| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.972307692307693, | |
| "eval_steps": 1000, | |
| "global_step": 505, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.009846153846153846, | |
| "grad_norm": 38.5, | |
| "learning_rate": 3.125e-06, | |
| "loss": 1.348, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.019692307692307693, | |
| "grad_norm": 28.625, | |
| "learning_rate": 6.25e-06, | |
| "loss": 1.3239, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.039384615384615386, | |
| "grad_norm": 18.375, | |
| "learning_rate": 1.25e-05, | |
| "loss": 1.3144, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.059076923076923075, | |
| "grad_norm": 14.3125, | |
| "learning_rate": 1.8750000000000002e-05, | |
| "loss": 1.4069, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.07876923076923077, | |
| "grad_norm": 11.5625, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.247, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.09846153846153846, | |
| "grad_norm": 11.25, | |
| "learning_rate": 3.125e-05, | |
| "loss": 1.2618, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.11815384615384615, | |
| "grad_norm": 10.25, | |
| "learning_rate": 3.7500000000000003e-05, | |
| "loss": 1.287, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.13784615384615384, | |
| "grad_norm": 5.9375, | |
| "learning_rate": 4.375e-05, | |
| "loss": 1.1393, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.15753846153846154, | |
| "grad_norm": 7.34375, | |
| "learning_rate": 5e-05, | |
| "loss": 1.1422, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.17723076923076922, | |
| "grad_norm": 7.84375, | |
| "learning_rate": 4.9997936302412985e-05, | |
| "loss": 1.0547, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.19692307692307692, | |
| "grad_norm": 7.0, | |
| "learning_rate": 4.9991745550359746e-05, | |
| "loss": 1.0486, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.21661538461538463, | |
| "grad_norm": 7.75, | |
| "learning_rate": 4.99814287659075e-05, | |
| "loss": 1.0206, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2363076923076923, | |
| "grad_norm": 13.0625, | |
| "learning_rate": 4.996698765231409e-05, | |
| "loss": 0.9878, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 4.59375, | |
| "learning_rate": 4.994842459374682e-05, | |
| "loss": 0.9275, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.2756923076923077, | |
| "grad_norm": 6.46875, | |
| "learning_rate": 4.992574265488883e-05, | |
| "loss": 0.9555, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.2953846153846154, | |
| "grad_norm": 3.890625, | |
| "learning_rate": 4.989894558043312e-05, | |
| "loss": 0.9275, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3150769230769231, | |
| "grad_norm": 4.75, | |
| "learning_rate": 4.986803779446432e-05, | |
| "loss": 0.9301, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.33476923076923076, | |
| "grad_norm": 2.71875, | |
| "learning_rate": 4.983302439972829e-05, | |
| "loss": 0.8875, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.35446153846153844, | |
| "grad_norm": 6.65625, | |
| "learning_rate": 4.979391117678969e-05, | |
| "loss": 0.8649, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.37415384615384617, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.975070458307763e-05, | |
| "loss": 0.8497, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.39384615384615385, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.970341175181956e-05, | |
| "loss": 0.8358, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.4135384615384615, | |
| "grad_norm": 4.1875, | |
| "learning_rate": 4.9652040490863624e-05, | |
| "loss": 0.8191, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.43323076923076925, | |
| "grad_norm": 4.75, | |
| "learning_rate": 4.95965992813896e-05, | |
| "loss": 0.8425, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.45292307692307693, | |
| "grad_norm": 2.796875, | |
| "learning_rate": 4.9537097276508704e-05, | |
| "loss": 0.8027, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.4726153846153846, | |
| "grad_norm": 2.453125, | |
| "learning_rate": 4.947354429975245e-05, | |
| "loss": 0.812, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.49230769230769234, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 4.940595084345082e-05, | |
| "loss": 0.7979, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 2.3125, | |
| "learning_rate": 4.933432806700004e-05, | |
| "loss": 0.7927, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.5316923076923077, | |
| "grad_norm": 2.953125, | |
| "learning_rate": 4.925868779502015e-05, | |
| "loss": 0.7773, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.5513846153846154, | |
| "grad_norm": 3.625, | |
| "learning_rate": 4.9179042515402926e-05, | |
| "loss": 0.7694, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.571076923076923, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.909540537725007e-05, | |
| "loss": 0.7703, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.5907692307692308, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 4.900779018870239e-05, | |
| "loss": 0.8162, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.6104615384615385, | |
| "grad_norm": 2.15625, | |
| "learning_rate": 4.891621141466014e-05, | |
| "loss": 0.743, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.6301538461538462, | |
| "grad_norm": 1.6875, | |
| "learning_rate": 4.882068417439493e-05, | |
| "loss": 0.7572, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.6498461538461539, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 4.872122423905358e-05, | |
| "loss": 0.7445, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.6695384615384615, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 4.8617848029054354e-05, | |
| "loss": 0.7419, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.6892307692307692, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 4.851057261137608e-05, | |
| "loss": 0.7402, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.7089230769230769, | |
| "grad_norm": 4.09375, | |
| "learning_rate": 4.839941569674041e-05, | |
| "loss": 0.7131, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.7286153846153847, | |
| "grad_norm": 4.25, | |
| "learning_rate": 4.8284395636687854e-05, | |
| "loss": 0.6954, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.7483076923076923, | |
| "grad_norm": 3.21875, | |
| "learning_rate": 4.816553142054805e-05, | |
| "loss": 0.699, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 4.125, | |
| "learning_rate": 4.804284267230468e-05, | |
| "loss": 0.6775, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 0.7876923076923077, | |
| "grad_norm": 2.984375, | |
| "learning_rate": 4.791634964735564e-05, | |
| "loss": 0.7056, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.8073846153846154, | |
| "grad_norm": 2.59375, | |
| "learning_rate": 4.778607322916896e-05, | |
| "loss": 0.6944, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 0.827076923076923, | |
| "grad_norm": 1.859375, | |
| "learning_rate": 4.765203492583502e-05, | |
| "loss": 0.668, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 0.8467692307692307, | |
| "grad_norm": 2.109375, | |
| "learning_rate": 4.751425686651568e-05, | |
| "loss": 0.673, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 0.8664615384615385, | |
| "grad_norm": 1.84375, | |
| "learning_rate": 4.737276179779083e-05, | |
| "loss": 0.7153, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 0.8861538461538462, | |
| "grad_norm": 2.125, | |
| "learning_rate": 4.722757307990302e-05, | |
| "loss": 0.7234, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.9058461538461539, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 4.707871468290078e-05, | |
| "loss": 0.6231, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 0.9255384615384615, | |
| "grad_norm": 1.6484375, | |
| "learning_rate": 4.69262111826813e-05, | |
| "loss": 0.642, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 0.9452307692307692, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 4.6770087756932995e-05, | |
| "loss": 0.6231, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 0.9649230769230769, | |
| "grad_norm": 1.78125, | |
| "learning_rate": 4.661037018097884e-05, | |
| "loss": 0.671, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 0.9846153846153847, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 4.6447084823520926e-05, | |
| "loss": 0.6657, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.0043076923076923, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 4.62802586422871e-05, | |
| "loss": 0.607, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 4.610991917958037e-05, | |
| "loss": 0.5739, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.0436923076923077, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 4.593609455773181e-05, | |
| "loss": 0.6011, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.0633846153846154, | |
| "grad_norm": 1.5234375, | |
| "learning_rate": 4.5758813474457606e-05, | |
| "loss": 0.5776, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.083076923076923, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 4.557810519812128e-05, | |
| "loss": 0.5808, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.1027692307692307, | |
| "grad_norm": 1.7890625, | |
| "learning_rate": 4.539399956290152e-05, | |
| "loss": 0.5965, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.1224615384615384, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 4.520652696386677e-05, | |
| "loss": 0.608, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.142153846153846, | |
| "grad_norm": 1.625, | |
| "learning_rate": 4.5015718351957015e-05, | |
| "loss": 0.5714, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.1618461538461538, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 4.482160522887403e-05, | |
| "loss": 0.5876, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.1815384615384614, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 4.462421964188052e-05, | |
| "loss": 0.5835, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.2012307692307693, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 4.442359417850924e-05, | |
| "loss": 0.5881, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.220923076923077, | |
| "grad_norm": 1.9140625, | |
| "learning_rate": 4.421976196118297e-05, | |
| "loss": 0.5471, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.2406153846153847, | |
| "grad_norm": 1.625, | |
| "learning_rate": 4.401275664174611e-05, | |
| "loss": 0.5417, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.2603076923076924, | |
| "grad_norm": 1.625, | |
| "learning_rate": 4.380261239590892e-05, | |
| "loss": 0.5337, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 4.358936391760524e-05, | |
| "loss": 0.5731, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.2996923076923077, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 4.337304641326467e-05, | |
| "loss": 0.5363, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.3193846153846154, | |
| "grad_norm": 1.515625, | |
| "learning_rate": 4.315369559600018e-05, | |
| "loss": 0.5566, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.339076923076923, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 4.2931347679711924e-05, | |
| "loss": 0.586, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.3587692307692307, | |
| "grad_norm": 1.6015625, | |
| "learning_rate": 4.270603937310859e-05, | |
| "loss": 0.5535, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.3784615384615384, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 4.2477807873646845e-05, | |
| "loss": 0.5788, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.398153846153846, | |
| "grad_norm": 1.484375, | |
| "learning_rate": 4.2246690861390294e-05, | |
| "loss": 0.538, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.417846153846154, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 4.201272649278856e-05, | |
| "loss": 0.5531, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.4375384615384617, | |
| "grad_norm": 1.5, | |
| "learning_rate": 4.177595339437789e-05, | |
| "loss": 0.55, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.4572307692307693, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 4.153641065640402e-05, | |
| "loss": 0.5333, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.476923076923077, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 4.129413782636859e-05, | |
| "loss": 0.5372, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.4966153846153847, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 4.1049174902499974e-05, | |
| "loss": 0.5575, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.5163076923076924, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 4.080156232714976e-05, | |
| "loss": 0.5571, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 4.055134098011589e-05, | |
| "loss": 0.5246, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 1.5556923076923077, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 4.0298552171893576e-05, | |
| "loss": 0.5597, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 1.5753846153846154, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 4.0043237636855116e-05, | |
| "loss": 0.5536, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.595076923076923, | |
| "grad_norm": 1.5078125, | |
| "learning_rate": 3.978543952635967e-05, | |
| "loss": 0.5527, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 1.6147692307692307, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 3.952520040179434e-05, | |
| "loss": 0.5137, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 1.6344615384615384, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 3.92625632275474e-05, | |
| "loss": 0.5795, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 1.654153846153846, | |
| "grad_norm": 1.46875, | |
| "learning_rate": 3.899757136391507e-05, | |
| "loss": 0.5237, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 1.6738461538461538, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 3.873026855994292e-05, | |
| "loss": 0.5326, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.6935384615384614, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 3.8460698946203054e-05, | |
| "loss": 0.5231, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 1.7132307692307691, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 3.818890702750841e-05, | |
| "loss": 0.5492, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 1.7329230769230768, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 3.791493767556511e-05, | |
| "loss": 0.6126, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 1.7526153846153845, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 3.7638836121564415e-05, | |
| "loss": 0.5463, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 1.7723076923076924, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 3.7360647948715164e-05, | |
| "loss": 0.515, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 3.708041908471827e-05, | |
| "loss": 0.5259, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 1.8116923076923077, | |
| "grad_norm": 1.578125, | |
| "learning_rate": 3.679819579418414e-05, | |
| "loss": 0.5059, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 1.8313846153846154, | |
| "grad_norm": 1.375, | |
| "learning_rate": 3.651402467099468e-05, | |
| "loss": 0.5709, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 1.851076923076923, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 3.622795263061079e-05, | |
| "loss": 0.5628, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 1.8707692307692307, | |
| "grad_norm": 1.4609375, | |
| "learning_rate": 3.594002690232682e-05, | |
| "loss": 0.5066, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.8904615384615384, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 3.565029502147323e-05, | |
| "loss": 0.5625, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 1.9101538461538463, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.53588048215687e-05, | |
| "loss": 0.5336, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 1.929846153846154, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 3.506560442642299e-05, | |
| "loss": 0.5215, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 1.9495384615384617, | |
| "grad_norm": 1.546875, | |
| "learning_rate": 3.4770742242191945e-05, | |
| "loss": 0.5296, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 1.9692307692307693, | |
| "grad_norm": 2.421875, | |
| "learning_rate": 3.4474266949385817e-05, | |
| "loss": 0.523, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.988923076923077, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 3.4176227494832305e-05, | |
| "loss": 0.4856, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.0086153846153847, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 3.387667308359568e-05, | |
| "loss": 0.5298, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.0283076923076924, | |
| "grad_norm": 1.375, | |
| "learning_rate": 3.3575653170853175e-05, | |
| "loss": 0.4869, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 1.0703125, | |
| "learning_rate": 3.327321745373021e-05, | |
| "loss": 0.479, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.0676923076923077, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 3.2969415863095556e-05, | |
| "loss": 0.4935, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.0873846153846154, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 3.266429855531797e-05, | |
| "loss": 0.4773, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.107076923076923, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 3.2357915903985605e-05, | |
| "loss": 0.4611, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.1267692307692307, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 3.2050318491589506e-05, | |
| "loss": 0.469, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.1464615384615384, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 3.174155710117271e-05, | |
| "loss": 0.4758, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.166153846153846, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 3.143168270794612e-05, | |
| "loss": 0.4933, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.1858461538461538, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 3.112074647087274e-05, | |
| "loss": 0.4814, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.2055384615384614, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 3.080879972422154e-05, | |
| "loss": 0.5064, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.225230769230769, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 3.0495893969092392e-05, | |
| "loss": 0.4576, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.244923076923077, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 3.0182080864913452e-05, | |
| "loss": 0.4902, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.2646153846153845, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 2.9867412220912373e-05, | |
| "loss": 0.4486, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.284307692307692, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 2.9551939987562866e-05, | |
| "loss": 0.4786, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 2.923571624800787e-05, | |
| "loss": 0.4814, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 2.3236923076923075, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 2.891879320946086e-05, | |
| "loss": 0.4915, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 2.3433846153846156, | |
| "grad_norm": 1.2265625, | |
| "learning_rate": 2.8601223194586612e-05, | |
| "loss": 0.4931, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 2.363076923076923, | |
| "grad_norm": 1.4453125, | |
| "learning_rate": 2.8283058632863003e-05, | |
| "loss": 0.481, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.382769230769231, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 2.7964352051925103e-05, | |
| "loss": 0.4458, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 2.4024615384615386, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 2.7645156068893073e-05, | |
| "loss": 0.499, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 2.4221538461538463, | |
| "grad_norm": 2.078125, | |
| "learning_rate": 2.732552338168531e-05, | |
| "loss": 0.4937, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 2.441846153846154, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 2.7005506760318235e-05, | |
| "loss": 0.4628, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 2.4615384615384617, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 2.66851590381942e-05, | |
| "loss": 0.4741, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.4812307692307694, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 2.6364533103378896e-05, | |
| "loss": 0.4569, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 2.500923076923077, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 2.604368188986977e-05, | |
| "loss": 0.4851, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 2.5206153846153847, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 2.5722658368856816e-05, | |
| "loss": 0.4935, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 2.5403076923076924, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 2.5401515539977305e-05, | |
| "loss": 0.4947, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 2.5080306422565707e-05, | |
| "loss": 0.4642, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.5796923076923077, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 2.4759084046900486e-05, | |
| "loss": 0.5064, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 2.5993846153846154, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 2.4437901445448936e-05, | |
| "loss": 0.4376, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 2.619076923076923, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 2.4116811644111852e-05, | |
| "loss": 0.4861, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 2.6387692307692308, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 2.379586765346907e-05, | |
| "loss": 0.4878, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 2.6584615384615384, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 2.347512246002774e-05, | |
| "loss": 0.4827, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.678153846153846, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 2.3154629017474384e-05, | |
| "loss": 0.4769, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 2.697846153846154, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 2.2834440237932536e-05, | |
| "loss": 0.5063, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 2.7175384615384615, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 2.251460898322712e-05, | |
| "loss": 0.4483, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 2.737230769230769, | |
| "grad_norm": 1.3984375, | |
| "learning_rate": 2.219518805615724e-05, | |
| "loss": 0.4855, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 2.756923076923077, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 2.1876230191778598e-05, | |
| "loss": 0.4663, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.7766153846153845, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 2.155778804869721e-05, | |
| "loss": 0.5065, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 2.796307692307692, | |
| "grad_norm": 1.25, | |
| "learning_rate": 2.123991420037565e-05, | |
| "loss": 0.4757, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 1.359375, | |
| "learning_rate": 2.0922661126453432e-05, | |
| "loss": 0.4768, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 2.835692307692308, | |
| "grad_norm": 1.265625, | |
| "learning_rate": 2.0606081204082797e-05, | |
| "loss": 0.4383, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 2.855384615384615, | |
| "grad_norm": 1.5625, | |
| "learning_rate": 2.02902266992815e-05, | |
| "loss": 0.4976, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.8750769230769233, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.9975149758303883e-05, | |
| "loss": 0.4871, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 2.8947692307692305, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.9660902399031782e-05, | |
| "loss": 0.4807, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 2.9144615384615387, | |
| "grad_norm": 1.4765625, | |
| "learning_rate": 1.9347536502386553e-05, | |
| "loss": 0.4544, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 2.934153846153846, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.9035103803763792e-05, | |
| "loss": 0.4924, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 2.953846153846154, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.8723655884491982e-05, | |
| "loss": 0.4846, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.9735384615384617, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.8413244163316696e-05, | |
| "loss": 0.4921, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 2.9932307692307694, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 1.8103919887911526e-05, | |
| "loss": 0.4728, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.012923076923077, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.7795734126417326e-05, | |
| "loss": 0.4531, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.0326153846153847, | |
| "grad_norm": 1.328125, | |
| "learning_rate": 1.7488737759011105e-05, | |
| "loss": 0.4468, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.0523076923076924, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.718298146950585e-05, | |
| "loss": 0.4727, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 1.6878515736982915e-05, | |
| "loss": 0.4429, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 3.0916923076923077, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.657539082745811e-05, | |
| "loss": 0.4304, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 3.1113846153846154, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.6273656785582986e-05, | |
| "loss": 0.4814, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 3.131076923076923, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.597336342638266e-05, | |
| "loss": 0.411, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 3.1507692307692308, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.5674560327031613e-05, | |
| "loss": 0.4318, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.1704615384615384, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.5377296818668638e-05, | |
| "loss": 0.4685, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 3.190153846153846, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 1.5081621978252548e-05, | |
| "loss": 0.423, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 3.209846153846154, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.47875846204597e-05, | |
| "loss": 0.4587, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 3.2295384615384615, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 1.449523328962496e-05, | |
| "loss": 0.4341, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 3.249230769230769, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 1.420461625172721e-05, | |
| "loss": 0.4596, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.268923076923077, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.3915781486420848e-05, | |
| "loss": 0.4357, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 3.2886153846153845, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.3628776679114517e-05, | |
| "loss": 0.4672, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 3.308307692307692, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 1.3343649213098486e-05, | |
| "loss": 0.4494, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.3060446161721855e-05, | |
| "loss": 0.4619, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 3.3476923076923075, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.277921428062091e-05, | |
| "loss": 0.4561, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 3.367384615384615, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 1.2500000000000006e-05, | |
| "loss": 0.4275, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 3.387076923076923, | |
| "grad_norm": 1.390625, | |
| "learning_rate": 1.2222849416966117e-05, | |
| "loss": 0.4704, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 3.406769230769231, | |
| "grad_norm": 1.296875, | |
| "learning_rate": 1.1947808287918404e-05, | |
| "loss": 0.4283, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 3.4264615384615382, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 1.1674922020994022e-05, | |
| "loss": 0.4346, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 3.4461538461538463, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 1.14042356685714e-05, | |
| "loss": 0.4613, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 3.465846153846154, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 1.1135793919832336e-05, | |
| "loss": 0.4634, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 3.4855384615384617, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 1.0869641093383962e-05, | |
| "loss": 0.4702, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 3.5052307692307694, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 1.0605821129941934e-05, | |
| "loss": 0.458, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 3.524923076923077, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 1.0344377585075998e-05, | |
| "loss": 0.4286, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 3.5446153846153847, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 1.0085353622019175e-05, | |
| "loss": 0.46, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 3.5643076923076924, | |
| "grad_norm": 1.25, | |
| "learning_rate": 9.82879200454167e-06, | |
| "loss": 0.4323, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 9.574735089890766e-06, | |
| "loss": 0.4452, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 3.6036923076923078, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 9.323224821797782e-06, | |
| "loss": 0.4605, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 3.6233846153846154, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 9.074302723553398e-06, | |
| "loss": 0.4871, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 3.643076923076923, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 8.8280098911523e-06, | |
| "loss": 0.4801, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.6627692307692308, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 8.584386986508388e-06, | |
| "loss": 0.4666, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 3.6824615384615385, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 8.343474230741715e-06, | |
| "loss": 0.4404, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 3.702153846153846, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 8.105311397538085e-06, | |
| "loss": 0.4526, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 3.721846153846154, | |
| "grad_norm": 1.4296875, | |
| "learning_rate": 7.869937806582642e-06, | |
| "loss": 0.4433, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 3.7415384615384615, | |
| "grad_norm": 1.1484375, | |
| "learning_rate": 7.63739231706833e-06, | |
| "loss": 0.4287, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.761230769230769, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 7.407713321280377e-06, | |
| "loss": 0.465, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 3.780923076923077, | |
| "grad_norm": 1.203125, | |
| "learning_rate": 7.180938738257944e-06, | |
| "loss": 0.445, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 3.8006153846153845, | |
| "grad_norm": 1.3203125, | |
| "learning_rate": 6.957106007533826e-06, | |
| "loss": 0.4544, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 3.820307692307692, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 6.736252082953307e-06, | |
| "loss": 0.4508, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 1.3671875, | |
| "learning_rate": 6.5184134265733e-06, | |
| "loss": 0.4575, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.8596923076923075, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 6.303626002642554e-06, | |
| "loss": 0.4432, | |
| "step": 392 | |
| }, | |
| { | |
| "epoch": 3.879384615384615, | |
| "grad_norm": 1.34375, | |
| "learning_rate": 6.091925271664156e-06, | |
| "loss": 0.4614, | |
| "step": 394 | |
| }, | |
| { | |
| "epoch": 3.8990769230769233, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 5.883346184541128e-06, | |
| "loss": 0.4645, | |
| "step": 396 | |
| }, | |
| { | |
| "epoch": 3.9187692307692306, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 5.67792317680616e-06, | |
| "loss": 0.4533, | |
| "step": 398 | |
| }, | |
| { | |
| "epoch": 3.9384615384615387, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 5.475690162936489e-06, | |
| "loss": 0.4232, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.958153846153846, | |
| "grad_norm": 2.046875, | |
| "learning_rate": 5.27668053075474e-06, | |
| "loss": 0.4266, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 3.977846153846154, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 5.0809271359167215e-06, | |
| "loss": 0.4529, | |
| "step": 404 | |
| }, | |
| { | |
| "epoch": 3.9975384615384613, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 4.888462296487128e-06, | |
| "loss": 0.4429, | |
| "step": 406 | |
| }, | |
| { | |
| "epoch": 4.017230769230769, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 4.699317787603927e-06, | |
| "loss": 0.4537, | |
| "step": 408 | |
| }, | |
| { | |
| "epoch": 4.036923076923077, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 4.513524836232458e-06, | |
| "loss": 0.4659, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.056615384615385, | |
| "grad_norm": 1.09375, | |
| "learning_rate": 4.331114116009938e-06, | |
| "loss": 0.4156, | |
| "step": 412 | |
| }, | |
| { | |
| "epoch": 4.076307692307692, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 4.152115742181434e-06, | |
| "loss": 0.4561, | |
| "step": 414 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 3.97655926662791e-06, | |
| "loss": 0.4438, | |
| "step": 416 | |
| }, | |
| { | |
| "epoch": 4.115692307692307, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 3.80447367298738e-06, | |
| "loss": 0.4331, | |
| "step": 418 | |
| }, | |
| { | |
| "epoch": 4.135384615384615, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 3.6358873718697726e-06, | |
| "loss": 0.4261, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.155076923076923, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 3.470828196166523e-06, | |
| "loss": 0.4629, | |
| "step": 422 | |
| }, | |
| { | |
| "epoch": 4.174769230769231, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 3.3093233964554466e-06, | |
| "loss": 0.4271, | |
| "step": 424 | |
| }, | |
| { | |
| "epoch": 4.194461538461539, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 3.151399636501773e-06, | |
| "loss": 0.4229, | |
| "step": 426 | |
| }, | |
| { | |
| "epoch": 4.214153846153846, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 2.997082988856087e-06, | |
| "loss": 0.4504, | |
| "step": 428 | |
| }, | |
| { | |
| "epoch": 4.233846153846154, | |
| "grad_norm": 1.4921875, | |
| "learning_rate": 2.8463989305498596e-06, | |
| "loss": 0.428, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 4.2535384615384615, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 2.699372338889297e-06, | |
| "loss": 0.4399, | |
| "step": 432 | |
| }, | |
| { | |
| "epoch": 4.27323076923077, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 2.5560274873481975e-06, | |
| "loss": 0.4375, | |
| "step": 434 | |
| }, | |
| { | |
| "epoch": 4.292923076923077, | |
| "grad_norm": 1.1875, | |
| "learning_rate": 2.416388041560491e-06, | |
| "loss": 0.4231, | |
| "step": 436 | |
| }, | |
| { | |
| "epoch": 4.312615384615385, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 2.2804770554131686e-06, | |
| "loss": 0.4409, | |
| "step": 438 | |
| }, | |
| { | |
| "epoch": 4.332307692307692, | |
| "grad_norm": 1.25, | |
| "learning_rate": 2.1483169672401686e-06, | |
| "loss": 0.4693, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 2.0199295961178893e-06, | |
| "loss": 0.4454, | |
| "step": 442 | |
| }, | |
| { | |
| "epoch": 4.3716923076923075, | |
| "grad_norm": 1.40625, | |
| "learning_rate": 1.895336138262968e-06, | |
| "loss": 0.4543, | |
| "step": 444 | |
| }, | |
| { | |
| "epoch": 4.391384615384616, | |
| "grad_norm": 1.1015625, | |
| "learning_rate": 1.7745571635328723e-06, | |
| "loss": 0.4302, | |
| "step": 446 | |
| }, | |
| { | |
| "epoch": 4.411076923076923, | |
| "grad_norm": 1.125, | |
| "learning_rate": 1.6576126120299045e-06, | |
| "loss": 0.4325, | |
| "step": 448 | |
| }, | |
| { | |
| "epoch": 4.430769230769231, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 1.5445217908091613e-06, | |
| "loss": 0.4406, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 4.450461538461538, | |
| "grad_norm": 1.21875, | |
| "learning_rate": 1.4353033706910296e-06, | |
| "loss": 0.4631, | |
| "step": 452 | |
| }, | |
| { | |
| "epoch": 4.470153846153846, | |
| "grad_norm": 1.15625, | |
| "learning_rate": 1.3299753831787192e-06, | |
| "loss": 0.4466, | |
| "step": 454 | |
| }, | |
| { | |
| "epoch": 4.489846153846154, | |
| "grad_norm": 1.4140625, | |
| "learning_rate": 1.2285552174813225e-06, | |
| "loss": 0.4379, | |
| "step": 456 | |
| }, | |
| { | |
| "epoch": 4.509538461538462, | |
| "grad_norm": 1.171875, | |
| "learning_rate": 1.131059617642935e-06, | |
| "loss": 0.443, | |
| "step": 458 | |
| }, | |
| { | |
| "epoch": 4.529230769230769, | |
| "grad_norm": 1.1796875, | |
| "learning_rate": 1.0375046797782866e-06, | |
| "loss": 0.4793, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 4.548923076923077, | |
| "grad_norm": 1.3046875, | |
| "learning_rate": 9.479058494153425e-07, | |
| "loss": 0.4512, | |
| "step": 462 | |
| }, | |
| { | |
| "epoch": 4.568615384615384, | |
| "grad_norm": 1.1328125, | |
| "learning_rate": 8.622779189453007e-07, | |
| "loss": 0.4558, | |
| "step": 464 | |
| }, | |
| { | |
| "epoch": 4.588307692307692, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 7.806350251804484e-07, | |
| "loss": 0.4365, | |
| "step": 466 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 1.3359375, | |
| "learning_rate": 7.029906470202046e-07, | |
| "loss": 0.4499, | |
| "step": 468 | |
| }, | |
| { | |
| "epoch": 4.627692307692308, | |
| "grad_norm": 1.421875, | |
| "learning_rate": 6.293576032258413e-07, | |
| "loss": 0.4228, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 4.647384615384615, | |
| "grad_norm": 1.3828125, | |
| "learning_rate": 5.597480503041486e-07, | |
| "loss": 0.4443, | |
| "step": 472 | |
| }, | |
| { | |
| "epoch": 4.667076923076923, | |
| "grad_norm": 1.375, | |
| "learning_rate": 4.941734805004289e-07, | |
| "loss": 0.4462, | |
| "step": 474 | |
| }, | |
| { | |
| "epoch": 4.686769230769231, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 4.326447199012068e-07, | |
| "loss": 0.4136, | |
| "step": 476 | |
| }, | |
| { | |
| "epoch": 4.7064615384615385, | |
| "grad_norm": 1.078125, | |
| "learning_rate": 3.751719266468584e-07, | |
| "loss": 0.418, | |
| "step": 478 | |
| }, | |
| { | |
| "epoch": 4.726153846153846, | |
| "grad_norm": 1.28125, | |
| "learning_rate": 3.217645892545695e-07, | |
| "loss": 0.437, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 4.745846153846154, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 2.724315250518056e-07, | |
| "loss": 0.4599, | |
| "step": 482 | |
| }, | |
| { | |
| "epoch": 4.765538461538462, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 2.271808787206092e-07, | |
| "loss": 0.4741, | |
| "step": 484 | |
| }, | |
| { | |
| "epoch": 4.785230769230769, | |
| "grad_norm": 1.0625, | |
| "learning_rate": 1.860201209529483e-07, | |
| "loss": 0.454, | |
| "step": 486 | |
| }, | |
| { | |
| "epoch": 4.804923076923077, | |
| "grad_norm": 1.2421875, | |
| "learning_rate": 1.489560472173468e-07, | |
| "loss": 0.4625, | |
| "step": 488 | |
| }, | |
| { | |
| "epoch": 4.8246153846153845, | |
| "grad_norm": 1.5390625, | |
| "learning_rate": 1.1599477663696845e-07, | |
| "loss": 0.443, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 4.844307692307693, | |
| "grad_norm": 1.109375, | |
| "learning_rate": 8.714175097937204e-08, | |
| "loss": 0.4617, | |
| "step": 492 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 1.2578125, | |
| "learning_rate": 6.240173375811343e-08, | |
| "loss": 0.4432, | |
| "step": 494 | |
| }, | |
| { | |
| "epoch": 4.883692307692308, | |
| "grad_norm": 1.3125, | |
| "learning_rate": 4.1778809446302304e-08, | |
| "loss": 0.4661, | |
| "step": 496 | |
| }, | |
| { | |
| "epoch": 4.903384615384615, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 2.5276382802272292e-08, | |
| "loss": 0.4307, | |
| "step": 498 | |
| }, | |
| { | |
| "epoch": 4.923076923076923, | |
| "grad_norm": 1.2890625, | |
| "learning_rate": 1.2897178307461067e-08, | |
| "loss": 0.4554, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.942769230769231, | |
| "grad_norm": 1.140625, | |
| "learning_rate": 4.6432397166285e-09, | |
| "loss": 0.4637, | |
| "step": 502 | |
| }, | |
| { | |
| "epoch": 4.962461538461539, | |
| "grad_norm": 1.3515625, | |
| "learning_rate": 5.159297204238023e-10, | |
| "loss": 0.459, | |
| "step": 504 | |
| } | |
| ], | |
| "logging_steps": 2, | |
| "max_steps": 505, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.4919954461790044e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |