| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.6734006734006734, | |
| "eval_steps": 100, | |
| "global_step": 200, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003367003367003367, | |
| "grad_norm": 190.2997283935547, | |
| "learning_rate": 6.711409395973154e-07, | |
| "loss": 13.9272, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006734006734006734, | |
| "grad_norm": 196.5933074951172, | |
| "learning_rate": 1.3422818791946309e-06, | |
| "loss": 14.3753, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010101010101010102, | |
| "grad_norm": 198.02767944335938, | |
| "learning_rate": 2.013422818791946e-06, | |
| "loss": 14.4143, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013468013468013467, | |
| "grad_norm": 186.30801391601562, | |
| "learning_rate": 2.6845637583892617e-06, | |
| "loss": 13.7729, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.016835016835016835, | |
| "grad_norm": 129.32237243652344, | |
| "learning_rate": 3.3557046979865773e-06, | |
| "loss": 11.4082, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.020202020202020204, | |
| "grad_norm": 123.9930191040039, | |
| "learning_rate": 4.026845637583892e-06, | |
| "loss": 11.5581, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02356902356902357, | |
| "grad_norm": 102.4565658569336, | |
| "learning_rate": 4.697986577181209e-06, | |
| "loss": 9.8311, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.026936026936026935, | |
| "grad_norm": 98.7117919921875, | |
| "learning_rate": 5.3691275167785235e-06, | |
| "loss": 9.825, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 121.9065170288086, | |
| "learning_rate": 6.04026845637584e-06, | |
| "loss": 8.5157, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03367003367003367, | |
| "grad_norm": 93.352294921875, | |
| "learning_rate": 6.7114093959731546e-06, | |
| "loss": 7.6328, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.037037037037037035, | |
| "grad_norm": 108.89420318603516, | |
| "learning_rate": 7.382550335570471e-06, | |
| "loss": 7.1598, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 191.65274047851562, | |
| "learning_rate": 8.053691275167785e-06, | |
| "loss": 6.237, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04377104377104377, | |
| "grad_norm": 150.62646484375, | |
| "learning_rate": 8.724832214765101e-06, | |
| "loss": 5.7063, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04713804713804714, | |
| "grad_norm": 185.48080444335938, | |
| "learning_rate": 9.395973154362418e-06, | |
| "loss": 5.093, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.050505050505050504, | |
| "grad_norm": 1576.556640625, | |
| "learning_rate": 1.006711409395973e-05, | |
| "loss": 8.3575, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05387205387205387, | |
| "grad_norm": 441.4505310058594, | |
| "learning_rate": 1.0738255033557047e-05, | |
| "loss": 4.679, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05723905723905724, | |
| "grad_norm": 499.8016357421875, | |
| "learning_rate": 1.1409395973154363e-05, | |
| "loss": 3.1432, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 472.59747314453125, | |
| "learning_rate": 1.208053691275168e-05, | |
| "loss": 2.9237, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06397306397306397, | |
| "grad_norm": 506.6687927246094, | |
| "learning_rate": 1.2751677852348994e-05, | |
| "loss": 2.6882, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06734006734006734, | |
| "grad_norm": 494.16949462890625, | |
| "learning_rate": 1.3422818791946309e-05, | |
| "loss": 2.4807, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0707070707070707, | |
| "grad_norm": 463.3478698730469, | |
| "learning_rate": 1.4093959731543624e-05, | |
| "loss": 2.2508, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 422.92401123046875, | |
| "learning_rate": 1.4765100671140942e-05, | |
| "loss": 1.9202, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07744107744107744, | |
| "grad_norm": 417.1321105957031, | |
| "learning_rate": 1.5436241610738255e-05, | |
| "loss": 1.6106, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 360.2781677246094, | |
| "learning_rate": 1.610738255033557e-05, | |
| "loss": 1.2741, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08417508417508418, | |
| "grad_norm": 297.3291015625, | |
| "learning_rate": 1.6778523489932888e-05, | |
| "loss": 1.0282, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08754208754208755, | |
| "grad_norm": 195.75958251953125, | |
| "learning_rate": 1.7449664429530202e-05, | |
| "loss": 0.799, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 116.36829376220703, | |
| "learning_rate": 1.8120805369127517e-05, | |
| "loss": 0.6593, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09427609427609428, | |
| "grad_norm": 70.56578063964844, | |
| "learning_rate": 1.8791946308724835e-05, | |
| "loss": 0.5787, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09764309764309764, | |
| "grad_norm": 45.22296905517578, | |
| "learning_rate": 1.946308724832215e-05, | |
| "loss": 0.5196, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10101010101010101, | |
| "grad_norm": 20.37734603881836, | |
| "learning_rate": 2.013422818791946e-05, | |
| "loss": 0.4681, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10437710437710437, | |
| "grad_norm": 7.735367298126221, | |
| "learning_rate": 2.080536912751678e-05, | |
| "loss": 0.4318, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10774410774410774, | |
| "grad_norm": 4.360243797302246, | |
| "learning_rate": 2.1476510067114094e-05, | |
| "loss": 0.4276, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 4.440345287322998, | |
| "learning_rate": 2.2147651006711412e-05, | |
| "loss": 0.4463, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11447811447811448, | |
| "grad_norm": 26.992700576782227, | |
| "learning_rate": 2.2818791946308727e-05, | |
| "loss": 0.4394, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11784511784511785, | |
| "grad_norm": 33.81399917602539, | |
| "learning_rate": 2.348993288590604e-05, | |
| "loss": 0.6005, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 7.8905029296875, | |
| "learning_rate": 2.416107382550336e-05, | |
| "loss": 0.4963, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12457912457912458, | |
| "grad_norm": 2.6311209201812744, | |
| "learning_rate": 2.4832214765100674e-05, | |
| "loss": 0.39, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.12794612794612795, | |
| "grad_norm": 2.389883041381836, | |
| "learning_rate": 2.550335570469799e-05, | |
| "loss": 0.3782, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13131313131313133, | |
| "grad_norm": 2.070525646209717, | |
| "learning_rate": 2.6174496644295304e-05, | |
| "loss": 0.3592, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13468013468013468, | |
| "grad_norm": 5.955089569091797, | |
| "learning_rate": 2.6845637583892618e-05, | |
| "loss": 0.3777, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13804713804713806, | |
| "grad_norm": 6.50673770904541, | |
| "learning_rate": 2.7516778523489933e-05, | |
| "loss": 0.389, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1414141414141414, | |
| "grad_norm": 2.0794308185577393, | |
| "learning_rate": 2.8187919463087248e-05, | |
| "loss": 0.3618, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1447811447811448, | |
| "grad_norm": 1.5477614402770996, | |
| "learning_rate": 2.885906040268457e-05, | |
| "loss": 0.3593, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 10.740438461303711, | |
| "learning_rate": 2.9530201342281884e-05, | |
| "loss": 0.3805, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 2.993213176727295, | |
| "learning_rate": 3.02013422818792e-05, | |
| "loss": 0.3673, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15488215488215487, | |
| "grad_norm": 17.512208938598633, | |
| "learning_rate": 3.087248322147651e-05, | |
| "loss": 0.3922, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15824915824915825, | |
| "grad_norm": 2.5222012996673584, | |
| "learning_rate": 3.1543624161073825e-05, | |
| "loss": 0.3873, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 0.8730729222297668, | |
| "learning_rate": 3.221476510067114e-05, | |
| "loss": 0.3593, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.16498316498316498, | |
| "grad_norm": 0.8050268292427063, | |
| "learning_rate": 3.288590604026846e-05, | |
| "loss": 0.3491, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.16835016835016836, | |
| "grad_norm": 0.7536938190460205, | |
| "learning_rate": 3.3557046979865775e-05, | |
| "loss": 0.3469, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1717171717171717, | |
| "grad_norm": 0.9090268015861511, | |
| "learning_rate": 3.422818791946309e-05, | |
| "loss": 0.3663, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1750841750841751, | |
| "grad_norm": 0.8775368928909302, | |
| "learning_rate": 3.4899328859060405e-05, | |
| "loss": 0.3489, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.17845117845117844, | |
| "grad_norm": 0.5326427221298218, | |
| "learning_rate": 3.557046979865772e-05, | |
| "loss": 0.3466, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.561137318611145, | |
| "learning_rate": 3.6241610738255034e-05, | |
| "loss": 0.3393, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 0.8053128123283386, | |
| "learning_rate": 3.6912751677852356e-05, | |
| "loss": 0.352, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18855218855218855, | |
| "grad_norm": 0.5964087843894958, | |
| "learning_rate": 3.758389261744967e-05, | |
| "loss": 0.3507, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1919191919191919, | |
| "grad_norm": 0.5998376607894897, | |
| "learning_rate": 3.8255033557046985e-05, | |
| "loss": 0.3504, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19528619528619529, | |
| "grad_norm": 1.2634875774383545, | |
| "learning_rate": 3.89261744966443e-05, | |
| "loss": 0.337, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.19865319865319866, | |
| "grad_norm": 0.5703901648521423, | |
| "learning_rate": 3.959731543624161e-05, | |
| "loss": 0.3408, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 0.7656762003898621, | |
| "learning_rate": 4.026845637583892e-05, | |
| "loss": 0.3206, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2053872053872054, | |
| "grad_norm": 0.6210582852363586, | |
| "learning_rate": 4.0939597315436244e-05, | |
| "loss": 0.354, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.20875420875420875, | |
| "grad_norm": 0.6622840166091919, | |
| "learning_rate": 4.161073825503356e-05, | |
| "loss": 0.3439, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 0.46426376700401306, | |
| "learning_rate": 4.228187919463087e-05, | |
| "loss": 0.3434, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.21548821548821548, | |
| "grad_norm": 0.38662126660346985, | |
| "learning_rate": 4.295302013422819e-05, | |
| "loss": 0.3362, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.21885521885521886, | |
| "grad_norm": 0.5812459588050842, | |
| "learning_rate": 4.36241610738255e-05, | |
| "loss": 0.323, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.626932680606842, | |
| "learning_rate": 4.4295302013422824e-05, | |
| "loss": 0.3427, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2255892255892256, | |
| "grad_norm": 0.5491658449172974, | |
| "learning_rate": 4.496644295302014e-05, | |
| "loss": 0.3406, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.22895622895622897, | |
| "grad_norm": 0.4023520052433014, | |
| "learning_rate": 4.5637583892617453e-05, | |
| "loss": 0.3328, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.23232323232323232, | |
| "grad_norm": 0.478535532951355, | |
| "learning_rate": 4.630872483221477e-05, | |
| "loss": 0.3402, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2356902356902357, | |
| "grad_norm": 0.44869011640548706, | |
| "learning_rate": 4.697986577181208e-05, | |
| "loss": 0.3516, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23905723905723905, | |
| "grad_norm": 0.4810108244419098, | |
| "learning_rate": 4.76510067114094e-05, | |
| "loss": 0.3411, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 0.3956281542778015, | |
| "learning_rate": 4.832214765100672e-05, | |
| "loss": 0.3395, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.24579124579124578, | |
| "grad_norm": 0.40301939845085144, | |
| "learning_rate": 4.8993288590604034e-05, | |
| "loss": 0.3217, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.24915824915824916, | |
| "grad_norm": 0.44550034403800964, | |
| "learning_rate": 4.966442953020135e-05, | |
| "loss": 0.3257, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.25252525252525254, | |
| "grad_norm": 0.5890341997146606, | |
| "learning_rate": 5.033557046979866e-05, | |
| "loss": 0.3335, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2558922558922559, | |
| "grad_norm": 0.8096022009849548, | |
| "learning_rate": 5.100671140939598e-05, | |
| "loss": 0.3421, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.25925925925925924, | |
| "grad_norm": 0.6044747829437256, | |
| "learning_rate": 5.167785234899329e-05, | |
| "loss": 0.3266, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.26262626262626265, | |
| "grad_norm": 0.5191451907157898, | |
| "learning_rate": 5.234899328859061e-05, | |
| "loss": 0.331, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.265993265993266, | |
| "grad_norm": 1.0799261331558228, | |
| "learning_rate": 5.302013422818792e-05, | |
| "loss": 0.3243, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.26936026936026936, | |
| "grad_norm": 5.513405799865723, | |
| "learning_rate": 5.3691275167785237e-05, | |
| "loss": 0.379, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.673650860786438, | |
| "learning_rate": 5.436241610738255e-05, | |
| "loss": 0.3482, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2760942760942761, | |
| "grad_norm": 1.1485897302627563, | |
| "learning_rate": 5.5033557046979866e-05, | |
| "loss": 0.3351, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.27946127946127947, | |
| "grad_norm": 0.5018780827522278, | |
| "learning_rate": 5.570469798657718e-05, | |
| "loss": 0.3077, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 4.367802619934082, | |
| "learning_rate": 5.6375838926174495e-05, | |
| "loss": 0.3284, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.28619528619528617, | |
| "grad_norm": 33.46516036987305, | |
| "learning_rate": 5.704697986577181e-05, | |
| "loss": 1.0651, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2895622895622896, | |
| "grad_norm": 91.36512756347656, | |
| "learning_rate": 5.771812080536914e-05, | |
| "loss": 1.7174, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.29292929292929293, | |
| "grad_norm": 9.666085243225098, | |
| "learning_rate": 5.838926174496645e-05, | |
| "loss": 0.5601, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 8.608613967895508, | |
| "learning_rate": 5.906040268456377e-05, | |
| "loss": 0.3865, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2996632996632997, | |
| "grad_norm": 3.025059223175049, | |
| "learning_rate": 5.973154362416108e-05, | |
| "loss": 0.358, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 9.862916946411133, | |
| "learning_rate": 6.04026845637584e-05, | |
| "loss": 0.4464, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3063973063973064, | |
| "grad_norm": 11.05635929107666, | |
| "learning_rate": 6.107382550335571e-05, | |
| "loss": 0.3977, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.30976430976430974, | |
| "grad_norm": 1.0226973295211792, | |
| "learning_rate": 6.174496644295302e-05, | |
| "loss": 0.3206, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.31313131313131315, | |
| "grad_norm": 1.007895827293396, | |
| "learning_rate": 6.241610738255034e-05, | |
| "loss": 0.3355, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3164983164983165, | |
| "grad_norm": 1.5956454277038574, | |
| "learning_rate": 6.308724832214765e-05, | |
| "loss": 0.3408, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.31986531986531985, | |
| "grad_norm": 21.75948715209961, | |
| "learning_rate": 6.375838926174497e-05, | |
| "loss": 0.4627, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 5.754608154296875, | |
| "learning_rate": 6.442953020134228e-05, | |
| "loss": 0.3818, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3265993265993266, | |
| "grad_norm": 3.1888318061828613, | |
| "learning_rate": 6.51006711409396e-05, | |
| "loss": 0.3713, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.32996632996632996, | |
| "grad_norm": 4.586446762084961, | |
| "learning_rate": 6.577181208053692e-05, | |
| "loss": 0.3394, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.9332061409950256, | |
| "learning_rate": 6.644295302013423e-05, | |
| "loss": 0.3267, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "grad_norm": 4.119638442993164, | |
| "learning_rate": 6.711409395973155e-05, | |
| "loss": 0.3825, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "eval_loss": 0.16432031989097595, | |
| "eval_runtime": 33.0116, | |
| "eval_samples_per_second": 30.292, | |
| "eval_steps_per_second": 1.908, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3400673400673401, | |
| "grad_norm": 1.244138240814209, | |
| "learning_rate": 6.778523489932886e-05, | |
| "loss": 0.3191, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3434343434343434, | |
| "grad_norm": 4.564449310302734, | |
| "learning_rate": 6.845637583892618e-05, | |
| "loss": 0.3792, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3468013468013468, | |
| "grad_norm": 71.92516326904297, | |
| "learning_rate": 6.912751677852349e-05, | |
| "loss": 0.9526, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3501683501683502, | |
| "grad_norm": 6.8141374588012695, | |
| "learning_rate": 6.979865771812081e-05, | |
| "loss": 0.4132, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.35353535353535354, | |
| "grad_norm": 4.9158616065979, | |
| "learning_rate": 7.046979865771812e-05, | |
| "loss": 0.3847, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3569023569023569, | |
| "grad_norm": 0.9838681221008301, | |
| "learning_rate": 7.114093959731544e-05, | |
| "loss": 0.335, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3602693602693603, | |
| "grad_norm": 0.44024720788002014, | |
| "learning_rate": 7.181208053691275e-05, | |
| "loss": 0.3159, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.5798377394676208, | |
| "learning_rate": 7.248322147651007e-05, | |
| "loss": 0.3308, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.367003367003367, | |
| "grad_norm": 0.5650081038475037, | |
| "learning_rate": 7.315436241610739e-05, | |
| "loss": 0.3161, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 0.5149471163749695, | |
| "learning_rate": 7.382550335570471e-05, | |
| "loss": 0.3291, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.37373737373737376, | |
| "grad_norm": 0.4448802173137665, | |
| "learning_rate": 7.449664429530202e-05, | |
| "loss": 0.3145, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3771043771043771, | |
| "grad_norm": 0.5278413891792297, | |
| "learning_rate": 7.516778523489934e-05, | |
| "loss": 0.3296, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.38047138047138046, | |
| "grad_norm": 0.455289363861084, | |
| "learning_rate": 7.583892617449665e-05, | |
| "loss": 0.318, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3838383838383838, | |
| "grad_norm": 0.5316647291183472, | |
| "learning_rate": 7.651006711409397e-05, | |
| "loss": 0.3117, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3872053872053872, | |
| "grad_norm": 0.43862929940223694, | |
| "learning_rate": 7.718120805369128e-05, | |
| "loss": 0.3062, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.39057239057239057, | |
| "grad_norm": 12.535127639770508, | |
| "learning_rate": 7.78523489932886e-05, | |
| "loss": 0.3557, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 15.351152420043945, | |
| "learning_rate": 7.852348993288591e-05, | |
| "loss": 0.5181, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.39730639730639733, | |
| "grad_norm": 11.918878555297852, | |
| "learning_rate": 7.919463087248322e-05, | |
| "loss": 0.4005, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4006734006734007, | |
| "grad_norm": 9.800668716430664, | |
| "learning_rate": 7.986577181208054e-05, | |
| "loss": 0.422, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 16.235355377197266, | |
| "learning_rate": 8.053691275167784e-05, | |
| "loss": 0.4051, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4074074074074074, | |
| "grad_norm": 1.8551958799362183, | |
| "learning_rate": 8.120805369127518e-05, | |
| "loss": 0.3506, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4107744107744108, | |
| "grad_norm": 3.990302562713623, | |
| "learning_rate": 8.187919463087249e-05, | |
| "loss": 0.3318, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.41414141414141414, | |
| "grad_norm": 22.28190040588379, | |
| "learning_rate": 8.255033557046981e-05, | |
| "loss": 0.4316, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4175084175084175, | |
| "grad_norm": 1.9532949924468994, | |
| "learning_rate": 8.322147651006712e-05, | |
| "loss": 0.3596, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4208754208754209, | |
| "grad_norm": 0.8453232645988464, | |
| "learning_rate": 8.389261744966444e-05, | |
| "loss": 0.3473, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 3.7085459232330322, | |
| "learning_rate": 8.456375838926175e-05, | |
| "loss": 0.3527, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4276094276094276, | |
| "grad_norm": 1.9306743144989014, | |
| "learning_rate": 8.523489932885907e-05, | |
| "loss": 0.3415, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.43097643097643096, | |
| "grad_norm": 5.023862361907959, | |
| "learning_rate": 8.590604026845638e-05, | |
| "loss": 0.3644, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.43434343434343436, | |
| "grad_norm": 4.241243362426758, | |
| "learning_rate": 8.65771812080537e-05, | |
| "loss": 0.4321, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4377104377104377, | |
| "grad_norm": 1.7396281957626343, | |
| "learning_rate": 8.7248322147651e-05, | |
| "loss": 0.3334, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.44107744107744107, | |
| "grad_norm": 8.367612838745117, | |
| "learning_rate": 8.791946308724833e-05, | |
| "loss": 0.3571, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 7.692532539367676, | |
| "learning_rate": 8.859060402684565e-05, | |
| "loss": 0.5196, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4478114478114478, | |
| "grad_norm": 12.191128730773926, | |
| "learning_rate": 8.926174496644296e-05, | |
| "loss": 0.6991, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4511784511784512, | |
| "grad_norm": 7.570639133453369, | |
| "learning_rate": 8.993288590604028e-05, | |
| "loss": 0.4818, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.7189193964004517, | |
| "learning_rate": 9.060402684563759e-05, | |
| "loss": 0.3728, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.45791245791245794, | |
| "grad_norm": 9.100985527038574, | |
| "learning_rate": 9.127516778523491e-05, | |
| "loss": 0.3869, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4612794612794613, | |
| "grad_norm": 9.76489543914795, | |
| "learning_rate": 9.194630872483221e-05, | |
| "loss": 0.3861, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.46464646464646464, | |
| "grad_norm": 3.834136962890625, | |
| "learning_rate": 9.261744966442954e-05, | |
| "loss": 0.4222, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.468013468013468, | |
| "grad_norm": 22.2440242767334, | |
| "learning_rate": 9.328859060402684e-05, | |
| "loss": 0.7935, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4713804713804714, | |
| "grad_norm": 1.4633365869522095, | |
| "learning_rate": 9.395973154362417e-05, | |
| "loss": 0.347, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.47474747474747475, | |
| "grad_norm": 1.9224159717559814, | |
| "learning_rate": 9.463087248322147e-05, | |
| "loss": 0.3299, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4781144781144781, | |
| "grad_norm": 3.4107277393341064, | |
| "learning_rate": 9.53020134228188e-05, | |
| "loss": 0.4199, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.48148148148148145, | |
| "grad_norm": 1.4255735874176025, | |
| "learning_rate": 9.59731543624161e-05, | |
| "loss": 0.3559, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 1.4576934576034546, | |
| "learning_rate": 9.664429530201344e-05, | |
| "loss": 0.3274, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4882154882154882, | |
| "grad_norm": 1.3531242609024048, | |
| "learning_rate": 9.731543624161075e-05, | |
| "loss": 0.3417, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.49158249158249157, | |
| "grad_norm": 13.97393798828125, | |
| "learning_rate": 9.798657718120807e-05, | |
| "loss": 0.3434, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.494949494949495, | |
| "grad_norm": 0.8413457870483398, | |
| "learning_rate": 9.865771812080538e-05, | |
| "loss": 0.3224, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4983164983164983, | |
| "grad_norm": 0.41903650760650635, | |
| "learning_rate": 9.93288590604027e-05, | |
| "loss": 0.3197, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5016835016835017, | |
| "grad_norm": 1.3428220748901367, | |
| "learning_rate": 0.0001, | |
| "loss": 0.3184, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5050505050505051, | |
| "grad_norm": 0.497494637966156, | |
| "learning_rate": 9.9999861762256e-05, | |
| "loss": 0.3064, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5084175084175084, | |
| "grad_norm": 0.5110116600990295, | |
| "learning_rate": 9.999944704978836e-05, | |
| "loss": 0.3195, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5117845117845118, | |
| "grad_norm": 0.4883813261985779, | |
| "learning_rate": 9.999875586489024e-05, | |
| "loss": 0.292, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 0.44456565380096436, | |
| "learning_rate": 9.999778821138357e-05, | |
| "loss": 0.3084, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 0.5006658434867859, | |
| "learning_rate": 9.999654409461896e-05, | |
| "loss": 0.3031, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5218855218855218, | |
| "grad_norm": 0.4398713707923889, | |
| "learning_rate": 9.999502352147583e-05, | |
| "loss": 0.3178, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 0.4853643476963043, | |
| "learning_rate": 9.999322650036214e-05, | |
| "loss": 0.3195, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5286195286195287, | |
| "grad_norm": 0.4636339545249939, | |
| "learning_rate": 9.999115304121457e-05, | |
| "loss": 0.3052, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.531986531986532, | |
| "grad_norm": 0.525205671787262, | |
| "learning_rate": 9.998880315549834e-05, | |
| "loss": 0.3133, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5353535353535354, | |
| "grad_norm": 0.40854206681251526, | |
| "learning_rate": 9.998617685620714e-05, | |
| "loss": 0.3076, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5387205387205387, | |
| "grad_norm": 0.5355719327926636, | |
| "learning_rate": 9.998327415786315e-05, | |
| "loss": 0.3052, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5420875420875421, | |
| "grad_norm": 0.3861645460128784, | |
| "learning_rate": 9.998009507651684e-05, | |
| "loss": 0.3099, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.5338487029075623, | |
| "learning_rate": 9.997663962974697e-05, | |
| "loss": 0.3052, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5488215488215489, | |
| "grad_norm": 0.45219364762306213, | |
| "learning_rate": 9.997290783666049e-05, | |
| "loss": 0.2948, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5521885521885522, | |
| "grad_norm": 0.5037462711334229, | |
| "learning_rate": 9.996889971789235e-05, | |
| "loss": 0.3019, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.3949816823005676, | |
| "learning_rate": 9.996461529560553e-05, | |
| "loss": 0.3028, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5589225589225589, | |
| "grad_norm": 0.3921789824962616, | |
| "learning_rate": 9.996005459349074e-05, | |
| "loss": 0.2982, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5622895622895623, | |
| "grad_norm": 0.4122919738292694, | |
| "learning_rate": 9.995521763676645e-05, | |
| "loss": 0.3071, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 0.4212525188922882, | |
| "learning_rate": 9.995010445217867e-05, | |
| "loss": 0.3086, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.569023569023569, | |
| "grad_norm": 0.3997049033641815, | |
| "learning_rate": 9.994471506800079e-05, | |
| "loss": 0.2957, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5723905723905723, | |
| "grad_norm": 0.34380048513412476, | |
| "learning_rate": 9.993904951403344e-05, | |
| "loss": 0.3122, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 0.40532243251800537, | |
| "learning_rate": 9.99331078216044e-05, | |
| "loss": 0.3055, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5791245791245792, | |
| "grad_norm": 0.4095707833766937, | |
| "learning_rate": 9.992689002356828e-05, | |
| "loss": 0.2868, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5824915824915825, | |
| "grad_norm": 0.41159185767173767, | |
| "learning_rate": 9.992039615430648e-05, | |
| "loss": 0.318, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5858585858585859, | |
| "grad_norm": 0.3728049397468567, | |
| "learning_rate": 9.991362624972688e-05, | |
| "loss": 0.309, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5892255892255892, | |
| "grad_norm": 0.3249180018901825, | |
| "learning_rate": 9.990658034726379e-05, | |
| "loss": 0.2818, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.35090282559394836, | |
| "learning_rate": 9.989925848587756e-05, | |
| "loss": 0.2839, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5959595959595959, | |
| "grad_norm": 0.3364333212375641, | |
| "learning_rate": 9.989166070605447e-05, | |
| "loss": 0.3063, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5993265993265994, | |
| "grad_norm": 0.4135960340499878, | |
| "learning_rate": 9.988378704980656e-05, | |
| "loss": 0.3085, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6026936026936027, | |
| "grad_norm": 0.35615649819374084, | |
| "learning_rate": 9.987563756067129e-05, | |
| "loss": 0.2955, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 0.3038477897644043, | |
| "learning_rate": 9.986721228371129e-05, | |
| "loss": 0.291, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6094276094276094, | |
| "grad_norm": 0.4663616120815277, | |
| "learning_rate": 9.985851126551428e-05, | |
| "loss": 0.3043, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6127946127946128, | |
| "grad_norm": 0.42187029123306274, | |
| "learning_rate": 9.984953455419258e-05, | |
| "loss": 0.2747, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6161616161616161, | |
| "grad_norm": 0.4150826334953308, | |
| "learning_rate": 9.9840282199383e-05, | |
| "loss": 0.2854, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6195286195286195, | |
| "grad_norm": 0.36844050884246826, | |
| "learning_rate": 9.983075425224653e-05, | |
| "loss": 0.2848, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.622895622895623, | |
| "grad_norm": 0.4171907603740692, | |
| "learning_rate": 9.982095076546807e-05, | |
| "loss": 0.3003, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6262626262626263, | |
| "grad_norm": 0.3620002269744873, | |
| "learning_rate": 9.981087179325608e-05, | |
| "loss": 0.3043, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6296296296296297, | |
| "grad_norm": 0.40760573744773865, | |
| "learning_rate": 9.980051739134233e-05, | |
| "loss": 0.3059, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.632996632996633, | |
| "grad_norm": 0.32069963216781616, | |
| "learning_rate": 9.978988761698161e-05, | |
| "loss": 0.2947, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 0.3327488303184509, | |
| "learning_rate": 9.977898252895134e-05, | |
| "loss": 0.2805, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6397306397306397, | |
| "grad_norm": 0.3968160152435303, | |
| "learning_rate": 9.976780218755131e-05, | |
| "loss": 0.2891, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6430976430976431, | |
| "grad_norm": 0.4018626809120178, | |
| "learning_rate": 9.975634665460332e-05, | |
| "loss": 0.2965, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 0.37805649638175964, | |
| "learning_rate": 9.974461599345088e-05, | |
| "loss": 0.3008, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6498316498316499, | |
| "grad_norm": 0.44425806403160095, | |
| "learning_rate": 9.973261026895877e-05, | |
| "loss": 0.2921, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6531986531986532, | |
| "grad_norm": 0.375931054353714, | |
| "learning_rate": 9.972032954751279e-05, | |
| "loss": 0.296, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6565656565656566, | |
| "grad_norm": 0.44635701179504395, | |
| "learning_rate": 9.970777389701926e-05, | |
| "loss": 0.29, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6599326599326599, | |
| "grad_norm": 0.28897619247436523, | |
| "learning_rate": 9.969494338690481e-05, | |
| "loss": 0.2895, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6632996632996633, | |
| "grad_norm": 0.4542882740497589, | |
| "learning_rate": 9.968183808811586e-05, | |
| "loss": 0.2887, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.3715568780899048, | |
| "learning_rate": 9.966845807311829e-05, | |
| "loss": 0.3038, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.67003367003367, | |
| "grad_norm": 0.36940261721611023, | |
| "learning_rate": 9.965480341589701e-05, | |
| "loss": 0.2934, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "grad_norm": 0.43046656250953674, | |
| "learning_rate": 9.96408741919556e-05, | |
| "loss": 0.2951, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "eval_loss": 0.1412619948387146, | |
| "eval_runtime": 32.6268, | |
| "eval_samples_per_second": 30.65, | |
| "eval_steps_per_second": 1.931, | |
| "step": 200 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 1485, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.6733116367097037e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |