| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 297, | |
| "global_step": 297, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.003367003367003367, | |
| "grad_norm": 0.8125, | |
| "learning_rate": 1e-05, | |
| "loss": 2.5758, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.006734006734006734, | |
| "grad_norm": 0.765625, | |
| "learning_rate": 9.966329966329968e-06, | |
| "loss": 2.5078, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.010101010101010102, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 9.932659932659933e-06, | |
| "loss": 2.5049, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.013468013468013467, | |
| "grad_norm": 0.734375, | |
| "learning_rate": 9.8989898989899e-06, | |
| "loss": 2.5518, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.016835016835016835, | |
| "grad_norm": 0.66796875, | |
| "learning_rate": 9.865319865319866e-06, | |
| "loss": 2.5427, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.020202020202020204, | |
| "grad_norm": 0.56640625, | |
| "learning_rate": 9.831649831649833e-06, | |
| "loss": 2.3623, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.02356902356902357, | |
| "grad_norm": 0.55859375, | |
| "learning_rate": 9.797979797979798e-06, | |
| "loss": 2.4237, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.026936026936026935, | |
| "grad_norm": 0.51953125, | |
| "learning_rate": 9.764309764309765e-06, | |
| "loss": 2.4176, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 0.5078125, | |
| "learning_rate": 9.730639730639732e-06, | |
| "loss": 2.3253, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.03367003367003367, | |
| "grad_norm": 0.478515625, | |
| "learning_rate": 9.696969696969698e-06, | |
| "loss": 2.3064, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.037037037037037035, | |
| "grad_norm": 0.482421875, | |
| "learning_rate": 9.663299663299665e-06, | |
| "loss": 2.2647, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 0.474609375, | |
| "learning_rate": 9.62962962962963e-06, | |
| "loss": 2.2657, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.04377104377104377, | |
| "grad_norm": 0.453125, | |
| "learning_rate": 9.595959595959597e-06, | |
| "loss": 2.2032, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.04713804713804714, | |
| "grad_norm": 0.46484375, | |
| "learning_rate": 9.562289562289562e-06, | |
| "loss": 2.1713, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.050505050505050504, | |
| "grad_norm": 0.484375, | |
| "learning_rate": 9.52861952861953e-06, | |
| "loss": 2.2073, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.05387205387205387, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 9.494949494949497e-06, | |
| "loss": 2.1354, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.05723905723905724, | |
| "grad_norm": 0.47265625, | |
| "learning_rate": 9.461279461279462e-06, | |
| "loss": 2.1304, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 0.466796875, | |
| "learning_rate": 9.427609427609429e-06, | |
| "loss": 2.0846, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.06397306397306397, | |
| "grad_norm": 0.451171875, | |
| "learning_rate": 9.393939393939396e-06, | |
| "loss": 2.0422, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.06734006734006734, | |
| "grad_norm": 0.44140625, | |
| "learning_rate": 9.360269360269361e-06, | |
| "loss": 2.0434, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0707070707070707, | |
| "grad_norm": 0.416015625, | |
| "learning_rate": 9.326599326599327e-06, | |
| "loss": 2.0235, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.07407407407407407, | |
| "grad_norm": 0.4296875, | |
| "learning_rate": 9.292929292929294e-06, | |
| "loss": 1.9988, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.07744107744107744, | |
| "grad_norm": 0.388671875, | |
| "learning_rate": 9.25925925925926e-06, | |
| "loss": 1.8975, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 0.3984375, | |
| "learning_rate": 9.225589225589226e-06, | |
| "loss": 1.9702, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.08417508417508418, | |
| "grad_norm": 0.396484375, | |
| "learning_rate": 9.191919191919193e-06, | |
| "loss": 1.9406, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.08754208754208755, | |
| "grad_norm": 0.392578125, | |
| "learning_rate": 9.15824915824916e-06, | |
| "loss": 1.8887, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 0.380859375, | |
| "learning_rate": 9.124579124579126e-06, | |
| "loss": 1.8769, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.09427609427609428, | |
| "grad_norm": 0.376953125, | |
| "learning_rate": 9.090909090909091e-06, | |
| "loss": 1.8423, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.09764309764309764, | |
| "grad_norm": 0.37890625, | |
| "learning_rate": 9.057239057239058e-06, | |
| "loss": 1.8612, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.10101010101010101, | |
| "grad_norm": 0.36328125, | |
| "learning_rate": 9.023569023569025e-06, | |
| "loss": 1.8174, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.10437710437710437, | |
| "grad_norm": 0.341796875, | |
| "learning_rate": 8.98989898989899e-06, | |
| "loss": 1.8132, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.10774410774410774, | |
| "grad_norm": 0.322265625, | |
| "learning_rate": 8.956228956228958e-06, | |
| "loss": 1.7676, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 8.922558922558923e-06, | |
| "loss": 1.7846, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.11447811447811448, | |
| "grad_norm": 0.306640625, | |
| "learning_rate": 8.888888888888888e-06, | |
| "loss": 1.7733, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.11784511784511785, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 8.855218855218855e-06, | |
| "loss": 1.739, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 0.3046875, | |
| "learning_rate": 8.821548821548822e-06, | |
| "loss": 1.725, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.12457912457912458, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 8.787878787878788e-06, | |
| "loss": 1.6985, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.12794612794612795, | |
| "grad_norm": 0.30078125, | |
| "learning_rate": 8.754208754208755e-06, | |
| "loss": 1.6951, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.13131313131313133, | |
| "grad_norm": 0.328125, | |
| "learning_rate": 8.720538720538722e-06, | |
| "loss": 1.6596, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.13468013468013468, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 8.686868686868687e-06, | |
| "loss": 1.6469, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.13804713804713806, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 8.653198653198653e-06, | |
| "loss": 1.6273, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.1414141414141414, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.61952861952862e-06, | |
| "loss": 1.6243, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.1447811447811448, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 8.585858585858587e-06, | |
| "loss": 1.5959, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.14814814814814814, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.552188552188552e-06, | |
| "loss": 1.6236, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 8.518518518518519e-06, | |
| "loss": 1.6213, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.15488215488215487, | |
| "grad_norm": 0.2490234375, | |
| "learning_rate": 8.484848484848486e-06, | |
| "loss": 1.5851, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.15824915824915825, | |
| "grad_norm": 0.23828125, | |
| "learning_rate": 8.451178451178452e-06, | |
| "loss": 1.5851, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 0.267578125, | |
| "learning_rate": 8.417508417508419e-06, | |
| "loss": 1.5747, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.16498316498316498, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 8.383838383838384e-06, | |
| "loss": 1.5184, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.16835016835016836, | |
| "grad_norm": 0.314453125, | |
| "learning_rate": 8.350168350168351e-06, | |
| "loss": 1.5904, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.1717171717171717, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 8.316498316498316e-06, | |
| "loss": 1.5713, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.1750841750841751, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 8.282828282828283e-06, | |
| "loss": 1.5267, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.17845117845117844, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 8.24915824915825e-06, | |
| "loss": 1.5543, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 8.215488215488216e-06, | |
| "loss": 1.5648, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.18518518518518517, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 8.181818181818183e-06, | |
| "loss": 1.4896, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.18855218855218855, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 8.148148148148148e-06, | |
| "loss": 1.5295, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.1919191919191919, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 8.114478114478115e-06, | |
| "loss": 1.5177, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.19528619528619529, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 8.08080808080808e-06, | |
| "loss": 1.4765, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.19865319865319866, | |
| "grad_norm": 0.19921875, | |
| "learning_rate": 8.047138047138048e-06, | |
| "loss": 1.4731, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 8.013468013468015e-06, | |
| "loss": 1.5113, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.2053872053872054, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 7.97979797979798e-06, | |
| "loss": 1.4836, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.20875420875420875, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 7.946127946127947e-06, | |
| "loss": 1.4953, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 7.912457912457913e-06, | |
| "loss": 1.5195, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.21548821548821548, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 7.87878787878788e-06, | |
| "loss": 1.4541, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.21885521885521886, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 7.845117845117845e-06, | |
| "loss": 1.4226, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 7.811447811447812e-06, | |
| "loss": 1.4715, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.2255892255892256, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 7.77777777777778e-06, | |
| "loss": 1.4595, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.22895622895622897, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 7.744107744107745e-06, | |
| "loss": 1.4243, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.23232323232323232, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 7.710437710437712e-06, | |
| "loss": 1.4426, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.2356902356902357, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.676767676767677e-06, | |
| "loss": 1.4317, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.23905723905723905, | |
| "grad_norm": 0.2041015625, | |
| "learning_rate": 7.643097643097644e-06, | |
| "loss": 1.4313, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 7.60942760942761e-06, | |
| "loss": 1.479, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.24579124579124578, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 7.5757575757575764e-06, | |
| "loss": 1.4102, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.24915824915824916, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 7.542087542087543e-06, | |
| "loss": 1.4277, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.25252525252525254, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 7.508417508417509e-06, | |
| "loss": 1.4199, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.2558922558922559, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 7.474747474747476e-06, | |
| "loss": 1.4312, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.25925925925925924, | |
| "grad_norm": 0.2138671875, | |
| "learning_rate": 7.441077441077442e-06, | |
| "loss": 1.4414, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.26262626262626265, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.4074074074074075e-06, | |
| "loss": 1.4225, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.265993265993266, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 7.373737373737374e-06, | |
| "loss": 1.4135, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.26936026936026936, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 7.340067340067341e-06, | |
| "loss": 1.4481, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 7.306397306397307e-06, | |
| "loss": 1.3645, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.2760942760942761, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 7.272727272727273e-06, | |
| "loss": 1.4644, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.27946127946127947, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 7.23905723905724e-06, | |
| "loss": 1.3992, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 7.2053872053872064e-06, | |
| "loss": 1.4122, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.28619528619528617, | |
| "grad_norm": 0.201171875, | |
| "learning_rate": 7.171717171717172e-06, | |
| "loss": 1.3658, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.2895622895622896, | |
| "grad_norm": 0.2255859375, | |
| "learning_rate": 7.138047138047138e-06, | |
| "loss": 1.4178, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.29292929292929293, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 7.104377104377105e-06, | |
| "loss": 1.3861, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2962962962962963, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 7.070707070707071e-06, | |
| "loss": 1.3704, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.2996632996632997, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 7.0370370370370375e-06, | |
| "loss": 1.41, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 0.208984375, | |
| "learning_rate": 7.0033670033670045e-06, | |
| "loss": 1.373, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.3063973063973064, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 6.969696969696971e-06, | |
| "loss": 1.3598, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.30976430976430974, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 6.936026936026936e-06, | |
| "loss": 1.3426, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.31313131313131315, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.902356902356902e-06, | |
| "loss": 1.3619, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.3164983164983165, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 6.868686868686869e-06, | |
| "loss": 1.3631, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.31986531986531985, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 6.835016835016836e-06, | |
| "loss": 1.4074, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 6.801346801346802e-06, | |
| "loss": 1.3986, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.3265993265993266, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 6.767676767676769e-06, | |
| "loss": 1.3347, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.32996632996632996, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 6.734006734006735e-06, | |
| "loss": 1.3661, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.271484375, | |
| "learning_rate": 6.7003367003367004e-06, | |
| "loss": 1.3536, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.3367003367003367, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 6.666666666666667e-06, | |
| "loss": 1.3298, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.3400673400673401, | |
| "grad_norm": 0.173828125, | |
| "learning_rate": 6.632996632996634e-06, | |
| "loss": 1.3568, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.3434343434343434, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 6.5993265993266e-06, | |
| "loss": 1.3967, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.3468013468013468, | |
| "grad_norm": 0.345703125, | |
| "learning_rate": 6.565656565656566e-06, | |
| "loss": 1.3748, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.3501683501683502, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.531986531986533e-06, | |
| "loss": 1.3578, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.35353535353535354, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 6.498316498316499e-06, | |
| "loss": 1.3561, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.3569023569023569, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 6.464646464646466e-06, | |
| "loss": 1.335, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.3602693602693603, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 6.430976430976431e-06, | |
| "loss": 1.3539, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 0.32421875, | |
| "learning_rate": 6.397306397306397e-06, | |
| "loss": 1.3218, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.367003367003367, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.363636363636364e-06, | |
| "loss": 1.3466, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 6.3299663299663304e-06, | |
| "loss": 1.3496, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.37373737373737376, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 6.296296296296297e-06, | |
| "loss": 1.3861, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.3771043771043771, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 6.262626262626264e-06, | |
| "loss": 1.3067, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.38047138047138046, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 6.22895622895623e-06, | |
| "loss": 1.3356, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.3838383838383838, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 6.195286195286195e-06, | |
| "loss": 1.3661, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.3872053872053872, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 6.1616161616161615e-06, | |
| "loss": 1.3279, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.39057239057239057, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 6.1279461279461286e-06, | |
| "loss": 1.3434, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 6.094276094276095e-06, | |
| "loss": 1.3545, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.39730639730639733, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 6.060606060606061e-06, | |
| "loss": 1.3066, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.4006734006734007, | |
| "grad_norm": 0.197265625, | |
| "learning_rate": 6.026936026936028e-06, | |
| "loss": 1.3392, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 0.24609375, | |
| "learning_rate": 5.993265993265994e-06, | |
| "loss": 1.3769, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.4074074074074074, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 5.95959595959596e-06, | |
| "loss": 1.324, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.4107744107744108, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 5.925925925925926e-06, | |
| "loss": 1.316, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.41414141414141414, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 5.892255892255893e-06, | |
| "loss": 1.3214, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.4175084175084175, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 5.858585858585859e-06, | |
| "loss": 1.313, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.4208754208754209, | |
| "grad_norm": 0.2265625, | |
| "learning_rate": 5.824915824915825e-06, | |
| "loss": 1.2991, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 0.228515625, | |
| "learning_rate": 5.791245791245792e-06, | |
| "loss": 1.3314, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.4276094276094276, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 5.7575757575757586e-06, | |
| "loss": 1.3422, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.43097643097643096, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 5.723905723905724e-06, | |
| "loss": 1.3403, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.43434343434343436, | |
| "grad_norm": 0.1689453125, | |
| "learning_rate": 5.69023569023569e-06, | |
| "loss": 1.3299, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.4377104377104377, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 5.656565656565657e-06, | |
| "loss": 1.2948, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.44107744107744107, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 5.622895622895623e-06, | |
| "loss": 1.3008, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 5.58922558922559e-06, | |
| "loss": 1.3109, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.4478114478114478, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 5.555555555555557e-06, | |
| "loss": 1.3173, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.4511784511784512, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 5.521885521885523e-06, | |
| "loss": 1.3035, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 5.488215488215489e-06, | |
| "loss": 1.3143, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.45791245791245794, | |
| "grad_norm": 0.3125, | |
| "learning_rate": 5.4545454545454545e-06, | |
| "loss": 1.3273, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.4612794612794613, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 5.420875420875421e-06, | |
| "loss": 1.2928, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.46464646464646464, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 5.387205387205388e-06, | |
| "loss": 1.3109, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.468013468013468, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 5.353535353535354e-06, | |
| "loss": 1.3421, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.4713804713804714, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 5.31986531986532e-06, | |
| "loss": 1.3012, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.47474747474747475, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 5.286195286195287e-06, | |
| "loss": 1.2922, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.4781144781144781, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 5.252525252525253e-06, | |
| "loss": 1.2697, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.48148148148148145, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 5.218855218855219e-06, | |
| "loss": 1.3034, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 0.2412109375, | |
| "learning_rate": 5.185185185185185e-06, | |
| "loss": 1.3159, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.4882154882154882, | |
| "grad_norm": 0.2119140625, | |
| "learning_rate": 5.151515151515152e-06, | |
| "loss": 1.3265, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.49158249158249157, | |
| "grad_norm": 0.2158203125, | |
| "learning_rate": 5.117845117845118e-06, | |
| "loss": 1.3025, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.494949494949495, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 5.0841750841750845e-06, | |
| "loss": 1.292, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.4983164983164983, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 5.0505050505050515e-06, | |
| "loss": 1.2741, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.5016835016835017, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 5.016835016835018e-06, | |
| "loss": 1.2658, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.5050505050505051, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 4.983164983164984e-06, | |
| "loss": 1.2984, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.5084175084175084, | |
| "grad_norm": 0.28515625, | |
| "learning_rate": 4.94949494949495e-06, | |
| "loss": 1.29, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.5117845117845118, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 4.915824915824916e-06, | |
| "loss": 1.2808, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.5151515151515151, | |
| "grad_norm": 0.17578125, | |
| "learning_rate": 4.8821548821548826e-06, | |
| "loss": 1.2975, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.5185185185185185, | |
| "grad_norm": 0.248046875, | |
| "learning_rate": 4.848484848484849e-06, | |
| "loss": 1.2837, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.5218855218855218, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 4.814814814814815e-06, | |
| "loss": 1.2542, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.5252525252525253, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 4.781144781144781e-06, | |
| "loss": 1.2679, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.5286195286195287, | |
| "grad_norm": 0.169921875, | |
| "learning_rate": 4.747474747474748e-06, | |
| "loss": 1.2849, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.531986531986532, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 4.7138047138047145e-06, | |
| "loss": 1.2927, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.5353535353535354, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 4.680134680134681e-06, | |
| "loss": 1.2951, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.5387205387205387, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 4.646464646464647e-06, | |
| "loss": 1.2925, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.5420875420875421, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 4.612794612794613e-06, | |
| "loss": 1.2964, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.5454545454545454, | |
| "grad_norm": 0.232421875, | |
| "learning_rate": 4.57912457912458e-06, | |
| "loss": 1.3198, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.5488215488215489, | |
| "grad_norm": 0.265625, | |
| "learning_rate": 4.5454545454545455e-06, | |
| "loss": 1.2564, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.5521885521885522, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 4.5117845117845126e-06, | |
| "loss": 1.2536, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.5555555555555556, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 4.478114478114479e-06, | |
| "loss": 1.2581, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.5589225589225589, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 4.444444444444444e-06, | |
| "loss": 1.2864, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.5622895622895623, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 4.410774410774411e-06, | |
| "loss": 1.2741, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.5656565656565656, | |
| "grad_norm": 0.171875, | |
| "learning_rate": 4.377104377104377e-06, | |
| "loss": 1.2903, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.569023569023569, | |
| "grad_norm": 0.2099609375, | |
| "learning_rate": 4.343434343434344e-06, | |
| "loss": 1.2491, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.5723905723905723, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 4.30976430976431e-06, | |
| "loss": 1.2999, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.5757575757575758, | |
| "grad_norm": 0.2578125, | |
| "learning_rate": 4.276094276094276e-06, | |
| "loss": 1.252, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.5791245791245792, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 4.242424242424243e-06, | |
| "loss": 1.2698, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.5824915824915825, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 4.208754208754209e-06, | |
| "loss": 1.2976, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.5858585858585859, | |
| "grad_norm": 0.279296875, | |
| "learning_rate": 4.1750841750841755e-06, | |
| "loss": 1.2809, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.5892255892255892, | |
| "grad_norm": 0.2421875, | |
| "learning_rate": 4.141414141414142e-06, | |
| "loss": 1.2793, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.5925925925925926, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 4.107744107744108e-06, | |
| "loss": 1.283, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.5959595959595959, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 4.074074074074074e-06, | |
| "loss": 1.2703, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.5993265993265994, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 4.04040404040404e-06, | |
| "loss": 1.2958, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.6026936026936027, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 4.0067340067340074e-06, | |
| "loss": 1.2795, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.6060606060606061, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 3.973063973063974e-06, | |
| "loss": 1.2627, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.6094276094276094, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 3.93939393939394e-06, | |
| "loss": 1.2868, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.6127946127946128, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 3.905723905723906e-06, | |
| "loss": 1.2585, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.6161616161616161, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 3.872053872053872e-06, | |
| "loss": 1.2739, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.6195286195286195, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 3.8383838383838385e-06, | |
| "loss": 1.2952, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.622895622895623, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 3.804713804713805e-06, | |
| "loss": 1.2613, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.6262626262626263, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 3.7710437710437713e-06, | |
| "loss": 1.2619, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.6296296296296297, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 3.737373737373738e-06, | |
| "loss": 1.2612, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.632996632996633, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.7037037037037037e-06, | |
| "loss": 1.2619, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.6363636363636364, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 3.6700336700336704e-06, | |
| "loss": 1.2732, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.6397306397306397, | |
| "grad_norm": 0.30859375, | |
| "learning_rate": 3.6363636363636366e-06, | |
| "loss": 1.2528, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.6430976430976431, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 3.6026936026936032e-06, | |
| "loss": 1.2548, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.6464646464646465, | |
| "grad_norm": 0.2001953125, | |
| "learning_rate": 3.569023569023569e-06, | |
| "loss": 1.2408, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.6498316498316499, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 3.5353535353535356e-06, | |
| "loss": 1.2915, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.6531986531986532, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 3.5016835016835023e-06, | |
| "loss": 1.2444, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.6565656565656566, | |
| "grad_norm": 0.205078125, | |
| "learning_rate": 3.468013468013468e-06, | |
| "loss": 1.2512, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.6599326599326599, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 3.4343434343434347e-06, | |
| "loss": 1.2214, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.6632996632996633, | |
| "grad_norm": 0.251953125, | |
| "learning_rate": 3.400673400673401e-06, | |
| "loss": 1.2741, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.259765625, | |
| "learning_rate": 3.3670033670033675e-06, | |
| "loss": 1.2408, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 0.67003367003367, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 1.2188, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 0.6734006734006734, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.2996632996633e-06, | |
| "loss": 1.2488, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.6767676767676768, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 3.2659932659932666e-06, | |
| "loss": 1.2758, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 0.6801346801346801, | |
| "grad_norm": 0.29296875, | |
| "learning_rate": 3.232323232323233e-06, | |
| "loss": 1.2792, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 0.6835016835016835, | |
| "grad_norm": 0.27734375, | |
| "learning_rate": 3.1986531986531986e-06, | |
| "loss": 1.2542, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 0.6868686868686869, | |
| "grad_norm": 0.2216796875, | |
| "learning_rate": 3.1649831649831652e-06, | |
| "loss": 1.2586, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 0.6902356902356902, | |
| "grad_norm": 0.3671875, | |
| "learning_rate": 3.131313131313132e-06, | |
| "loss": 1.268, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.6936026936026936, | |
| "grad_norm": 0.2021484375, | |
| "learning_rate": 3.0976430976430976e-06, | |
| "loss": 1.2389, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 0.696969696969697, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 3.0639730639730643e-06, | |
| "loss": 1.2394, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 0.7003367003367004, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 3.0303030303030305e-06, | |
| "loss": 1.2688, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 0.7037037037037037, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 2.996632996632997e-06, | |
| "loss": 1.2535, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 0.7070707070707071, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 2.962962962962963e-06, | |
| "loss": 1.2406, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.7104377104377104, | |
| "grad_norm": 0.1953125, | |
| "learning_rate": 2.9292929292929295e-06, | |
| "loss": 1.2489, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 0.7138047138047138, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 2.895622895622896e-06, | |
| "loss": 1.2694, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 0.7171717171717171, | |
| "grad_norm": 0.1806640625, | |
| "learning_rate": 2.861952861952862e-06, | |
| "loss": 1.2729, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 0.7205387205387206, | |
| "grad_norm": 0.236328125, | |
| "learning_rate": 2.8282828282828286e-06, | |
| "loss": 1.2425, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 0.7239057239057239, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 2.794612794612795e-06, | |
| "loss": 1.2807, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.7272727272727273, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 2.7609427609427614e-06, | |
| "loss": 1.255, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 0.7306397306397306, | |
| "grad_norm": 0.212890625, | |
| "learning_rate": 2.7272727272727272e-06, | |
| "loss": 1.2519, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 0.734006734006734, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 2.693602693602694e-06, | |
| "loss": 1.2452, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 0.7373737373737373, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 2.65993265993266e-06, | |
| "loss": 1.2769, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 2.6262626262626267e-06, | |
| "loss": 1.2567, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.7441077441077442, | |
| "grad_norm": 0.185546875, | |
| "learning_rate": 2.5925925925925925e-06, | |
| "loss": 1.2641, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 0.7474747474747475, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 2.558922558922559e-06, | |
| "loss": 1.2451, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 0.7508417508417509, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 2.5252525252525258e-06, | |
| "loss": 1.2237, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 0.7542087542087542, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 2.491582491582492e-06, | |
| "loss": 1.2465, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 0.7575757575757576, | |
| "grad_norm": 0.369140625, | |
| "learning_rate": 2.457912457912458e-06, | |
| "loss": 1.3071, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.7609427609427609, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 2.4242424242424244e-06, | |
| "loss": 1.2614, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 0.7643097643097643, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 2.3905723905723906e-06, | |
| "loss": 1.2847, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 0.7676767676767676, | |
| "grad_norm": 0.22265625, | |
| "learning_rate": 2.3569023569023572e-06, | |
| "loss": 1.2603, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 0.7710437710437711, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 2.3232323232323234e-06, | |
| "loss": 1.2442, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 0.7744107744107744, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 2.28956228956229e-06, | |
| "loss": 1.268, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.7777777777777778, | |
| "grad_norm": 0.1962890625, | |
| "learning_rate": 2.2558922558922563e-06, | |
| "loss": 1.238, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 0.7811447811447811, | |
| "grad_norm": 0.2431640625, | |
| "learning_rate": 2.222222222222222e-06, | |
| "loss": 1.2514, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 0.7845117845117845, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 2.1885521885521887e-06, | |
| "loss": 1.2194, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 0.7878787878787878, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 2.154882154882155e-06, | |
| "loss": 1.244, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 0.7912457912457912, | |
| "grad_norm": 0.203125, | |
| "learning_rate": 2.1212121212121216e-06, | |
| "loss": 1.225, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.7946127946127947, | |
| "grad_norm": 0.294921875, | |
| "learning_rate": 2.0875420875420878e-06, | |
| "loss": 1.2393, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 0.797979797979798, | |
| "grad_norm": 0.18359375, | |
| "learning_rate": 2.053872053872054e-06, | |
| "loss": 1.2488, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 0.8013468013468014, | |
| "grad_norm": 0.1845703125, | |
| "learning_rate": 2.02020202020202e-06, | |
| "loss": 1.2427, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 0.8047138047138047, | |
| "grad_norm": 0.23046875, | |
| "learning_rate": 1.986531986531987e-06, | |
| "loss": 1.2546, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 0.8080808080808081, | |
| "grad_norm": 0.1982421875, | |
| "learning_rate": 1.952861952861953e-06, | |
| "loss": 1.2521, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.8114478114478114, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 1.9191919191919192e-06, | |
| "loss": 1.2965, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 0.8148148148148148, | |
| "grad_norm": 0.26171875, | |
| "learning_rate": 1.8855218855218857e-06, | |
| "loss": 1.2844, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 0.8181818181818182, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 1.2445, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 0.8215488215488216, | |
| "grad_norm": 0.1728515625, | |
| "learning_rate": 1.8181818181818183e-06, | |
| "loss": 1.244, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 0.8249158249158249, | |
| "grad_norm": 0.224609375, | |
| "learning_rate": 1.7845117845117845e-06, | |
| "loss": 1.2586, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.8282828282828283, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 1.7508417508417511e-06, | |
| "loss": 1.2333, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 0.8316498316498316, | |
| "grad_norm": 0.2080078125, | |
| "learning_rate": 1.7171717171717173e-06, | |
| "loss": 1.2431, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 0.835016835016835, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 1.6835016835016838e-06, | |
| "loss": 1.2606, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 0.8383838383838383, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 1.64983164983165e-06, | |
| "loss": 1.2771, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 0.8417508417508418, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 1.6161616161616164e-06, | |
| "loss": 1.2611, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.8451178451178452, | |
| "grad_norm": 0.2060546875, | |
| "learning_rate": 1.5824915824915826e-06, | |
| "loss": 1.2421, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 0.8484848484848485, | |
| "grad_norm": 0.255859375, | |
| "learning_rate": 1.5488215488215488e-06, | |
| "loss": 1.2568, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 0.8518518518518519, | |
| "grad_norm": 0.2734375, | |
| "learning_rate": 1.5151515151515152e-06, | |
| "loss": 1.2583, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 0.8552188552188552, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 1.4814814814814815e-06, | |
| "loss": 1.2582, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 0.8585858585858586, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 1.447811447811448e-06, | |
| "loss": 1.2667, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.8619528619528619, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.4141414141414143e-06, | |
| "loss": 1.2687, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 0.8653198653198653, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.3804713804713807e-06, | |
| "loss": 1.2761, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 0.8686868686868687, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 1.346801346801347e-06, | |
| "loss": 1.2392, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 0.8720538720538721, | |
| "grad_norm": 0.2294921875, | |
| "learning_rate": 1.3131313131313134e-06, | |
| "loss": 1.2687, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 0.8754208754208754, | |
| "grad_norm": 0.2470703125, | |
| "learning_rate": 1.2794612794612796e-06, | |
| "loss": 1.2291, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.8787878787878788, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 1.245791245791246e-06, | |
| "loss": 1.239, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 0.8821548821548821, | |
| "grad_norm": 0.1748046875, | |
| "learning_rate": 1.2121212121212122e-06, | |
| "loss": 1.2599, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 0.8855218855218855, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.1784511784511786e-06, | |
| "loss": 1.2912, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 0.8888888888888888, | |
| "grad_norm": 0.216796875, | |
| "learning_rate": 1.144781144781145e-06, | |
| "loss": 1.2382, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 0.8922558922558923, | |
| "grad_norm": 0.2236328125, | |
| "learning_rate": 1.111111111111111e-06, | |
| "loss": 1.2655, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.8956228956228957, | |
| "grad_norm": 0.1943359375, | |
| "learning_rate": 1.0774410774410775e-06, | |
| "loss": 1.2769, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 0.898989898989899, | |
| "grad_norm": 0.1923828125, | |
| "learning_rate": 1.0437710437710439e-06, | |
| "loss": 1.2753, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 0.9023569023569024, | |
| "grad_norm": 0.1865234375, | |
| "learning_rate": 1.01010101010101e-06, | |
| "loss": 1.2405, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 0.9057239057239057, | |
| "grad_norm": 0.2451171875, | |
| "learning_rate": 9.764309764309765e-07, | |
| "loss": 1.2443, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.1796875, | |
| "learning_rate": 9.427609427609428e-07, | |
| "loss": 1.2461, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.9124579124579124, | |
| "grad_norm": 0.1787109375, | |
| "learning_rate": 9.090909090909091e-07, | |
| "loss": 1.2383, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 0.9158249158249159, | |
| "grad_norm": 0.1904296875, | |
| "learning_rate": 8.754208754208756e-07, | |
| "loss": 1.2546, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 0.9191919191919192, | |
| "grad_norm": 0.37109375, | |
| "learning_rate": 8.417508417508419e-07, | |
| "loss": 1.3184, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 0.9225589225589226, | |
| "grad_norm": 0.193359375, | |
| "learning_rate": 8.080808080808082e-07, | |
| "loss": 1.2659, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 0.9259259259259259, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 7.744107744107744e-07, | |
| "loss": 1.2878, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.9292929292929293, | |
| "grad_norm": 0.337890625, | |
| "learning_rate": 7.407407407407407e-07, | |
| "loss": 1.282, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 0.9326599326599326, | |
| "grad_norm": 0.19140625, | |
| "learning_rate": 7.070707070707071e-07, | |
| "loss": 1.2493, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 0.936026936026936, | |
| "grad_norm": 0.177734375, | |
| "learning_rate": 6.734006734006735e-07, | |
| "loss": 1.2419, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 0.9393939393939394, | |
| "grad_norm": 0.1826171875, | |
| "learning_rate": 6.397306397306398e-07, | |
| "loss": 1.2396, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 0.9427609427609428, | |
| "grad_norm": 0.189453125, | |
| "learning_rate": 6.060606060606061e-07, | |
| "loss": 1.2458, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.9461279461279462, | |
| "grad_norm": 0.33984375, | |
| "learning_rate": 5.723905723905725e-07, | |
| "loss": 1.2557, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 0.9494949494949495, | |
| "grad_norm": 0.1767578125, | |
| "learning_rate": 5.387205387205387e-07, | |
| "loss": 1.2546, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 0.9528619528619529, | |
| "grad_norm": 0.2275390625, | |
| "learning_rate": 5.05050505050505e-07, | |
| "loss": 1.2439, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 0.9562289562289562, | |
| "grad_norm": 0.1884765625, | |
| "learning_rate": 4.713804713804714e-07, | |
| "loss": 1.2948, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 0.9595959595959596, | |
| "grad_norm": 0.21875, | |
| "learning_rate": 4.377104377104378e-07, | |
| "loss": 1.215, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.9629629629629629, | |
| "grad_norm": 0.2109375, | |
| "learning_rate": 4.040404040404041e-07, | |
| "loss": 1.2779, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 0.9663299663299664, | |
| "grad_norm": 0.2353515625, | |
| "learning_rate": 3.7037037037037036e-07, | |
| "loss": 1.2564, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 0.9696969696969697, | |
| "grad_norm": 0.21484375, | |
| "learning_rate": 3.3670033670033673e-07, | |
| "loss": 1.2712, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 0.9730639730639731, | |
| "grad_norm": 0.25390625, | |
| "learning_rate": 3.0303030303030305e-07, | |
| "loss": 1.2408, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 0.9764309764309764, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 2.6936026936026936e-07, | |
| "loss": 1.2375, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.9797979797979798, | |
| "grad_norm": 0.353515625, | |
| "learning_rate": 2.356902356902357e-07, | |
| "loss": 1.2509, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 0.9831649831649831, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 2.0202020202020205e-07, | |
| "loss": 1.2597, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 0.9865319865319865, | |
| "grad_norm": 0.181640625, | |
| "learning_rate": 1.6835016835016837e-07, | |
| "loss": 1.2576, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 0.98989898989899, | |
| "grad_norm": 0.298828125, | |
| "learning_rate": 1.3468013468013468e-07, | |
| "loss": 1.2312, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 0.9932659932659933, | |
| "grad_norm": 0.20703125, | |
| "learning_rate": 1.0101010101010103e-07, | |
| "loss": 1.2391, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.9966329966329966, | |
| "grad_norm": 0.2177734375, | |
| "learning_rate": 6.734006734006734e-08, | |
| "loss": 1.2403, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.1875, | |
| "learning_rate": 3.367003367003367e-08, | |
| "loss": 1.2241, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.2526977062225342, | |
| "eval_runtime": 10.0634, | |
| "eval_samples_per_second": 2.782, | |
| "eval_steps_per_second": 0.397, | |
| "step": 297 | |
| } | |
| ], | |
| "logging_steps": 1.0, | |
| "max_steps": 297, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 0, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 9.429499898169917e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |