| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.992, | |
| "eval_steps": 500, | |
| "global_step": 390, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0128, | |
| "grad_norm": 6.324252051119927, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 1.0118, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 5.9688058893493015, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.9778, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.244871187955355, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0174, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.833516015151964, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 1.0153, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.2088745594806225, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9213, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.2480169976588855, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9111, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.0729457120394987, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8605, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.33434237598107, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.8544, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.72290955613629, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8669, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.106659922442911, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.791, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.129899115930081, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8698, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.228332758631176, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8335, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.3731347434300254, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8246, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.3321767271863747, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.765, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.322133937571792, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7677, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 2.137166466413172, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.8038, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.5077821136212597, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7408, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.3954266042040517, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7092, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.3678502303917293, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7416, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.0690503888860323, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7298, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.175688640252805, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7167, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 0.9309518689266235, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.6797, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.8242696146754106, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.6826, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.9649513002930041, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.6969, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7738353042967385, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.7236, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.0696118185416132, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.6863, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.9708847335653784, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6773, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.9335630064211746, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7053, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8868265607851216, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.7015, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7402426767040144, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.6618, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.6693978617985297, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6523, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.9529137417786196, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6841, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.8822470714498974, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6522, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.9000609238475252, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6584, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 1.1851467287982542, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.673, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 1.1997940767378104, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6816, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.1189537878350022, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6718, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.6893978184611891, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6099, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 1.2224211223243404, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6782, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.7430870996176311, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6495, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 1.1631964179009207, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6738, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.642535947337299, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6458, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9962446651136049, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6363, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.6529383216928, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6325, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.665318254171301, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6534, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.6165335027408949, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.653, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.5574173362541408, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.65, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.6280826462058071, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6541, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.6203280746694708, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.632, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.7007741257765994, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.656, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.693194050161898, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6726, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.6560371713795997, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6117, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.5849323604376722, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6447, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.6651148218627327, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6153, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.6642543008033831, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6518, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.8280440038955447, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6471, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.8056463476272753, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6171, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.9114939367475415, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6387, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.6745636018596052, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6236, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.752070194057599, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6597, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.8368601735451879, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6483, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.7071627306959537, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6507, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 3.800411031081067, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6328, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 1.3318246622346792, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6466, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 36.624122862808996, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6872, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 3.6106773044196645, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6639, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 2.8089784851013544, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6653, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 788.8153114703023, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6884, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 3.4468425171161203, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6679, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 2.4154982548981354, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6584, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 3.4397979540116843, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6647, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 2.4830540908326473, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6037, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 10.98602106549027, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6866, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 5.6181889292101195, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6853, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 17.36040099890159, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6571, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 1.652566119147992, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6696, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.7850084757079352, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6425, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 1.0797119553052676, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6081, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 2.3414404642730293, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.2019, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 1.2716766619944198, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5666, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 1.0290905482327615, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5182, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.2067987776778961, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5922, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.9803163108607817, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5332, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.8238209548230994, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5353, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.004416298955316, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5065, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.7211965998844605, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.5473, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 1.2436254609064055, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5244, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.9805090829463198, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5303, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 1.1046047507479493, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.515, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.965414772982106, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.5213, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 1.0721945559517085, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5098, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.742233783177718, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.5176, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 1.0414798010167508, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5222, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.7078245964036027, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5069, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.9057657006415607, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5293, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.7521752649808066, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5498, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.8904687712456775, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.5481, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.7845771995969328, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5098, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.6724403519697602, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5031, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.662965205080792, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5074, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5034752545406941, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5231, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6234343539559026, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5022, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5355447459662184, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5243, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.5225054680622154, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4882, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6110003038637313, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5357, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5842957309603038, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5094, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5504736920136277, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5191, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.6815278589698465, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5018, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.5121802313259853, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4776, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.589803047738699, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4861, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5158418545012493, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5401, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.7049801692159384, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5503, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.4763507525005289, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.5081, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5760616158786693, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5017, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.49570421076276977, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.522, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5385201660398796, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.4978, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.49638755874437296, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4928, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5173601779618554, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5265, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4462347902835942, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4909, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.5136298545214933, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5185, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.44845455485742997, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4978, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4821976201537964, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5078, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.4588415094532159, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4977, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.4306423128937503, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5287, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.4595823862828167, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.4822, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.4977274591989858, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5153, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.5295460450934338, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4771, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5027355161360124, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5172, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.4963205124518233, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5087, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.493355963617873, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.519, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.47644914245232, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5142, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.52917928043368, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.5001, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.46466848311758996, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.5137, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5473904898711462, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5163, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.4676146281145352, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.53, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.6010280801028324, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5337, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.4861120822978407, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4781, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5809382109321362, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5099, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.5045504493570258, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5027, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.5190861665820314, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5255, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.50179308962177, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5048, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.525310203263823, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5021, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.5051835108323796, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4765, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.4842217182471228, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4794, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.4674534463067605, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.4997, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.5561324491099091, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5225, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.5764827153627148, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5312, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.4669904355658487, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5309, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.5093472405434888, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5236, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4016156712538288, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4989, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.43441485786977085, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5211, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.512318190339783, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5139, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4880423520782735, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.5077, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.4837398747563273, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5255, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.5123310957913444, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.5124, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.4709397478767053, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.4858, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.4207707116922543, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.8495, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.5729581518724675, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3794, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.4011543110078455, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3605, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6923050280824109, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3671, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6638379166829008, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.347, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.4996825290947737, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3499, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.60863040070661, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3712, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.6210950367824146, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3703, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.5222476508115615, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.337, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5503313816811329, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3403, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5580536586074936, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3672, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5059233641942925, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3315, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.5084692075893241, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3641, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.6977546794079632, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3733, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5141701276060532, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3522, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5083976262661234, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3197, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5390737504337718, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3804, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.46761536620076877, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3664, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4886758953109607, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3332, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4988532657430669, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.379, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.4796126806857884, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3673, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.4795088339000365, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3479, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.481883320600888, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3586, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4827546151526932, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3699, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5987323339194617, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3803, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.460569275759714, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3512, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4566588410872878, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3327, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.56510264336169, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3728, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.41267925901617286, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3485, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.5108655076115967, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3324, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.4441743249786229, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3557, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.44445595970220786, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3478, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.4305206928781103, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.353, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.45495436048705845, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3773, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.4798767472271238, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3592, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.41336243573761117, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3366, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.5064603585459536, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3431, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.44772018238622424, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3602, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4595080497377754, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3532, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.5020966458621507, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.349, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.4194885717874874, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3587, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.5317909004393624, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3442, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.4150644644262044, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.354, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4701882497703371, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3384, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.42558943275434047, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3227, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.4833154639913761, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3949, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.45692304748578516, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3428, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.4266961048727715, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3603, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.3930624436830229, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3373, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.421201541612621, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3743, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4655099752218029, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3757, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.4105819963713678, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3587, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.43143315744228605, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3437, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.37270812230658906, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3336, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.43077744900598874, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3566, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.41535609186653355, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.351, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.42501681344614844, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3474, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.40390183276210745, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3529, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4317929856767169, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3536, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.4166330717960872, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3399, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4575213204869802, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3597, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.4200233056583318, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3589, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.42270639093493007, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3413, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.40192276924468645, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.361, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3936550575488538, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3638, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.41225989131078655, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3709, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.3777504210917383, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3473, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.43088560466303416, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3664, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.4441699690685119, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3609, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.4064661098758121, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3541, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3995158390100949, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3425, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.404359492106528, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3786, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.3948009095770409, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3423, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.39463488414779574, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3552, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.39938292896127475, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3468, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.382557471031179, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3593, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.40369692953579894, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.354, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3931581059893209, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3409, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.2987220349294752, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.5748, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.5668116963384578, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2413, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.4508255465227191, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2407, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.7561695394562868, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2206, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.6923858058648299, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.219, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4370442265219903, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2215, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.5037235840588825, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2247, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5395254419340598, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2257, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.46435251502023517, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.1997, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.4851995088021078, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.237, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4087707657119096, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2239, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4842697994619925, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2301, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.4270154774453825, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2107, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4253014801224203, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2476, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.41433560609729475, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2233, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.4320708737343593, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2401, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.39421182476471306, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2046, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.41748129609836426, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2441, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.36234686013969153, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2046, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.41227372495225545, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2225, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.43245284849830956, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2168, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.45204698938880367, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2349, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.39561318063946993, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2201, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3563963510037156, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.1845, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.3858057649661854, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.225, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.40436024094147116, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2272, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.39136900123560114, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2019, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.36398554728176497, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2288, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.3439157889105267, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2072, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.4076518339166455, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2416, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3623963936134355, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.1921, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.37872893292532045, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2102, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.3466735275648323, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2092, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.37556003977591806, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.237, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3570018229257897, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2096, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3569925639593487, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2051, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.39518912644546783, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2165, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3666082629558139, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2263, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.3481815873573447, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.2242, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.34999851581999175, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.1959, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.34705743997557603, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2204, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3560862543589103, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2248, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.4154146223058893, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2281, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3385571943955062, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2058, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.3730416638794807, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2388, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.34048172115616904, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2247, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.35346238730365476, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2246, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3691681824103454, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2236, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.34555779073027054, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.212, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.33228691677287675, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2163, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.33116808364349143, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2209, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3248075791375289, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.1963, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.3673420006713999, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.208, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.3524839209457635, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2047, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3561434502553937, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2431, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3206476573884973, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2041, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.33021616223600253, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.1947, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.3628644313055313, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2582, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.321381514400304, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.203, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.34191406782651407, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2159, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.356687314357332, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2257, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3264621744486339, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.1989, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.3507112339812231, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.219, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3379308464312734, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2148, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.33957852294210744, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2207, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.34706502355920943, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2223, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.337652308769353, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2068, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.3275040680509758, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2031, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.33929507675080023, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2093, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.3420758542066909, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.227, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.3287262744762415, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.22, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.3167874857575128, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2061, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.3629462876358366, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2541, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.32226911303205164, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.226, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.32487524993828437, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2119, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.32182246014789745, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2148, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.344609112403515, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2224, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.33126977578419603, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2102, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.8928180027786778, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3626, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5133052882940572, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.153, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.43952848015743234, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1465, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3615388753212631, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1505, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.34752704505542265, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1632, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3622026736292617, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1558, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.40613384813422454, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1434, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.49660776976314874, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1413, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.4850158257992296, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1489, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.407814257048164, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1399, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3659472791417999, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1452, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.3237648511804672, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1486, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3099402560550694, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1463, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.30923744110346674, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1307, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.39275645674895815, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.143, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.36687522069846723, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1416, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.35337012188142464, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1369, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.29714114688974724, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1253, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.2983297435113719, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.13, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.3360872849468959, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1497, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.32969176235041037, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1281, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.3331799957333484, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1437, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3199313310693573, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1378, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.3599164741601635, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.152, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.3127820085228739, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1166, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.3171705067489213, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1283, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.29653073797418444, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1252, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.3398709251581674, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1602, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.323874486342724, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1502, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.29206602816578614, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1414, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.3044301707532178, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1545, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.283173557135327, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1236, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2957509665366772, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1371, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.3140500718880482, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1569, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.29237677131436585, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1237, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.30893018591907706, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1414, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.3020127755524499, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1508, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.2772043949464147, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1362, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.29664609896083083, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1539, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2904219062510931, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1312, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2976163514043663, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1313, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.3098027741238999, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1539, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.31562023869446976, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.157, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.28750651075875144, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1341, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2965240716314853, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1489, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.27107178393363657, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1114, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.29405680231526776, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1624, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.2878107351951627, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1321, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2819176186542553, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1208, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.27164935545415797, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1367, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.29365323896535034, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1424, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.29219021088964675, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1482, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.29328744769931114, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1384, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.2897195967651209, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1418, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.3039354513226491, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1546, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.27153358694788043, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1294, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2962140421998007, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1433, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.2898192512721484, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1385, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2965185514539473, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1511, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.28412045908542694, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1336, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.3067646457003337, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1509, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.2857789123764494, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1244, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.3093513670201484, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1529, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.28794825402823093, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1434, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.27964844546586815, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.126, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.3150346673765271, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1626, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2639673603000804, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1107, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.301802170262654, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1549, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.3030519612218664, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1479, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.27912680019639596, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1267, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.301239301045312, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1442, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2831791987871808, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.1301, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.3071896028636736, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1588, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2935549135659552, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1432, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2968680639534526, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1372, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2952253724744376, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1426, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2975852496245735, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1316, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.3081936101192452, | |
| "learning_rate": 0.0, | |
| "loss": 0.1378, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 6.670095881926083e+17, | |
| "train_loss": 0.39215392955602746, | |
| "train_runtime": 12456.6735, | |
| "train_samples_per_second": 4.014, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "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": 6.670095881926083e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |