| { | |
| "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": 6.115225457318272, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.7972, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.4778649832938, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.8223, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.439432589198134, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 0.8336, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.916085347437607, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.794, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.649690599890324, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.7651, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.3516014711724074, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.6905, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 2.0210867581306613, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.6952, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.1647802159924066, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.7169, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.285557439271826, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.6915, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.742700936931223, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.6619, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.345916968810201, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.6244, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.1198253262604476, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.6451, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 1.854722364751461, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.5837, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.6202210991889734, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.6217, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.467827204969234, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.5798, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 1.2901839247708227, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.5836, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.0524795187246585, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.5692, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.6043224605320856, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.5896, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.083519275986709, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.5607, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 1.1676451919151445, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.5338, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.1649317866906146, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.5467, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.0070139617846363, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.5649, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.8808640916185065, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.5375, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 0.6891911050508679, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.4995, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8034299306922472, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.5509, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 0.6810478528470438, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.5251, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.6628937158980546, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.493, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7071258758781976, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.5124, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.5730112406456523, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.4934, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.7591360874451095, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.4959, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.7433414478275835, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.5276, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.5230218083058822, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.5056, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.5788315940951135, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.4775, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.628996593385306, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.5025, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.6364819152597294, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.4956, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.6952439014406236, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.4795, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.5879249245725728, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.4886, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.7143567047490685, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.4662, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.754270169595988, | |
| "learning_rate": 4e-05, | |
| "loss": 0.4845, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.6916230753184576, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.4946, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.6655929376045815, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.4751, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 0.7019272485859203, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.4879, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.8696723571749855, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.4667, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7429977461233525, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.4761, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 0.6918166294558654, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.4654, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.7171915391313837, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.4543, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.9295428779797352, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.4758, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.7984379574635976, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.4906, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.9454295106930493, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.4632, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.8457972079509006, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.4629, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.8777267910421587, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.4996, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7765226591426262, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.4734, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.9927734864349484, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.4543, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 0.9443366075284568, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.4908, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 1.0531525900028036, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.4663, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.799320060729199, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.4809, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.1131030340708192, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.4784, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.7567089549779984, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.4806, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 1.0788676254710285, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.4854, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.7991461438625849, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.4903, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6199089194762689, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.4707, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.7896919100970462, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.4474, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.45780085481027216, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.4529, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.7221706609456762, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.445, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.5165115410834582, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.4589, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.7202953596591821, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.4857, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.5705334637961889, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.4618, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.5500129664597081, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.4492, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5744757957783347, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.4429, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5802945866116739, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.4579, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.7235041394840783, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.4445, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5041980422363816, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.4653, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.660792488065925, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.4588, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.5486994848030444, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.4429, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.7034073232264785, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.4826, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.5059700872962927, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.4544, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.6369868161296368, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.4613, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 0.48870471079255756, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.4449, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 1.2635183664979024, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 0.7069, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.8268755545602047, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.3875, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.6721621975488452, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.3507, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 0.6347440615314595, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4196, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 0.7367054221518949, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.3929, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 0.6821773954639229, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.4045, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 0.7612172229842501, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.3792, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.5906116108616847, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.3781, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 0.6312131712709363, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.3997, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.6329032589193756, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.416, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.5997371849044506, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.3655, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.6121363195901529, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.3895, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.7461859611022873, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.403, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.6457462210535605, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.3995, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.6276460566924693, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.3953, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.6651405841975484, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.4154, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5545620200098397, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.3607, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.6397790668991523, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.416, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.4720374118182995, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.3556, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.6346584954709299, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.4022, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.46025361233210943, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.3802, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.6179922371263346, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.4032, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.5123787588585933, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.3724, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.5016395997791134, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.3848, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5275969489140777, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.4048, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.4274247976232669, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.3684, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.43690661889705235, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.3689, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.5027782707045861, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.3929, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.4043604339052812, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.3837, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.45253437084829845, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.3822, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.5111440287154948, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.391, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.43656648959033406, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.3987, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.447739635005765, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.3669, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.4891443852288491, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.403, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5354777825209106, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4006, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.44665401713182695, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.3685, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.5653760562391243, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.3875, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.4540472373860856, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.3653, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.6113566659285303, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4005, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.4168488729472874, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.3654, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.4620977772662231, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.4001, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.476420893899808, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.3572, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.381303454265223, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.3467, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.4863320007755061, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.4082, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.381600758560905, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.3347, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.394620847076037, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.3879, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.44406295699549964, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.3642, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.39408607155522185, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4074, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.49314747982309703, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.3801, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.36062471985657507, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.3757, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.48026342944805833, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.4022, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.4489396246814716, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.3948, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.47748548923970513, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.3889, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.45412392757138254, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.3741, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.49636144268343996, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.3891, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.4390905719231864, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.3739, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5158833144634472, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.3685, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.4094994605665896, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.3637, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.46870316387196487, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.3995, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.39827114114257894, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.3717, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.452460089054947, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.3746, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.4742389634137148, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.3919, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.3725927079643266, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.3711, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.4331691811604596, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.3822, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.4419833649320582, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.3996, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.42256653425091645, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.3613, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.48066707088078137, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.3801, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.43537246948490294, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.3793, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.4202591485432943, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.3649, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.394523012460404, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.3448, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.39136239240309023, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.3763, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.41271571999739015, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.3694, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.40875329398478466, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.3931, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.3883380354441773, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.3628, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.4091328267428603, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4089, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.3570274352018894, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.3163, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.443729606957663, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.413, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.3818874557795846, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.3589, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.029327384874714, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.616, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.7112912072965769, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.2665, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.3570251966441835, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3094, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6606729021015588, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.2876, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6071440995571158, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3013, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.5168033718618819, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.2793, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.5597858928265351, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.2753, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.4676542419821463, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.2938, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.624694783602782, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.3133, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.4508439076138672, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.2637, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5508607119104917, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.3112, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5334197735316446, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3028, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.5087020493527552, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.2822, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.5113084597267213, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3134, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5879580442573731, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.2833, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.4073874911657843, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.2672, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.6308940531620888, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3206, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.48145709078390986, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.2456, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.5086807605931843, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.2885, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.5065318303439502, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3178, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.5214188023044475, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.2685, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.48564632364190025, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3006, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.5081858291521164, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.2559, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.4873213781242658, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3054, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.4970841886138032, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.2812, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.48557798780934575, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3065, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4882395532803412, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3054, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.42507926185662265, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.2787, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.4522538873448131, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.2745, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.39747296793816345, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.2837, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.43180929878370683, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.2776, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.422046755560608, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.289, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.3939134040727102, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3012, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.4360103933699082, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.2589, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.43222381961527234, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.2996, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.39947532282855386, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.2895, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.468193590172406, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3278, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.37028122335191704, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.2592, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.40521601819241465, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.2738, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.39880849277987285, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.2889, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.4360800776688632, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3022, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.332952630190702, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.2595, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.35302482121437867, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.2641, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.3671929433847647, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.2933, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.3810125942961015, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.304, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.3571039273418628, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.2741, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.3797320221202612, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.2851, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.36043504682960464, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.2889, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.40755688277419305, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.2663, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.34514855584879406, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.2848, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.3838068439029518, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.2879, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.3847498131506985, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.27, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.34940071010776225, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.2778, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.4223538244246661, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.2938, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.35980587807766456, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.2721, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.36119284083682335, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.276, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.3682108196916936, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.2421, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.3538115363318626, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.2822, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.36504269855271493, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.2569, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.36439562322493063, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.2765, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.3366573070490289, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.2883, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.3366306462927351, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.2828, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.36513014328007964, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.2852, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3117418449051438, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.2689, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.339060400239575, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.2986, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3666591657683088, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.314, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.32066906189526106, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.2704, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.33772346517378604, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3129, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.35495336862264104, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.2541, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3010062534177195, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.2781, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.37288601253513487, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.2693, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.3280334445303061, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.2794, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.34011244409998703, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.2702, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.3294569180096905, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.2684, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.3418098557951546, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.2886, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.322402298730085, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.2757, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.34587493019466187, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.2928, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3177547577906029, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.2677, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.0570934459380867, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.4456, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.4983219007754451, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.1978, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.40594125810558024, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.1948, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.6097176434241903, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2023, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.540156981197435, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.1948, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.449124003378837, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2137, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.44005144456712886, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.1894, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.38990337829232735, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.1736, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.5144918977495327, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2129, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.3957954029581393, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2238, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.3991613810037469, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2079, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.4364261994509852, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2064, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.36711860268112606, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.1754, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.41000829446821285, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.1975, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.37471080060425027, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2116, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.3752427978399487, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.1939, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.38504540888291255, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2053, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.344247206773842, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.189, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3655250175724291, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2025, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.3845876904971133, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2034, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.3554121162054489, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.1916, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.37347476325986834, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2182, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.3341073684904624, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.1814, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.3504981331843485, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2023, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.34781101624044813, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.187, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.3127735331627072, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.1638, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.33265237990198615, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.1862, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.35015832867925345, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.206, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.32908095503184986, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.1778, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.37537224375824596, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2185, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.30543957154824575, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.1874, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.33672682853021696, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2084, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.2955685048428476, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.1635, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.3171529594970444, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.1842, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.34736581552759127, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2027, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.3150102183187486, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.1928, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.3329004389577658, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.1914, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.3423245818888095, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.1857, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.3075317039384064, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.1992, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.31724733790920756, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.1926, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.30076557190569597, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.1809, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.30519863258037544, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2143, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.29506570414695543, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.1767, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.37820461333233696, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.1928, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.30779340816976763, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.1775, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.34605248083621654, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.2056, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3114467514691629, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.1774, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.3115269201330818, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.1832, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3450165912378589, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2018, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.3227156104513421, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.1753, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3029791254137128, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.1857, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.3088465643618022, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.1887, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.305263460649981, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.1732, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.30080533998033043, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.1835, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3101978906920816, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.1975, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.31151270349369586, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.1742, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.31759513152883234, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.1773, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.32097522054765193, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.1854, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.30042618451980785, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.1913, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.30588959543731825, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.1897, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.2877406552784323, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.1834, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.30319113463160563, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.1979, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.3026029014241974, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.1795, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3220822432074593, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2122, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.286174071468055, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.1689, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.2976224108721899, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.1721, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.3026819830570282, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2155, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.2771619310052799, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.1855, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.3110088221545863, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.1899, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.27848627362089545, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.1703, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.32841441923674536, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.201, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.29364084745073127, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.1994, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.28912769750239053, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.183, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.29953889869329897, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.1912, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.30085837546384075, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.1889, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.28765182024887026, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.1804, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.28983659076395446, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.1988, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.2797250120247476, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.1933, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 0.8579904807362196, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3106, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.3686807073622712, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1257, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.37815873740386835, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1444, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3134039760178023, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1559, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.259911145416525, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1183, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.3560684167412651, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1478, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.3388086260272147, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1135, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.4595029727093649, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1565, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.3829975297149748, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1394, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3557753474699705, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1577, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.277898611629972, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1368, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.3009807450851778, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1494, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.27962571461409447, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1328, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.29841854123122513, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.1379, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.31715843446203096, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1294, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.2866957811931399, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1412, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.27339643221481014, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1187, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.2767366552244504, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1256, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.3235338230264642, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1462, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.29220332383844744, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1346, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.2620875042907759, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1181, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.2753430473262122, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1264, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3055838045897359, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1352, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.26108815755898396, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.111, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.27344902870130877, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1304, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.28104922812306843, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.1442, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.2589803368614472, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1325, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.26619056445422595, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.1259, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.2717324953630556, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1401, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.2608867229914016, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1285, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.26744213800042205, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1465, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.2654610799356427, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1279, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2537264472875052, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1243, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.27063300969206494, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1417, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.2715344987158459, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1368, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.2774586022594309, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1451, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.25823280715305535, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.122, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.24994646655316943, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.1203, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.26102830614544353, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1366, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.2621874678042739, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1446, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.2523200217151637, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1444, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.26357906948775117, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1483, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.254595565148833, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1464, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.25541870607323525, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1296, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.2678438520080487, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.1367, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.27246834098654793, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1376, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.2545316898387819, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1234, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.23980914455097604, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1272, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.27533603789694006, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.1409, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.24256452436055875, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1184, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.26512113226814327, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.137, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.25057089470542104, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1343, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.26905994457509647, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1384, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.2515567887129915, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1183, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.25860107318444187, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1222, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.2647431542641901, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1415, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.24625643686363832, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1234, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.24904863719886972, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1197, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.23781029435536066, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1261, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.2634085996145061, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1454, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.2533454000515048, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1444, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.24985204283286847, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.129, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.2559467668931813, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1268, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.2626131517894016, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1374, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.25223516188684386, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1344, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.24653382543023958, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1445, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2487358839721311, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.126, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.27092325982020843, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.1449, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.2419114914044283, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1231, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.28368695654925413, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1322, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.2615847280303799, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1363, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.23348986784414286, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.1184, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.2612408310339395, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1392, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.2624637156145282, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1204, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.2800766114957828, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1354, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.25414283692757583, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1161, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.2637008974252742, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1446, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2402179661973949, | |
| "learning_rate": 0.0, | |
| "loss": 0.1216, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 6.874049729748009e+17, | |
| "train_loss": 0.30696559743239327, | |
| "train_runtime": 15599.7333, | |
| "train_samples_per_second": 3.205, | |
| "train_steps_per_second": 0.025 | |
| } | |
| ], | |
| "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": 6.874049729748009e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |