188633e-08, "loss": 2.1642, "step": 1497 }, { "epoch": 2.854146806482364, "grad_norm": 6.581394356234226, "learning_rate": 7.45773036740255e-08, "loss": 2.1051, "step": 1498 }, { "epoch": 2.856053384175405, "grad_norm": 6.856641083855075, "learning_rate": 7.268193981473548e-08, "loss": 2.2814, "step": 1499 }, { "epoch": 2.8579599618684464, "grad_norm": 6.615798894579222, "learning_rate": 7.081079575051053e-08, "loss": 2.1829, "step": 1500 }, { "epoch": 2.8598665395614873, "grad_norm": 6.652648661280042, "learning_rate": 6.896388067879233e-08, "loss": 2.0796, "step": 1501 }, { "epoch": 2.861773117254528, "grad_norm": 7.181887530409291, "learning_rate": 6.714120367792897e-08, "loss": 2.1742, "step": 1502 }, { "epoch": 2.863679694947569, "grad_norm": 6.767113324406131, "learning_rate": 6.534277370712771e-08, "loss": 2.0749, "step": 1503 }, { "epoch": 2.86558627264061, "grad_norm": 6.853966511860557, "learning_rate": 6.356859960641116e-08, "loss": 2.2366, "step": 1504 }, { "epoch": 2.867492850333651, "grad_norm": 7.116851660867472, "learning_rate": 6.181869009657448e-08, "loss": 2.0008, "step": 1505 }, { "epoch": 2.869399428026692, "grad_norm": 6.3071636370749244, "learning_rate": 6.009305377914387e-08, "loss": 2.0602, "step": 1506 }, { "epoch": 2.871306005719733, "grad_norm": 6.802117752906428, "learning_rate": 5.839169913633036e-08, "loss": 2.2371, "step": 1507 }, { "epoch": 2.873212583412774, "grad_norm": 6.990085317222018, "learning_rate": 5.671463453099324e-08, "loss": 2.1654, "step": 1508 }, { "epoch": 2.875119161105815, "grad_norm": 7.1027985521984975, "learning_rate": 5.5061868206593425e-08, "loss": 2.2483, "step": 1509 }, { "epoch": 2.877025738798856, "grad_norm": 6.807755404243176, "learning_rate": 5.3433408287158486e-08, "loss": 2.2783, "step": 1510 }, { "epoch": 2.878932316491897, "grad_norm": 6.618566875284323, "learning_rate": 5.182926277723821e-08, "loss": 2.0719, "step": 1511 }, { "epoch": 2.880838894184938, "grad_norm": 6.450248562477616, "learning_rate": 5.02494395618669e-08, "loss": 2.1869, "step": 1512 }, { "epoch": 2.8827454718779792, "grad_norm": 6.52601366577178, "learning_rate": 4.8693946406526695e-08, "loss": 2.0352, "step": 1513 }, { "epoch": 2.88465204957102, "grad_norm": 6.525661998217902, "learning_rate": 4.7162790957106516e-08, "loss": 1.9684, "step": 1514 }, { "epoch": 2.886558627264061, "grad_norm": 6.5481179181414495, "learning_rate": 4.565598073986377e-08, "loss": 2.1363, "step": 1515 }, { "epoch": 2.888465204957102, "grad_norm": 6.693801472243049, "learning_rate": 4.4173523161391566e-08, "loss": 2.1311, "step": 1516 }, { "epoch": 2.890371782650143, "grad_norm": 6.090902227196953, "learning_rate": 4.271542550857821e-08, "loss": 2.1467, "step": 1517 }, { "epoch": 2.892278360343184, "grad_norm": 6.588272267272736, "learning_rate": 4.128169494857393e-08, "loss": 2.1049, "step": 1518 }, { "epoch": 2.894184938036225, "grad_norm": 6.46431418612604, "learning_rate": 3.9872338528753604e-08, "loss": 2.1377, "step": 1519 }, { "epoch": 2.8960915157292657, "grad_norm": 6.444773978299536, "learning_rate": 3.848736317668411e-08, "loss": 2.0634, "step": 1520 }, { "epoch": 2.897998093422307, "grad_norm": 7.293240444552128, "learning_rate": 3.712677570008927e-08, "loss": 2.038, "step": 1521 }, { "epoch": 2.899904671115348, "grad_norm": 6.806721347260175, "learning_rate": 3.579058278681546e-08, "loss": 2.1005, "step": 1522 }, { "epoch": 2.901811248808389, "grad_norm": 6.311204356503324, "learning_rate": 3.447879100480056e-08, "loss": 2.0242, "step": 1523 }, { "epoch": 2.90371782650143, "grad_norm": 6.618052892444979, "learning_rate": 3.3191406802041693e-08, "loss": 2.0065, "step": 1524 }, { "epoch": 2.9056244041944708, "grad_norm": 7.470686056032922, "learning_rate": 3.1928436506560854e-08, "loss": 2.1149, "step": 1525 }, { "epoch": 2.907530981887512, "grad_norm": 7.634714801278908, "learning_rate": 3.0689886326376593e-08, "loss": 2.1607, "step": 1526 }, { "epoch": 2.909437559580553, "grad_norm": 6.444174733685101, "learning_rate": 2.9475762349472915e-08, "loss": 1.9434, "step": 1527 }, { "epoch": 2.911344137273594, "grad_norm": 7.0332172179237, "learning_rate": 2.828607054376875e-08, "loss": 2.0809, "step": 1528 }, { "epoch": 2.913250714966635, "grad_norm": 6.0864457148376765, "learning_rate": 2.7120816757087997e-08, "loss": 2.1189, "step": 1529 }, { "epoch": 2.915157292659676, "grad_norm": 6.8793465169048496, "learning_rate": 2.598000671713341e-08, "loss": 2.2595, "step": 1530 }, { "epoch": 2.9170638703527167, "grad_norm": 6.639857758400512, "learning_rate": 2.4863646031453857e-08, "loss": 1.9909, "step": 1531 }, { "epoch": 2.9189704480457577, "grad_norm": 6.504330133873522, "learning_rate": 2.377174018742323e-08, "loss": 2.188, "step": 1532 }, { "epoch": 2.9208770257387986, "grad_norm": 6.775693481372834, "learning_rate": 2.2704294552206574e-08, "loss": 2.0795, "step": 1533 }, { "epoch": 2.92278360343184, "grad_norm": 6.825967013917976, "learning_rate": 2.1661314372738995e-08, "loss": 2.1774, "step": 1534 }, { "epoch": 2.924690181124881, "grad_norm": 7.173672561537188, "learning_rate": 2.0642804775696245e-08, "loss": 2.1094, "step": 1535 }, { "epoch": 2.926596758817922, "grad_norm": 6.415384877468813, "learning_rate": 1.9648770767472515e-08, "loss": 2.1976, "step": 1536 }, { "epoch": 2.9285033365109627, "grad_norm": 6.517782181588419, "learning_rate": 1.8679217234154335e-08, "loss": 2.0461, "step": 1537 }, { "epoch": 2.9304099142040037, "grad_norm": 6.479407207037839, "learning_rate": 1.773414894149561e-08, "loss": 2.1086, "step": 1538 }, { "epoch": 2.932316491897045, "grad_norm": 6.555053182280121, "learning_rate": 1.6813570534895406e-08, "loss": 1.9817, "step": 1539 }, { "epoch": 2.934223069590086, "grad_norm": 7.2786230973318915, "learning_rate": 1.5917486539375747e-08, "loss": 2.361, "step": 1540 }, { "epoch": 2.936129647283127, "grad_norm": 6.618634514202705, "learning_rate": 1.5045901359558302e-08, "loss": 2.054, "step": 1541 }, { "epoch": 2.938036224976168, "grad_norm": 6.32306479729753, "learning_rate": 1.4198819279641618e-08, "loss": 2.0359, "step": 1542 }, { "epoch": 2.9399428026692087, "grad_norm": 6.470513054255001, "learning_rate": 1.3376244463383926e-08, "loss": 2.2292, "step": 1543 }, { "epoch": 2.9418493803622496, "grad_norm": 6.671588785295492, "learning_rate": 1.2578180954078145e-08, "loss": 2.1124, "step": 1544 }, { "epoch": 2.9437559580552906, "grad_norm": 6.696349223267095, "learning_rate": 1.1804632674534688e-08, "loss": 2.1574, "step": 1545 }, { "epoch": 2.9456625357483315, "grad_norm": 6.261378367884958, "learning_rate": 1.1055603427061468e-08, "loss": 1.901, "step": 1546 }, { "epoch": 2.947569113441373, "grad_norm": 6.587168081742577, "learning_rate": 1.033109689344558e-08, "loss": 2.1776, "step": 1547 }, { "epoch": 2.9494756911344138, "grad_norm": 7.993812268090312, "learning_rate": 9.631116634934434e-09, "loss": 2.1233, "step": 1548 }, { "epoch": 2.9513822688274547, "grad_norm": 6.482423974056659, "learning_rate": 8.955666092219094e-09, "loss": 2.0926, "step": 1549 }, { "epoch": 2.9532888465204956, "grad_norm": 6.567020869316282, "learning_rate": 8.304748585417077e-09, "loss": 2.1096, "step": 1550 }, { "epoch": 2.9551954242135365, "grad_norm": 7.050654183748193, "learning_rate": 7.678367314055136e-09, "loss": 2.2116, "step": 1551 }, { "epoch": 2.957102001906578, "grad_norm": 7.333854326087091, "learning_rate": 7.076525357054831e-09, "loss": 2.0994, "step": 1552 }, { "epoch": 2.959008579599619, "grad_norm": 6.986938172299948, "learning_rate": 6.499225672717546e-09, "loss": 2.0188, "step": 1553 }, { "epoch": 2.9609151572926597, "grad_norm": 7.078041643706314, "learning_rate": 5.946471098708384e-09, "loss": 2.1329, "step": 1554 }, { "epoch": 2.9628217349857007, "grad_norm": 7.556569185050528, "learning_rate": 5.418264352043401e-09, "loss": 2.0365, "step": 1555 }, { "epoch": 2.9647283126787416, "grad_norm": 6.891856056100627, "learning_rate": 4.914608029076284e-09, "loss": 2.0674, "step": 1556 }, { "epoch": 2.9666348903717825, "grad_norm": 6.769445300264954, "learning_rate": 4.4355046054844755e-09, "loss": 2.0452, "step": 1557 }, { "epoch": 2.9685414680648234, "grad_norm": 9.85285901784855, "learning_rate": 3.9809564362586205e-09, "loss": 2.204, "step": 1558 }, { "epoch": 2.9704480457578644, "grad_norm": 7.348812826678112, "learning_rate": 3.550965755690361e-09, "loss": 2.2939, "step": 1559 }, { "epoch": 2.9723546234509057, "grad_norm": 7.190857668907964, "learning_rate": 3.145534677360118e-09, "loss": 2.0811, "step": 1560 }, { "epoch": 2.9742612011439467, "grad_norm": 6.524449209757054, "learning_rate": 2.7646651941287682e-09, "loss": 2.0713, "step": 1561 }, { "epoch": 2.9761677788369876, "grad_norm": 6.659730329621004, "learning_rate": 2.4083591781265404e-09, "loss": 2.1711, "step": 1562 }, { "epoch": 2.9780743565300285, "grad_norm": 8.003118094847625, "learning_rate": 2.076618380744133e-09, "loss": 2.2369, "step": 1563 }, { "epoch": 2.97998093422307, "grad_norm": 6.84954861587305, "learning_rate": 1.7694444326238346e-09, "loss": 2.0816, "step": 1564 }, { "epoch": 2.981887511916111, "grad_norm": 6.158951045284458, "learning_rate": 1.48683884365175e-09, "loss": 2.1474, "step": 1565 }, { "epoch": 2.9837940896091517, "grad_norm": 6.767361855112665, "learning_rate": 1.22880300295114e-09, "loss": 2.2001, "step": 1566 }, { "epoch": 2.9857006673021926, "grad_norm": 6.693500517168756, "learning_rate": 9.953381788735394e-10, "loss": 2.1493, "step": 1567 }, { "epoch": 2.9876072449952336, "grad_norm": 6.640213214570163, "learning_rate": 7.864455189954268e-10, "loss": 2.0978, "step": 1568 }, { "epoch": 2.9895138226882745, "grad_norm": 6.30197945499572, "learning_rate": 6.021260501098969e-10, "loss": 2.0546, "step": 1569 }, { "epoch": 2.9914204003813154, "grad_norm": 6.492093256753609, "learning_rate": 4.4238067822222064e-10, "loss": 2.1628, "step": 1570 }, { "epoch": 2.9933269780743563, "grad_norm": 6.798597326581562, "learning_rate": 3.0721018854762417e-10, "loss": 2.1711, "step": 1571 }, { "epoch": 2.9952335557673972, "grad_norm": 6.376220835038747, "learning_rate": 1.9661524550407262e-10, "loss": 2.1868, "step": 1572 }, { "epoch": 2.9971401334604386, "grad_norm": 6.522442619414145, "learning_rate": 1.1059639271171485e-10, "loss": 2.2643, "step": 1573 }, { "epoch": 2.9990467111534795, "grad_norm": 7.545957601725168, "learning_rate": 4.915405298844267e-11, "loss": 2.1834, "step": 1574 }, { "epoch": 3.0, "grad_norm": 7.545957601725168, "learning_rate": 1.228852834767036e-11, "loss": 2.2598, "step": 1575 }, { "epoch": 3.0, "step": 1575, "total_flos": 4403523932160.0, "train_loss": 2.7719489691749453, "train_runtime": 8882.769, "train_samples_per_second": 11.335, "train_steps_per_second": 0.177 } ], "logging_steps": 1, "max_steps": 1575, "num_input_tokens_seen": 0, "num_train_epochs": 3, "save_steps": 500, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 4403523932160.0, "train_batch_size": 1, "trial_name": null, "trial_params": null }