| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.49747474747474746, | |
| "eval_steps": 98, | |
| "global_step": 197, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0025252525252525255, | |
| "grad_norm": 17.0, | |
| "learning_rate": 1.0000000000000002e-06, | |
| "loss": 1.9033, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005050505050505051, | |
| "grad_norm": 19.5, | |
| "learning_rate": 2.0000000000000003e-06, | |
| "loss": 2.1028, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.007575757575757576, | |
| "grad_norm": 17.125, | |
| "learning_rate": 3e-06, | |
| "loss": 2.1133, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.010101010101010102, | |
| "grad_norm": 15.8125, | |
| "learning_rate": 4.000000000000001e-06, | |
| "loss": 2.0188, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.012626262626262626, | |
| "grad_norm": 14.375, | |
| "learning_rate": 5e-06, | |
| "loss": 1.9426, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.015151515151515152, | |
| "grad_norm": 11.0, | |
| "learning_rate": 6e-06, | |
| "loss": 1.7974, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.017676767676767676, | |
| "grad_norm": 9.1875, | |
| "learning_rate": 7e-06, | |
| "loss": 1.7561, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.020202020202020204, | |
| "grad_norm": 8.4375, | |
| "learning_rate": 8.000000000000001e-06, | |
| "loss": 1.7865, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.022727272727272728, | |
| "grad_norm": 9.5, | |
| "learning_rate": 9e-06, | |
| "loss": 1.7445, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.025252525252525252, | |
| "grad_norm": 8.6875, | |
| "learning_rate": 1e-05, | |
| "loss": 1.6558, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.027777777777777776, | |
| "grad_norm": 6.625, | |
| "learning_rate": 1.1000000000000001e-05, | |
| "loss": 1.2742, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.030303030303030304, | |
| "grad_norm": 6.28125, | |
| "learning_rate": 1.2e-05, | |
| "loss": 1.2157, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.03282828282828283, | |
| "grad_norm": 6.0, | |
| "learning_rate": 1.3000000000000001e-05, | |
| "loss": 1.24, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.03535353535353535, | |
| "grad_norm": 5.15625, | |
| "learning_rate": 1.4e-05, | |
| "loss": 1.1103, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.03787878787878788, | |
| "grad_norm": 4.0, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.9098, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.04040404040404041, | |
| "grad_norm": 3.90625, | |
| "learning_rate": 1.6000000000000003e-05, | |
| "loss": 1.005, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.04292929292929293, | |
| "grad_norm": 3.203125, | |
| "learning_rate": 1.7e-05, | |
| "loss": 0.7924, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.045454545454545456, | |
| "grad_norm": 3.078125, | |
| "learning_rate": 1.8e-05, | |
| "loss": 0.8178, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.047979797979797977, | |
| "grad_norm": 2.46875, | |
| "learning_rate": 1.9e-05, | |
| "loss": 0.7241, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.050505050505050504, | |
| "grad_norm": 2.90625, | |
| "learning_rate": 2e-05, | |
| "loss": 0.9651, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.05303030303030303, | |
| "grad_norm": 2.53125, | |
| "learning_rate": 1.99468085106383e-05, | |
| "loss": 0.7474, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.05555555555555555, | |
| "grad_norm": 2.296875, | |
| "learning_rate": 1.9893617021276595e-05, | |
| "loss": 0.7503, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.05808080808080808, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 1.9840425531914894e-05, | |
| "loss": 0.7411, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.06060606060606061, | |
| "grad_norm": 2.265625, | |
| "learning_rate": 1.9787234042553193e-05, | |
| "loss": 0.7489, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.06313131313131314, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.973404255319149e-05, | |
| "loss": 0.7545, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.06565656565656566, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.968085106382979e-05, | |
| "loss": 0.6775, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.06818181818181818, | |
| "grad_norm": 2.25, | |
| "learning_rate": 1.962765957446809e-05, | |
| "loss": 0.7888, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.0707070707070707, | |
| "grad_norm": 2.21875, | |
| "learning_rate": 1.9574468085106384e-05, | |
| "loss": 0.8352, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.07323232323232323, | |
| "grad_norm": 2.125, | |
| "learning_rate": 1.9521276595744682e-05, | |
| "loss": 0.6698, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.07575757575757576, | |
| "grad_norm": 2.5625, | |
| "learning_rate": 1.946808510638298e-05, | |
| "loss": 0.8309, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.07828282828282829, | |
| "grad_norm": 2.234375, | |
| "learning_rate": 1.9414893617021276e-05, | |
| "loss": 0.6307, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.08080808080808081, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 1.9361702127659575e-05, | |
| "loss": 0.6557, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.08333333333333333, | |
| "grad_norm": 2.328125, | |
| "learning_rate": 1.9308510638297873e-05, | |
| "loss": 0.6325, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.08585858585858586, | |
| "grad_norm": 2.609375, | |
| "learning_rate": 1.9255319148936172e-05, | |
| "loss": 0.7254, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.08838383838383838, | |
| "grad_norm": 1.9453125, | |
| "learning_rate": 1.920212765957447e-05, | |
| "loss": 0.6736, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.09090909090909091, | |
| "grad_norm": 1.7265625, | |
| "learning_rate": 1.914893617021277e-05, | |
| "loss": 0.74, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.09343434343434344, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 1.9095744680851064e-05, | |
| "loss": 0.6725, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.09595959595959595, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.9042553191489363e-05, | |
| "loss": 0.631, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.09848484848484848, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.898936170212766e-05, | |
| "loss": 0.6208, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.10101010101010101, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.893617021276596e-05, | |
| "loss": 0.5535, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.10353535353535354, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.888297872340426e-05, | |
| "loss": 0.6425, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.10606060606060606, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.8829787234042557e-05, | |
| "loss": 0.5682, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.10858585858585859, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.8776595744680852e-05, | |
| "loss": 0.6053, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.1111111111111111, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.872340425531915e-05, | |
| "loss": 0.5894, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.11363636363636363, | |
| "grad_norm": 1.6640625, | |
| "learning_rate": 1.8670212765957446e-05, | |
| "loss": 0.523, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.11616161616161616, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.8617021276595745e-05, | |
| "loss": 0.5638, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.11868686868686869, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.8563829787234043e-05, | |
| "loss": 0.533, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.12121212121212122, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.8510638297872342e-05, | |
| "loss": 0.5829, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.12373737373737374, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 1.845744680851064e-05, | |
| "loss": 0.6252, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.12626262626262627, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.840425531914894e-05, | |
| "loss": 0.641, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.12878787878787878, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.8351063829787234e-05, | |
| "loss": 0.4742, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.13131313131313133, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.8297872340425533e-05, | |
| "loss": 0.5409, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.13383838383838384, | |
| "grad_norm": 1.640625, | |
| "learning_rate": 1.824468085106383e-05, | |
| "loss": 0.6769, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.13636363636363635, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 1.8191489361702127e-05, | |
| "loss": 0.5134, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.1388888888888889, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.8138297872340425e-05, | |
| "loss": 0.6641, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.1414141414141414, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.8085106382978724e-05, | |
| "loss": 0.6039, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.14393939393939395, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.8031914893617023e-05, | |
| "loss": 0.5456, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.14646464646464646, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.797872340425532e-05, | |
| "loss": 0.6185, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.14898989898989898, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.792553191489362e-05, | |
| "loss": 0.6234, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.15151515151515152, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.7872340425531915e-05, | |
| "loss": 0.6293, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.15404040404040403, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.7819148936170214e-05, | |
| "loss": 0.5785, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.15656565656565657, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.7765957446808512e-05, | |
| "loss": 0.5021, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.1590909090909091, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.7712765957446807e-05, | |
| "loss": 0.6317, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.16161616161616163, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 1.765957446808511e-05, | |
| "loss": 0.56, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.16414141414141414, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 1.7606382978723408e-05, | |
| "loss": 0.5755, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 1.6171875, | |
| "learning_rate": 1.7553191489361703e-05, | |
| "loss": 0.6197, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.1691919191919192, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.7500000000000002e-05, | |
| "loss": 0.5798, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.1717171717171717, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.74468085106383e-05, | |
| "loss": 0.5906, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.17424242424242425, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.7393617021276596e-05, | |
| "loss": 0.7807, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.17676767676767677, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.7340425531914894e-05, | |
| "loss": 0.626, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.17929292929292928, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.7287234042553193e-05, | |
| "loss": 0.6278, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.18181818181818182, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.723404255319149e-05, | |
| "loss": 0.5956, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.18434343434343434, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.718085106382979e-05, | |
| "loss": 0.5488, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.18686868686868688, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 1.712765957446809e-05, | |
| "loss": 0.5977, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.1893939393939394, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.7074468085106384e-05, | |
| "loss": 0.5672, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.1919191919191919, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.7021276595744682e-05, | |
| "loss": 0.5855, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.19444444444444445, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.696808510638298e-05, | |
| "loss": 0.5899, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.19696969696969696, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.6914893617021276e-05, | |
| "loss": 0.5164, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.1994949494949495, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.6861702127659575e-05, | |
| "loss": 0.7067, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 0.20202020202020202, | |
| "grad_norm": 1.828125, | |
| "learning_rate": 1.6808510638297873e-05, | |
| "loss": 0.7673, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.20454545454545456, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.6755319148936172e-05, | |
| "loss": 0.5957, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 0.20707070707070707, | |
| "grad_norm": 1.59375, | |
| "learning_rate": 1.670212765957447e-05, | |
| "loss": 0.593, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.20959595959595959, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.664893617021277e-05, | |
| "loss": 0.6161, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 0.21212121212121213, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.6595744680851064e-05, | |
| "loss": 0.5788, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.21464646464646464, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.6542553191489363e-05, | |
| "loss": 0.5945, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.21717171717171718, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 1.648936170212766e-05, | |
| "loss": 0.5936, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.2196969696969697, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.6436170212765957e-05, | |
| "loss": 0.5393, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 0.2222222222222222, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.6382978723404255e-05, | |
| "loss": 0.5319, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.22474747474747475, | |
| "grad_norm": 1.734375, | |
| "learning_rate": 1.6329787234042554e-05, | |
| "loss": 0.7014, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 0.22727272727272727, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.6276595744680853e-05, | |
| "loss": 0.6423, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.2297979797979798, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.622340425531915e-05, | |
| "loss": 0.7358, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 0.23232323232323232, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.6170212765957446e-05, | |
| "loss": 0.4773, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.23484848484848486, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.6117021276595745e-05, | |
| "loss": 0.5955, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 0.23737373737373738, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 1.6063829787234044e-05, | |
| "loss": 0.6716, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.2398989898989899, | |
| "grad_norm": 1.6328125, | |
| "learning_rate": 1.6010638297872342e-05, | |
| "loss": 0.6673, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.24242424242424243, | |
| "grad_norm": 1.90625, | |
| "learning_rate": 1.595744680851064e-05, | |
| "loss": 0.7431, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.24494949494949494, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.590425531914894e-05, | |
| "loss": 0.5161, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 0.2474747474747475, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.5851063829787235e-05, | |
| "loss": 0.5427, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.2474747474747475, | |
| "eval_loss": 0.4802989363670349, | |
| "eval_model_preparation_time": 0.0001, | |
| "eval_runtime": 4.4682, | |
| "eval_samples_per_second": 23.723, | |
| "eval_steps_per_second": 3.133, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.5797872340425533e-05, | |
| "loss": 0.5947, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 0.25252525252525254, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.5744680851063832e-05, | |
| "loss": 0.6192, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.255050505050505, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 1.5691489361702127e-05, | |
| "loss": 0.568, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 0.25757575757575757, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 1.5638297872340426e-05, | |
| "loss": 0.6521, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 0.2601010101010101, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.5585106382978724e-05, | |
| "loss": 0.65, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 0.26262626262626265, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.5531914893617023e-05, | |
| "loss": 0.4976, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 0.26515151515151514, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.547872340425532e-05, | |
| "loss": 0.5354, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.2676767676767677, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 1.542553191489362e-05, | |
| "loss": 0.5232, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 0.2702020202020202, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.5372340425531915e-05, | |
| "loss": 0.4453, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 0.2727272727272727, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.5319148936170214e-05, | |
| "loss": 0.6146, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 0.27525252525252525, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.5265957446808512e-05, | |
| "loss": 0.4385, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 0.2777777777777778, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.521276595744681e-05, | |
| "loss": 0.5709, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.2803030303030303, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.5159574468085108e-05, | |
| "loss": 0.5547, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 0.2828282828282828, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.5106382978723407e-05, | |
| "loss": 0.5413, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 0.28535353535353536, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 1.5053191489361702e-05, | |
| "loss": 0.5048, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 0.2878787878787879, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.5000000000000002e-05, | |
| "loss": 0.6009, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 0.2904040404040404, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.49468085106383e-05, | |
| "loss": 0.472, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.29292929292929293, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 1.4893617021276596e-05, | |
| "loss": 0.6086, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 0.29545454545454547, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.4840425531914894e-05, | |
| "loss": 0.5837, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 0.29797979797979796, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 1.4787234042553193e-05, | |
| "loss": 0.5546, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 0.3005050505050505, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.473404255319149e-05, | |
| "loss": 0.4623, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 0.30303030303030304, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.4680851063829789e-05, | |
| "loss": 0.492, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.3055555555555556, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.4627659574468087e-05, | |
| "loss": 0.5731, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 0.30808080808080807, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.4574468085106384e-05, | |
| "loss": 0.5626, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 0.3106060606060606, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.4521276595744683e-05, | |
| "loss": 0.475, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 0.31313131313131315, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.4468085106382981e-05, | |
| "loss": 0.552, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 0.31565656565656564, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 1.4414893617021276e-05, | |
| "loss": 0.525, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.3181818181818182, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.4361702127659575e-05, | |
| "loss": 0.5764, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 0.3207070707070707, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.4308510638297874e-05, | |
| "loss": 0.5816, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 0.32323232323232326, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.425531914893617e-05, | |
| "loss": 0.5908, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 0.32575757575757575, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.420212765957447e-05, | |
| "loss": 0.5943, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 0.3282828282828283, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.4148936170212768e-05, | |
| "loss": 0.4884, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.33080808080808083, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 1.4095744680851065e-05, | |
| "loss": 0.5799, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.4042553191489363e-05, | |
| "loss": 0.5285, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 0.33585858585858586, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.3989361702127662e-05, | |
| "loss": 0.4816, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 0.3383838383838384, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.3936170212765959e-05, | |
| "loss": 0.6294, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 0.3409090909090909, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.3882978723404257e-05, | |
| "loss": 0.4614, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.3434343434343434, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.3829787234042556e-05, | |
| "loss": 0.5877, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 0.34595959595959597, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 1.3776595744680851e-05, | |
| "loss": 0.4327, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 0.3484848484848485, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.372340425531915e-05, | |
| "loss": 0.5889, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 0.351010101010101, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.3670212765957447e-05, | |
| "loss": 0.5389, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 0.35353535353535354, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.3617021276595745e-05, | |
| "loss": 0.4373, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.3560606060606061, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.3563829787234044e-05, | |
| "loss": 0.5761, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 0.35858585858585856, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.351063829787234e-05, | |
| "loss": 0.7717, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 0.3611111111111111, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.345744680851064e-05, | |
| "loss": 0.4974, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 0.36363636363636365, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.3404255319148938e-05, | |
| "loss": 0.5615, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 0.3661616161616162, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.3351063829787235e-05, | |
| "loss": 0.5081, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.3686868686868687, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 1.3297872340425533e-05, | |
| "loss": 0.486, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 0.3712121212121212, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.3244680851063832e-05, | |
| "loss": 0.4847, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 0.37373737373737376, | |
| "grad_norm": 1.53125, | |
| "learning_rate": 1.3191489361702127e-05, | |
| "loss": 0.5955, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 0.37626262626262624, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.3138297872340426e-05, | |
| "loss": 0.46, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 0.3787878787878788, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 1.3085106382978724e-05, | |
| "loss": 0.5995, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.3813131313131313, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 1.3031914893617021e-05, | |
| "loss": 0.468, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 0.3838383838383838, | |
| "grad_norm": 1.984375, | |
| "learning_rate": 1.297872340425532e-05, | |
| "loss": 0.5832, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 0.38636363636363635, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.2925531914893619e-05, | |
| "loss": 0.6264, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 0.3888888888888889, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.2872340425531915e-05, | |
| "loss": 0.5987, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 0.39141414141414144, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.2819148936170214e-05, | |
| "loss": 0.5271, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.3939393939393939, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 1.2765957446808513e-05, | |
| "loss": 0.4722, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 0.39646464646464646, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 1.2712765957446808e-05, | |
| "loss": 0.6128, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 0.398989898989899, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 1.2659574468085108e-05, | |
| "loss": 0.7296, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 0.4015151515151515, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.2606382978723407e-05, | |
| "loss": 0.658, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 0.40404040404040403, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.2553191489361702e-05, | |
| "loss": 0.5477, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.4065656565656566, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 1.25e-05, | |
| "loss": 0.5191, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 0.4090909090909091, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.24468085106383e-05, | |
| "loss": 0.6534, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 0.4116161616161616, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.2393617021276596e-05, | |
| "loss": 0.5652, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 0.41414141414141414, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.2340425531914895e-05, | |
| "loss": 0.7201, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 0.4166666666666667, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 1.2287234042553193e-05, | |
| "loss": 0.6026, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.41919191919191917, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.223404255319149e-05, | |
| "loss": 0.5286, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 0.4217171717171717, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.2180851063829789e-05, | |
| "loss": 0.5045, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 0.42424242424242425, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.2127659574468087e-05, | |
| "loss": 0.5062, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 0.42676767676767674, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.2074468085106383e-05, | |
| "loss": 0.5936, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 0.4292929292929293, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.2021276595744681e-05, | |
| "loss": 0.4753, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.4318181818181818, | |
| "grad_norm": 1.5859375, | |
| "learning_rate": 1.196808510638298e-05, | |
| "loss": 0.5516, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 0.43434343434343436, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 1.1914893617021277e-05, | |
| "loss": 0.571, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 0.43686868686868685, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 1.1861702127659575e-05, | |
| "loss": 0.5234, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 0.4393939393939394, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 1.1808510638297874e-05, | |
| "loss": 0.5271, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 0.44191919191919193, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.175531914893617e-05, | |
| "loss": 0.6933, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.4444444444444444, | |
| "grad_norm": 1.5, | |
| "learning_rate": 1.170212765957447e-05, | |
| "loss": 0.5311, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 0.44696969696969696, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.1648936170212768e-05, | |
| "loss": 0.4916, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 0.4494949494949495, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.1595744680851065e-05, | |
| "loss": 0.6005, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 0.45202020202020204, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.1542553191489364e-05, | |
| "loss": 0.4982, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 1.1489361702127662e-05, | |
| "loss": 0.5284, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.45707070707070707, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.1436170212765957e-05, | |
| "loss": 0.4811, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 0.4595959595959596, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 1.1382978723404256e-05, | |
| "loss": 0.4331, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 0.4621212121212121, | |
| "grad_norm": 1.609375, | |
| "learning_rate": 1.1329787234042555e-05, | |
| "loss": 0.5808, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 0.46464646464646464, | |
| "grad_norm": 1.71875, | |
| "learning_rate": 1.1276595744680851e-05, | |
| "loss": 0.7746, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 0.4671717171717172, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.122340425531915e-05, | |
| "loss": 0.5737, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.4696969696969697, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 1.1170212765957447e-05, | |
| "loss": 0.5221, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 0.4722222222222222, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.1117021276595746e-05, | |
| "loss": 0.6547, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 0.47474747474747475, | |
| "grad_norm": 1.375, | |
| "learning_rate": 1.1063829787234044e-05, | |
| "loss": 0.5264, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 0.4772727272727273, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 1.1010638297872341e-05, | |
| "loss": 0.5393, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 0.4797979797979798, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.095744680851064e-05, | |
| "loss": 0.5669, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.4823232323232323, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 1.0904255319148938e-05, | |
| "loss": 0.5213, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 0.48484848484848486, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.0851063829787233e-05, | |
| "loss": 0.5955, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 0.48737373737373735, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.0797872340425532e-05, | |
| "loss": 0.5273, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 0.4898989898989899, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.074468085106383e-05, | |
| "loss": 0.4816, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 0.49242424242424243, | |
| "grad_norm": 2.0, | |
| "learning_rate": 1.0691489361702128e-05, | |
| "loss": 0.6121, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.494949494949495, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 1.0638297872340426e-05, | |
| "loss": 0.6325, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.494949494949495, | |
| "eval_loss": 0.45528531074523926, | |
| "eval_model_preparation_time": 0.0001, | |
| "eval_runtime": 4.4693, | |
| "eval_samples_per_second": 23.717, | |
| "eval_steps_per_second": 3.132, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 0.49747474747474746, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.0585106382978725e-05, | |
| "loss": 0.4857, | |
| "step": 197 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 396, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 197, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.091750925875282e+17, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |