| { | |
| "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": 6.139900313284506, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.982, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.473622382094706, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 1.0269, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 5.910909581534967, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.9791, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.431408087708716, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9529, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.334703544282361, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.9795, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.26958420542499, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.9199, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.2197406984704307, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8975, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.884094266166332, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.948, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.667464842732564, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8651, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.529371731289261, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8713, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.06753130984254, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8642, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.2447955376241544, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8517, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.36535604507628, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.8085, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.526891956841021, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7782, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.1647490207967346, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.7673, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.7399334665681228, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7761, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.5439752698084486, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7331, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.3812194097093384, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7492, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.4798548278502144, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.7403, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.0526178978200404, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7199, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.1577653852079144, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7152, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.0397874713592166, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7239, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.8574629739887523, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.7395, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.8584041733532946, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7081, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.7914487070712474, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.6798, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.8622551307947883, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.693, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.7141709276324931, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.7096, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7826771910410102, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.6783, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8196104708149089, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6886, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7678152225350511, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.7314, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.640089937892778, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6605, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.8767392026657923, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.6807, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7468693961838233, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6938, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 1.16314941785704, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6883, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7026278298844031, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6628, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.9461024061575943, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6729, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 1.2545201195647278, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6961, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.7875132453325014, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6582, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.7733392202468674, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6751, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.738302250579463, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6526, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.9196908996293846, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6743, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.9763670802760736, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6779, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.9431069206539577, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.6843, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7819206381595187, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6414, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.798766102120932, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6344, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7434111193997881, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6546, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.7622058324923021, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6794, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.7165251815424691, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.686, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.7666598707959372, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.671, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.822114138575259, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6406, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8304109518162148, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.6581, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7423622892079009, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6752, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.9244733683833497, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6779, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.8120899578253081, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6766, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.7396789785604715, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6175, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.7113013922011991, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6569, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 0.8416428672715566, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6572, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.8190171093043638, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6466, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.7377344598987262, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6558, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.6403193422024179, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6354, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6418318989162312, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6268, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.6466592208882427, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6589, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.559884936938963, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6265, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.5726799662645751, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6442, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.7368110470438402, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6425, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6627118682726961, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.6435, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.622824162799881, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6568, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.7152104954553946, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6303, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5115533412954613, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.657, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.7131340772915977, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6172, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.45376432530352945, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6516, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.6686657742052962, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.6116, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.5698483066198811, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.684, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.6622553669065425, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.6489, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.6163366290149376, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.6448, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.5879803664030521, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6467, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.7470416927518838, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6737, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.7633746768246295, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6534, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.6328206856822884, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.1966, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.6809959381505128, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.5308, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.7108664225775927, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5528, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.5690362601094925, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.5199, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.6877437567830023, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.6027, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6432448956921576, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5079, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.61240321338029, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5514, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.6495384017336064, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.521, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.5515222668973156, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5112, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6557901750788028, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5385, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.678073957221363, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5203, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6738015149322208, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.575, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6924736609760833, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5509, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6053215881817227, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4836, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6409549917926998, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.5257, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6388886427293657, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5103, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5936906156210888, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.5584, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.6219367151228186, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5223, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5998003093584972, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.544, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.53934222313818, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5268, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.5786943481965757, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5233, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.4852746743697653, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5219, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5568093534765395, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.5342, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5900775911771164, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.5508, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5564911637444528, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.532, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.4738239679386976, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.5167, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6054046039712183, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5126, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5313249790157942, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.5147, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5656201581530086, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.5415, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.7540038112567037, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5364, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.5574157401204813, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.5204, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.5795114768460621, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.5244, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5899632241274246, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5362, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.6232312143547775, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.5284, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.43533930034130625, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4916, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5943074651459542, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.5183, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.6220991957899387, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.5475, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.47120782275329426, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.527, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.6598175298583286, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.5145, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.5407768307770716, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.5142, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.5180342583907449, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.51, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.471400620593815, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.5329, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.5839612982923811, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.5138, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.580364050557502, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5476, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5926105498575768, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.541, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.5406108444064682, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.4943, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.5421264786835023, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.5328, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5228632295441396, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.5432, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.4990238336942576, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.5602, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.5399681282004855, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.519, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.440617066226563, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5049, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5212220366780365, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.5196, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5902994703466927, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.5132, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.4683027745405901, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.513, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.5574702026227369, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.542, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.43647529415507, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.4811, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.6876855413605403, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5371, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5609940338725775, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5239, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.5651266791805881, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.512, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.4816258629758263, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.5029, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.6118922184860773, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.5176, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.4659771506765139, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.5152, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.6380533273212581, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5329, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.7435818393803963, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.5288, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.46224785706377947, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.5246, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.7625858979093663, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.5053, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5937164437831631, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.525, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.6586299029487668, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.5405, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.7165219101216355, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.5421, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.577436645474833, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.5028, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.6469818834033737, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.5673, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.4944699950408038, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.5157, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.49832676352746946, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.5209, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5657882920267109, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.5341, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.42657603946839406, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.517, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.5702179709147125, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.5279, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.42790970957491925, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.4982, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5833807456376587, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5073, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.382854032367803, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.949, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6021437111476249, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3609, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.6365129317567508, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.421, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.8414906913897036, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.4307, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.8014686895578418, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3598, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.6232847673678833, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.4099, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6477273935703248, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.375, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.8372381943320925, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.4337, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.7456930993505484, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3786, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.621740680516258, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3849, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.8056442824166912, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3796, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.590643603055589, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3878, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.6981743964305072, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.3898, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5660004809953806, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3826, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5494053076582944, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.3772, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.697896268022206, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3887, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.5917790844528241, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.4056, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.6333276316810618, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3811, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5206503855507073, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3622, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.6293182752692434, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3704, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.5356258671041978, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.3764, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.5111433443431654, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3701, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5026478617605791, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3568, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.49371109027204074, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3902, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.5323557517059921, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3778, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.49459861006905453, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.4065, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.5290369090539728, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3633, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.4941638139989671, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.415, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.5705981660680955, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3671, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.4361858145773666, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3587, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.7360021293702974, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.3951, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.45095818052494235, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3355, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.6356752877203468, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3792, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.462190668442627, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3937, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.525602730135968, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3595, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.408941022265278, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3529, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.44857848693349356, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3854, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.35705359896891403, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.3517, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.365435879439438, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3531, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.38444723536304637, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3701, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.42636636474465983, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3718, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.40112451176008684, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.4039, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.41417640033535225, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3847, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.3828957431795322, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3764, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.35665689173097015, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3552, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.3435643831700335, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3565, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.3786752581896812, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3776, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.385068667799081, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3707, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.38566486096960945, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3622, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.38118601793316903, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3712, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3992700053019032, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3809, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.3435263019065936, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3651, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.3965688643433488, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.398, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.36190663496273573, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3753, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.3661346371065075, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3594, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.39248111500085064, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3958, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.35855092384056914, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3624, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.34151475671906895, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3456, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.3862807689652061, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.4113, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.38219624353153225, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3658, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4032351956658789, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.395, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.37995525447353606, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3791, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.3382288828909466, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3538, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.42387717817282067, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3803, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.3554948928537524, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3655, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3682029796161342, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3717, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.41226031268261126, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.39, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.36008077206631883, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3556, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3843785833818051, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3805, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.36248959871970426, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3778, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.38243368425583524, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.3752, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3782656481654194, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3734, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.35160882734595184, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3571, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3534239598287396, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3614, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.37128532250247387, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3665, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.39290459892523866, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.3656, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.3351183261832644, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3588, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.37868875230821064, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3736, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.2225278041694008, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.6085, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.6053479435655329, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2532, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.4960349269967615, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.2594, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 1.094336972920146, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2901, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.5623869298091444, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.2419, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4605589086986637, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2263, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.6113060381787875, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2826, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.5498176595225925, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2494, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.49462495276119733, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2713, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.40611803510055516, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2329, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4626974488623721, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2314, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.5727202481863393, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2624, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.3952638735914865, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2319, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.4151601627234044, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2435, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.44574932050922483, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.249, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.4202213295671942, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2405, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.4119069382746125, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2565, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.3997971965406744, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.235, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.4121201525803537, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2649, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.427764097481975, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2708, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3383493217142155, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2193, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.40262438337775447, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2557, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.37942071316306536, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2557, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.34798108627478713, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2305, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.396243492215517, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2635, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3766943443838131, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2525, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.32986392716560653, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.2307, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3900246469971243, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2728, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.36817582539460925, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2474, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3492778365503047, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2367, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.35795971328526993, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2485, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.3362446480479155, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2453, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.3808204119941009, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.2579, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.32199981853606435, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2149, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.34055185666395094, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2674, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.34319675875279254, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.25, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.35450735922153487, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.2572, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.32301229807323845, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2426, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.33360571369987735, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.244, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.35475372146450995, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2519, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.33027102870941494, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2362, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3537481325853907, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2617, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.34919045586331077, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.2512, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3487481679637291, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2368, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.34781919100875563, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2517, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.3553434464222214, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2602, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.33290663380142466, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.2437, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.31417822911534743, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2398, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3501054412092899, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2659, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.34330275642918745, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2412, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3489691146370825, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2561, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.32326217619302994, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2322, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.3153886513180333, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.2504, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.33112840526232246, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2497, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3499125610855062, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.2516, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.3482579304430075, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2736, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.307584837957961, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2226, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.31813523929278154, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2267, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.35074114352131847, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.2672, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.3137130420840304, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.234, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.31999076131727244, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2419, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3286879120075679, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2387, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.3482525695657893, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2657, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3153274429334436, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2062, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.32997236092801496, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2509, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.3383457685502815, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2544, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.3291807639457476, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2234, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.3221378820137518, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.241, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.31855378367692677, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.2397, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.30917825057221815, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.2334, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.31832109090187977, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2185, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.3291207731984391, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2514, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.31732383870143727, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2408, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.3306355602610903, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2586, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.337537976669306, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2388, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.3452438405556631, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.2619, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.31799925944004126, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2409, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.3237025330240859, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.2497, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 1.1580735050841415, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.4501, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5349320336111352, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1831, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.4680550632896594, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1957, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3422229498066871, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1792, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.28659057616263406, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.152, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.40407672427764263, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1833, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.506605614272818, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1803, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.5502359359170463, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1862, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.4627040316305199, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1793, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3685148418808579, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1779, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.3260570230802829, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1795, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.3305558750830745, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1662, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.348046057734535, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1786, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3428019535424454, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1796, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.3488041862399251, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1695, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.3112195656613761, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1625, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.3624118374015353, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1802, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.3210111371550455, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1664, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.3208395212417022, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1805, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.30152701938808846, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1574, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.30495257711676116, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1791, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.31714190640037976, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1731, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3081705729124222, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1588, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.3277678143873994, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1782, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.3273647166019751, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.171, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.3060170821767323, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.171, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.28265569851455236, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1643, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.28294087109523286, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1784, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2974833917357202, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1719, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.29803036395489585, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1819, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.2857724687150965, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1603, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.29218941322791125, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1702, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2856185836716623, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1518, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.2896695368323106, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1678, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.30122355165183445, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1676, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.31143686572757046, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.18, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.28814648850515306, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1603, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.29270987825615, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1733, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.30384190367308855, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.178, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.29966938514840474, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1709, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2738927246761317, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1558, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.28800396870610384, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1806, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2795288357570854, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1678, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.3035823411646313, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1792, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2868531728401176, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1584, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2908081669532781, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1585, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2913224049259, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1756, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.288553017272967, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1599, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.304959135466857, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.178, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.28971946845924185, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1775, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.28823728241901475, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1739, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.29665753453484345, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1635, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.2884202099523914, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1707, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.2709423202270636, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1663, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.27973649405640016, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1684, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.3008493990397883, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1661, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.2859739452422636, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1641, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.28489504380487607, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1709, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2828402125078142, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1705, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2789185090773921, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1582, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.3097260254470352, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1769, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.29042603958469415, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1705, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2852345385355369, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1735, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2948683935024499, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1711, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.2945063554480941, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1794, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.2979265454392112, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1807, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.28965682187212005, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1598, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.28447178550873164, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1622, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.3090272020586994, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1785, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.2848764300570547, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1462, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.30718607049058466, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1897, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.2976727734081797, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.182, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2856637809819105, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1495, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.29865264342007625, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1831, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2649530999654682, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1579, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2828777671993983, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1786, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.28880989111643857, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1642, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2815283257193674, | |
| "learning_rate": 0.0, | |
| "loss": 0.152, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 8.312172410338017e+17, | |
| "train_loss": 0.4122787300592814, | |
| "train_runtime": 14321.8441, | |
| "train_samples_per_second": 3.491, | |
| "train_steps_per_second": 0.027 | |
| } | |
| ], | |
| "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": 8.312172410338017e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |