| { | |
| "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": 6.185333324633181, | |
| "learning_rate": 1.0256410256410257e-06, | |
| "loss": 0.9737, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.0256, | |
| "grad_norm": 6.1945774440782655, | |
| "learning_rate": 2.0512820512820513e-06, | |
| "loss": 0.9912, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.0384, | |
| "grad_norm": 6.223501938809448, | |
| "learning_rate": 3.0769230769230774e-06, | |
| "loss": 1.0245, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.0512, | |
| "grad_norm": 5.4144431888891145, | |
| "learning_rate": 4.102564102564103e-06, | |
| "loss": 0.9471, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.064, | |
| "grad_norm": 4.412365404101135, | |
| "learning_rate": 5.128205128205128e-06, | |
| "loss": 0.924, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0768, | |
| "grad_norm": 2.285373080334339, | |
| "learning_rate": 6.153846153846155e-06, | |
| "loss": 0.8638, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 0.0896, | |
| "grad_norm": 1.9637742801474023, | |
| "learning_rate": 7.17948717948718e-06, | |
| "loss": 0.8318, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 0.1024, | |
| "grad_norm": 4.6987012118754965, | |
| "learning_rate": 8.205128205128205e-06, | |
| "loss": 0.8969, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 0.1152, | |
| "grad_norm": 4.7711009077321345, | |
| "learning_rate": 9.230769230769232e-06, | |
| "loss": 0.8692, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 0.128, | |
| "grad_norm": 4.584697561643143, | |
| "learning_rate": 1.0256410256410256e-05, | |
| "loss": 0.8198, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.1408, | |
| "grad_norm": 4.200681545194696, | |
| "learning_rate": 1.1282051282051283e-05, | |
| "loss": 0.8257, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 0.1536, | |
| "grad_norm": 3.39230288406121, | |
| "learning_rate": 1.230769230769231e-05, | |
| "loss": 0.8344, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 0.1664, | |
| "grad_norm": 2.6190174686376912, | |
| "learning_rate": 1.3333333333333333e-05, | |
| "loss": 0.7994, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 0.1792, | |
| "grad_norm": 2.499659214535984, | |
| "learning_rate": 1.435897435897436e-05, | |
| "loss": 0.7838, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 0.192, | |
| "grad_norm": 2.4507415997384014, | |
| "learning_rate": 1.5384615384615387e-05, | |
| "loss": 0.767, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.2048, | |
| "grad_norm": 2.2889249708462267, | |
| "learning_rate": 1.641025641025641e-05, | |
| "loss": 0.7927, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 0.2176, | |
| "grad_norm": 1.7305999661235856, | |
| "learning_rate": 1.7435897435897438e-05, | |
| "loss": 0.7801, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 0.2304, | |
| "grad_norm": 1.6298474108806031, | |
| "learning_rate": 1.8461538461538465e-05, | |
| "loss": 0.7705, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 0.2432, | |
| "grad_norm": 1.5534009848127368, | |
| "learning_rate": 1.9487179487179488e-05, | |
| "loss": 0.742, | |
| "step": 19 | |
| }, | |
| { | |
| "epoch": 0.256, | |
| "grad_norm": 0.909976652304247, | |
| "learning_rate": 2.0512820512820512e-05, | |
| "loss": 0.7244, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.2688, | |
| "grad_norm": 1.2637053044293773, | |
| "learning_rate": 2.153846153846154e-05, | |
| "loss": 0.7519, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 0.2816, | |
| "grad_norm": 1.1709309424629075, | |
| "learning_rate": 2.2564102564102566e-05, | |
| "loss": 0.7267, | |
| "step": 22 | |
| }, | |
| { | |
| "epoch": 0.2944, | |
| "grad_norm": 0.7691595088162027, | |
| "learning_rate": 2.3589743589743593e-05, | |
| "loss": 0.6866, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 0.3072, | |
| "grad_norm": 1.0881124444251495, | |
| "learning_rate": 2.461538461538462e-05, | |
| "loss": 0.7298, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.8477235922479758, | |
| "learning_rate": 2.5641025641025646e-05, | |
| "loss": 0.71, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.3328, | |
| "grad_norm": 1.0721876420763274, | |
| "learning_rate": 2.6666666666666667e-05, | |
| "loss": 0.6829, | |
| "step": 26 | |
| }, | |
| { | |
| "epoch": 0.3456, | |
| "grad_norm": 0.915516947924937, | |
| "learning_rate": 2.7692307692307694e-05, | |
| "loss": 0.6871, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 0.3584, | |
| "grad_norm": 0.7272634912902047, | |
| "learning_rate": 2.871794871794872e-05, | |
| "loss": 0.7112, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 0.3712, | |
| "grad_norm": 0.8799669751396201, | |
| "learning_rate": 2.9743589743589747e-05, | |
| "loss": 0.6907, | |
| "step": 29 | |
| }, | |
| { | |
| "epoch": 0.384, | |
| "grad_norm": 0.9033085740150314, | |
| "learning_rate": 3.0769230769230774e-05, | |
| "loss": 0.683, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.3968, | |
| "grad_norm": 0.7797177277305145, | |
| "learning_rate": 3.1794871794871795e-05, | |
| "loss": 0.6589, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 0.4096, | |
| "grad_norm": 0.7593251968764766, | |
| "learning_rate": 3.282051282051282e-05, | |
| "loss": 0.691, | |
| "step": 32 | |
| }, | |
| { | |
| "epoch": 0.4224, | |
| "grad_norm": 0.7555285034164372, | |
| "learning_rate": 3.384615384615385e-05, | |
| "loss": 0.6474, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 0.4352, | |
| "grad_norm": 0.7736084962659345, | |
| "learning_rate": 3.4871794871794875e-05, | |
| "loss": 0.6898, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 0.448, | |
| "grad_norm": 0.7333425229852718, | |
| "learning_rate": 3.58974358974359e-05, | |
| "loss": 0.6597, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.4608, | |
| "grad_norm": 0.9573288921061265, | |
| "learning_rate": 3.692307692307693e-05, | |
| "loss": 0.6326, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 0.4736, | |
| "grad_norm": 0.8330117835886696, | |
| "learning_rate": 3.794871794871795e-05, | |
| "loss": 0.6529, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 0.4864, | |
| "grad_norm": 0.9650435928465799, | |
| "learning_rate": 3.8974358974358976e-05, | |
| "loss": 0.6656, | |
| "step": 38 | |
| }, | |
| { | |
| "epoch": 0.4992, | |
| "grad_norm": 0.7308809439867704, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6394, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 0.512, | |
| "grad_norm": 0.772489225675158, | |
| "learning_rate": 3.9999198907597046e-05, | |
| "loss": 0.6372, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5248, | |
| "grad_norm": 0.8225659600250838, | |
| "learning_rate": 3.9996795694563096e-05, | |
| "loss": 0.6408, | |
| "step": 41 | |
| }, | |
| { | |
| "epoch": 0.5376, | |
| "grad_norm": 1.1704941197058623, | |
| "learning_rate": 3.999279055341771e-05, | |
| "loss": 0.6479, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 0.5504, | |
| "grad_norm": 0.7688938172286541, | |
| "learning_rate": 3.998718380500971e-05, | |
| "loss": 0.613, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 0.5632, | |
| "grad_norm": 0.7644467787743866, | |
| "learning_rate": 3.997997589849145e-05, | |
| "loss": 0.6645, | |
| "step": 44 | |
| }, | |
| { | |
| "epoch": 0.576, | |
| "grad_norm": 1.0796533930436027, | |
| "learning_rate": 3.9971167411282835e-05, | |
| "loss": 0.6731, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.5888, | |
| "grad_norm": 0.6635670900762584, | |
| "learning_rate": 3.99607590490251e-05, | |
| "loss": 0.6506, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 0.6016, | |
| "grad_norm": 0.9348948092683446, | |
| "learning_rate": 3.9948751645524235e-05, | |
| "loss": 0.6319, | |
| "step": 47 | |
| }, | |
| { | |
| "epoch": 0.6144, | |
| "grad_norm": 0.7102552570269351, | |
| "learning_rate": 3.9935146162684206e-05, | |
| "loss": 0.6494, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 0.6272, | |
| "grad_norm": 0.7333027311998618, | |
| "learning_rate": 3.9919943690429906e-05, | |
| "loss": 0.6069, | |
| "step": 49 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 0.6558597880206618, | |
| "learning_rate": 3.9903145446619837e-05, | |
| "loss": 0.6166, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6528, | |
| "grad_norm": 0.796316918502133, | |
| "learning_rate": 3.9884752776948564e-05, | |
| "loss": 0.643, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 0.6656, | |
| "grad_norm": 0.7074831658562292, | |
| "learning_rate": 3.9864767154838864e-05, | |
| "loss": 0.6394, | |
| "step": 52 | |
| }, | |
| { | |
| "epoch": 0.6784, | |
| "grad_norm": 0.7513148641045853, | |
| "learning_rate": 3.9843190181323744e-05, | |
| "loss": 0.6608, | |
| "step": 53 | |
| }, | |
| { | |
| "epoch": 0.6912, | |
| "grad_norm": 1.0100131696295715, | |
| "learning_rate": 3.982002358491817e-05, | |
| "loss": 0.6714, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 0.704, | |
| "grad_norm": 0.8417766033343338, | |
| "learning_rate": 3.979526922148058e-05, | |
| "loss": 0.6409, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.7168, | |
| "grad_norm": 0.7676794703538521, | |
| "learning_rate": 3.9768929074064206e-05, | |
| "loss": 0.6221, | |
| "step": 56 | |
| }, | |
| { | |
| "epoch": 0.7296, | |
| "grad_norm": 1.0095483592083567, | |
| "learning_rate": 3.9741005252758255e-05, | |
| "loss": 0.6314, | |
| "step": 57 | |
| }, | |
| { | |
| "epoch": 0.7424, | |
| "grad_norm": 0.7906695135585969, | |
| "learning_rate": 3.971149999451886e-05, | |
| "loss": 0.6366, | |
| "step": 58 | |
| }, | |
| { | |
| "epoch": 0.7552, | |
| "grad_norm": 0.608086473476843, | |
| "learning_rate": 3.9680415662989806e-05, | |
| "loss": 0.6139, | |
| "step": 59 | |
| }, | |
| { | |
| "epoch": 0.768, | |
| "grad_norm": 0.9993911293389912, | |
| "learning_rate": 3.9647754748313294e-05, | |
| "loss": 0.6264, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.7808, | |
| "grad_norm": 0.6372308983756567, | |
| "learning_rate": 3.96135198669304e-05, | |
| "loss": 0.6102, | |
| "step": 61 | |
| }, | |
| { | |
| "epoch": 0.7936, | |
| "grad_norm": 0.8081205926237475, | |
| "learning_rate": 3.957771376137144e-05, | |
| "loss": 0.6078, | |
| "step": 62 | |
| }, | |
| { | |
| "epoch": 0.8064, | |
| "grad_norm": 0.7080215534287139, | |
| "learning_rate": 3.954033930003634e-05, | |
| "loss": 0.6274, | |
| "step": 63 | |
| }, | |
| { | |
| "epoch": 0.8192, | |
| "grad_norm": 0.6803619452668841, | |
| "learning_rate": 3.9501399476964806e-05, | |
| "loss": 0.6043, | |
| "step": 64 | |
| }, | |
| { | |
| "epoch": 0.832, | |
| "grad_norm": 0.590256397423044, | |
| "learning_rate": 3.946089741159648e-05, | |
| "loss": 0.6163, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.8448, | |
| "grad_norm": 0.6910981832934187, | |
| "learning_rate": 3.9418836348521045e-05, | |
| "loss": 0.5869, | |
| "step": 66 | |
| }, | |
| { | |
| "epoch": 0.8576, | |
| "grad_norm": 0.6450510323447164, | |
| "learning_rate": 3.937521965721831e-05, | |
| "loss": 0.6112, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 0.8704, | |
| "grad_norm": 0.5267305427416011, | |
| "learning_rate": 3.933005083178828e-05, | |
| "loss": 0.6274, | |
| "step": 68 | |
| }, | |
| { | |
| "epoch": 0.8832, | |
| "grad_norm": 0.5994605831636954, | |
| "learning_rate": 3.928333349067125e-05, | |
| "loss": 0.6152, | |
| "step": 69 | |
| }, | |
| { | |
| "epoch": 0.896, | |
| "grad_norm": 0.5791744550352241, | |
| "learning_rate": 3.923507137635792e-05, | |
| "loss": 0.6465, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9088, | |
| "grad_norm": 0.6539444809440274, | |
| "learning_rate": 3.9185268355089606e-05, | |
| "loss": 0.6089, | |
| "step": 71 | |
| }, | |
| { | |
| "epoch": 0.9216, | |
| "grad_norm": 0.5895792876182698, | |
| "learning_rate": 3.913392841654851e-05, | |
| "loss": 0.5969, | |
| "step": 72 | |
| }, | |
| { | |
| "epoch": 0.9344, | |
| "grad_norm": 0.6749554556089646, | |
| "learning_rate": 3.9081055673538093e-05, | |
| "loss": 0.6427, | |
| "step": 73 | |
| }, | |
| { | |
| "epoch": 0.9472, | |
| "grad_norm": 0.8220067642013372, | |
| "learning_rate": 3.902665436165364e-05, | |
| "loss": 0.5949, | |
| "step": 74 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.5663697701858186, | |
| "learning_rate": 3.897072883894291e-05, | |
| "loss": 0.592, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.9728, | |
| "grad_norm": 0.6609780785777901, | |
| "learning_rate": 3.8913283585557054e-05, | |
| "loss": 0.6089, | |
| "step": 76 | |
| }, | |
| { | |
| "epoch": 0.9856, | |
| "grad_norm": 0.748590011671772, | |
| "learning_rate": 3.885432320339167e-05, | |
| "loss": 0.6513, | |
| "step": 77 | |
| }, | |
| { | |
| "epoch": 0.9984, | |
| "grad_norm": 1.02273797642161, | |
| "learning_rate": 3.879385241571817e-05, | |
| "loss": 0.6463, | |
| "step": 78 | |
| }, | |
| { | |
| "epoch": 1.0112, | |
| "grad_norm": 2.377936577052311, | |
| "learning_rate": 3.873187606680543e-05, | |
| "loss": 1.1577, | |
| "step": 79 | |
| }, | |
| { | |
| "epoch": 1.024, | |
| "grad_norm": 0.621808171404831, | |
| "learning_rate": 3.866839912153168e-05, | |
| "loss": 0.4876, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0368, | |
| "grad_norm": 0.8765798797027711, | |
| "learning_rate": 3.860342666498677e-05, | |
| "loss": 0.5173, | |
| "step": 81 | |
| }, | |
| { | |
| "epoch": 1.0496, | |
| "grad_norm": 1.2613734582204674, | |
| "learning_rate": 3.853696390206484e-05, | |
| "loss": 0.4947, | |
| "step": 82 | |
| }, | |
| { | |
| "epoch": 1.0624, | |
| "grad_norm": 1.4032381011617667, | |
| "learning_rate": 3.846901615704734e-05, | |
| "loss": 0.5052, | |
| "step": 83 | |
| }, | |
| { | |
| "epoch": 1.0752, | |
| "grad_norm": 1.1687977564396996, | |
| "learning_rate": 3.839958887317649e-05, | |
| "loss": 0.5346, | |
| "step": 84 | |
| }, | |
| { | |
| "epoch": 1.088, | |
| "grad_norm": 1.157552624204067, | |
| "learning_rate": 3.832868761221926e-05, | |
| "loss": 0.5389, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.1008, | |
| "grad_norm": 0.8126008532840939, | |
| "learning_rate": 3.825631805402182e-05, | |
| "loss": 0.4734, | |
| "step": 86 | |
| }, | |
| { | |
| "epoch": 1.1136, | |
| "grad_norm": 1.057710759722028, | |
| "learning_rate": 3.818248599605448e-05, | |
| "loss": 0.5369, | |
| "step": 87 | |
| }, | |
| { | |
| "epoch": 1.1264, | |
| "grad_norm": 0.7158375655650051, | |
| "learning_rate": 3.810719735294731e-05, | |
| "loss": 0.5345, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1392, | |
| "grad_norm": 0.8346681480039647, | |
| "learning_rate": 3.8030458156016326e-05, | |
| "loss": 0.5056, | |
| "step": 89 | |
| }, | |
| { | |
| "epoch": 1.152, | |
| "grad_norm": 0.5943794369581732, | |
| "learning_rate": 3.795227455278029e-05, | |
| "loss": 0.4714, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1648, | |
| "grad_norm": 0.6435250676511584, | |
| "learning_rate": 3.787265280646825e-05, | |
| "loss": 0.5027, | |
| "step": 91 | |
| }, | |
| { | |
| "epoch": 1.1776, | |
| "grad_norm": 0.7801501269989074, | |
| "learning_rate": 3.7791599295517825e-05, | |
| "loss": 0.4981, | |
| "step": 92 | |
| }, | |
| { | |
| "epoch": 1.1904, | |
| "grad_norm": 0.5704890419992142, | |
| "learning_rate": 3.7709120513064196e-05, | |
| "loss": 0.4598, | |
| "step": 93 | |
| }, | |
| { | |
| "epoch": 1.2032, | |
| "grad_norm": 0.901042097557887, | |
| "learning_rate": 3.762522306641998e-05, | |
| "loss": 0.5102, | |
| "step": 94 | |
| }, | |
| { | |
| "epoch": 1.216, | |
| "grad_norm": 0.5396386532211707, | |
| "learning_rate": 3.7539913676545874e-05, | |
| "loss": 0.4946, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 1.2288000000000001, | |
| "grad_norm": 0.8815686069713111, | |
| "learning_rate": 3.745319917751229e-05, | |
| "loss": 0.5399, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.2416, | |
| "grad_norm": 0.5932527143714663, | |
| "learning_rate": 3.736508651595188e-05, | |
| "loss": 0.4729, | |
| "step": 97 | |
| }, | |
| { | |
| "epoch": 1.2544, | |
| "grad_norm": 0.598101684595443, | |
| "learning_rate": 3.727558275050301e-05, | |
| "loss": 0.5102, | |
| "step": 98 | |
| }, | |
| { | |
| "epoch": 1.2671999999999999, | |
| "grad_norm": 0.6187773113343685, | |
| "learning_rate": 3.718469505124434e-05, | |
| "loss": 0.5209, | |
| "step": 99 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 0.5881369247632209, | |
| "learning_rate": 3.709243069912041e-05, | |
| "loss": 0.5004, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.2928, | |
| "grad_norm": 0.48969202113975935, | |
| "learning_rate": 3.699879708535838e-05, | |
| "loss": 0.4866, | |
| "step": 101 | |
| }, | |
| { | |
| "epoch": 1.3056, | |
| "grad_norm": 0.6021643586070218, | |
| "learning_rate": 3.69038017108759e-05, | |
| "loss": 0.4828, | |
| "step": 102 | |
| }, | |
| { | |
| "epoch": 1.3184, | |
| "grad_norm": 0.5187009826724566, | |
| "learning_rate": 3.680745218568026e-05, | |
| "loss": 0.5252, | |
| "step": 103 | |
| }, | |
| { | |
| "epoch": 1.3312, | |
| "grad_norm": 0.6002221945380577, | |
| "learning_rate": 3.6709756228258735e-05, | |
| "loss": 0.4942, | |
| "step": 104 | |
| }, | |
| { | |
| "epoch": 1.3439999999999999, | |
| "grad_norm": 0.6064290305259183, | |
| "learning_rate": 3.6610721664960236e-05, | |
| "loss": 0.5293, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 1.3568, | |
| "grad_norm": 0.569375651691498, | |
| "learning_rate": 3.65103564293684e-05, | |
| "loss": 0.4844, | |
| "step": 106 | |
| }, | |
| { | |
| "epoch": 1.3696, | |
| "grad_norm": 0.5515050057158436, | |
| "learning_rate": 3.640866856166601e-05, | |
| "loss": 0.4808, | |
| "step": 107 | |
| }, | |
| { | |
| "epoch": 1.3824, | |
| "grad_norm": 0.581049566341724, | |
| "learning_rate": 3.6305666207990886e-05, | |
| "loss": 0.5185, | |
| "step": 108 | |
| }, | |
| { | |
| "epoch": 1.3952, | |
| "grad_norm": 0.45623008025202666, | |
| "learning_rate": 3.6201357619783336e-05, | |
| "loss": 0.4662, | |
| "step": 109 | |
| }, | |
| { | |
| "epoch": 1.408, | |
| "grad_norm": 0.4967875514055768, | |
| "learning_rate": 3.609575115312511e-05, | |
| "loss": 0.4846, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.4208, | |
| "grad_norm": 0.5071568035344309, | |
| "learning_rate": 3.598885526807003e-05, | |
| "loss": 0.5027, | |
| "step": 111 | |
| }, | |
| { | |
| "epoch": 1.4336, | |
| "grad_norm": 0.48215964154791296, | |
| "learning_rate": 3.5880678527966224e-05, | |
| "loss": 0.4803, | |
| "step": 112 | |
| }, | |
| { | |
| "epoch": 1.4464000000000001, | |
| "grad_norm": 0.5320249466944718, | |
| "learning_rate": 3.577122959877017e-05, | |
| "loss": 0.4636, | |
| "step": 113 | |
| }, | |
| { | |
| "epoch": 1.4592, | |
| "grad_norm": 0.5860223197951501, | |
| "learning_rate": 3.566051724835245e-05, | |
| "loss": 0.4928, | |
| "step": 114 | |
| }, | |
| { | |
| "epoch": 1.472, | |
| "grad_norm": 0.45709471644107164, | |
| "learning_rate": 3.554855034579532e-05, | |
| "loss": 0.4678, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 1.4848, | |
| "grad_norm": 0.5975221302484784, | |
| "learning_rate": 3.5435337860682304e-05, | |
| "loss": 0.5261, | |
| "step": 116 | |
| }, | |
| { | |
| "epoch": 1.4976, | |
| "grad_norm": 0.4904420079297209, | |
| "learning_rate": 3.532088886237956e-05, | |
| "loss": 0.4669, | |
| "step": 117 | |
| }, | |
| { | |
| "epoch": 1.5104, | |
| "grad_norm": 0.6016075834204346, | |
| "learning_rate": 3.520521251930941e-05, | |
| "loss": 0.4869, | |
| "step": 118 | |
| }, | |
| { | |
| "epoch": 1.5232, | |
| "grad_norm": 0.6003716602429013, | |
| "learning_rate": 3.5088318098215805e-05, | |
| "loss": 0.538, | |
| "step": 119 | |
| }, | |
| { | |
| "epoch": 1.536, | |
| "grad_norm": 0.45179630491483885, | |
| "learning_rate": 3.497021496342203e-05, | |
| "loss": 0.478, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.5488, | |
| "grad_norm": 0.6312533084746175, | |
| "learning_rate": 3.485091257608047e-05, | |
| "loss": 0.4976, | |
| "step": 121 | |
| }, | |
| { | |
| "epoch": 1.5615999999999999, | |
| "grad_norm": 0.5468738522922301, | |
| "learning_rate": 3.473042049341474e-05, | |
| "loss": 0.5104, | |
| "step": 122 | |
| }, | |
| { | |
| "epoch": 1.5744, | |
| "grad_norm": 0.5290138315773162, | |
| "learning_rate": 3.4608748367954064e-05, | |
| "loss": 0.4715, | |
| "step": 123 | |
| }, | |
| { | |
| "epoch": 1.5872000000000002, | |
| "grad_norm": 0.6055642144877327, | |
| "learning_rate": 3.4485905946759965e-05, | |
| "loss": 0.5115, | |
| "step": 124 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 0.4956689211193411, | |
| "learning_rate": 3.4361903070645484e-05, | |
| "loss": 0.459, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.6128, | |
| "grad_norm": 0.5379941260937106, | |
| "learning_rate": 3.423674967338681e-05, | |
| "loss": 0.4801, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.6256, | |
| "grad_norm": 0.49846690307976405, | |
| "learning_rate": 3.411045578092754e-05, | |
| "loss": 0.4946, | |
| "step": 127 | |
| }, | |
| { | |
| "epoch": 1.6383999999999999, | |
| "grad_norm": 0.595567053674627, | |
| "learning_rate": 3.398303151057543e-05, | |
| "loss": 0.5039, | |
| "step": 128 | |
| }, | |
| { | |
| "epoch": 1.6512, | |
| "grad_norm": 0.48221791160292277, | |
| "learning_rate": 3.385448707019199e-05, | |
| "loss": 0.5057, | |
| "step": 129 | |
| }, | |
| { | |
| "epoch": 1.6640000000000001, | |
| "grad_norm": 0.5823510592140544, | |
| "learning_rate": 3.372483275737468e-05, | |
| "loss": 0.4759, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6768, | |
| "grad_norm": 0.5004853756273299, | |
| "learning_rate": 3.359407895863199e-05, | |
| "loss": 0.4944, | |
| "step": 131 | |
| }, | |
| { | |
| "epoch": 1.6896, | |
| "grad_norm": 0.46420514589743705, | |
| "learning_rate": 3.34622361485514e-05, | |
| "loss": 0.4706, | |
| "step": 132 | |
| }, | |
| { | |
| "epoch": 1.7024, | |
| "grad_norm": 0.5946355031615173, | |
| "learning_rate": 3.332931488896029e-05, | |
| "loss": 0.4812, | |
| "step": 133 | |
| }, | |
| { | |
| "epoch": 1.7151999999999998, | |
| "grad_norm": 0.5532775880619796, | |
| "learning_rate": 3.319532582807977e-05, | |
| "loss": 0.5023, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 1.728, | |
| "grad_norm": 0.5104791221494308, | |
| "learning_rate": 3.30602796996717e-05, | |
| "loss": 0.5172, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 1.7408000000000001, | |
| "grad_norm": 0.5584122007297562, | |
| "learning_rate": 3.2924187322178865e-05, | |
| "loss": 0.5156, | |
| "step": 136 | |
| }, | |
| { | |
| "epoch": 1.7536, | |
| "grad_norm": 0.49411440039159626, | |
| "learning_rate": 3.278705959785821e-05, | |
| "loss": 0.4953, | |
| "step": 137 | |
| }, | |
| { | |
| "epoch": 1.7664, | |
| "grad_norm": 0.5564743907355297, | |
| "learning_rate": 3.2648907511907544e-05, | |
| "loss": 0.4708, | |
| "step": 138 | |
| }, | |
| { | |
| "epoch": 1.7792, | |
| "grad_norm": 0.5610929598358726, | |
| "learning_rate": 3.250974213158555e-05, | |
| "loss": 0.4976, | |
| "step": 139 | |
| }, | |
| { | |
| "epoch": 1.792, | |
| "grad_norm": 0.5159884426043321, | |
| "learning_rate": 3.23695746053251e-05, | |
| "loss": 0.4952, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.8048, | |
| "grad_norm": 0.5913168395108185, | |
| "learning_rate": 3.222841616184025e-05, | |
| "loss": 0.5272, | |
| "step": 141 | |
| }, | |
| { | |
| "epoch": 1.8176, | |
| "grad_norm": 0.5531214312337782, | |
| "learning_rate": 3.208627810922665e-05, | |
| "loss": 0.4762, | |
| "step": 142 | |
| }, | |
| { | |
| "epoch": 1.8304, | |
| "grad_norm": 0.5089692877560992, | |
| "learning_rate": 3.194317183405573e-05, | |
| "loss": 0.4827, | |
| "step": 143 | |
| }, | |
| { | |
| "epoch": 1.8432, | |
| "grad_norm": 0.5711028287489832, | |
| "learning_rate": 3.1799108800462466e-05, | |
| "loss": 0.4787, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 1.8559999999999999, | |
| "grad_norm": 0.5324663100558089, | |
| "learning_rate": 3.1654100549227024e-05, | |
| "loss": 0.4446, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 1.8688, | |
| "grad_norm": 0.4565377868212234, | |
| "learning_rate": 3.1508158696850275e-05, | |
| "loss": 0.4874, | |
| "step": 146 | |
| }, | |
| { | |
| "epoch": 1.8816000000000002, | |
| "grad_norm": 0.5587044713629004, | |
| "learning_rate": 3.136129493462312e-05, | |
| "loss": 0.514, | |
| "step": 147 | |
| }, | |
| { | |
| "epoch": 1.8944, | |
| "grad_norm": 0.4619094624226731, | |
| "learning_rate": 3.121352102768998e-05, | |
| "loss": 0.4908, | |
| "step": 148 | |
| }, | |
| { | |
| "epoch": 1.9072, | |
| "grad_norm": 0.46202787113546345, | |
| "learning_rate": 3.106484881410628e-05, | |
| "loss": 0.4862, | |
| "step": 149 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 0.46903423730237415, | |
| "learning_rate": 3.091529020389009e-05, | |
| "loss": 0.4912, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9327999999999999, | |
| "grad_norm": 0.4350523722666238, | |
| "learning_rate": 3.076485717806808e-05, | |
| "loss": 0.4973, | |
| "step": 151 | |
| }, | |
| { | |
| "epoch": 1.9456, | |
| "grad_norm": 0.5438177221688146, | |
| "learning_rate": 3.061356178771564e-05, | |
| "loss": 0.4952, | |
| "step": 152 | |
| }, | |
| { | |
| "epoch": 1.9584000000000001, | |
| "grad_norm": 0.44586519968529437, | |
| "learning_rate": 3.0461416152991555e-05, | |
| "loss": 0.4917, | |
| "step": 153 | |
| }, | |
| { | |
| "epoch": 1.9712, | |
| "grad_norm": 0.5280856279630843, | |
| "learning_rate": 3.0308432462167045e-05, | |
| "loss": 0.4691, | |
| "step": 154 | |
| }, | |
| { | |
| "epoch": 1.984, | |
| "grad_norm": 0.4556406462674358, | |
| "learning_rate": 3.015462297064936e-05, | |
| "loss": 0.4547, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 1.9968, | |
| "grad_norm": 0.5305472639384778, | |
| "learning_rate": 3.0000000000000004e-05, | |
| "loss": 0.5012, | |
| "step": 156 | |
| }, | |
| { | |
| "epoch": 2.0096, | |
| "grad_norm": 1.441334211808108, | |
| "learning_rate": 2.98445759369477e-05, | |
| "loss": 0.816, | |
| "step": 157 | |
| }, | |
| { | |
| "epoch": 2.0224, | |
| "grad_norm": 0.6135366697126325, | |
| "learning_rate": 2.9688363232396056e-05, | |
| "loss": 0.3254, | |
| "step": 158 | |
| }, | |
| { | |
| "epoch": 2.0352, | |
| "grad_norm": 1.576249715766764, | |
| "learning_rate": 2.9531374400426158e-05, | |
| "loss": 0.3906, | |
| "step": 159 | |
| }, | |
| { | |
| "epoch": 2.048, | |
| "grad_norm": 0.6961408362993984, | |
| "learning_rate": 2.9373622017294075e-05, | |
| "loss": 0.3556, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.0608, | |
| "grad_norm": 0.6201611938489392, | |
| "learning_rate": 2.9215118720423375e-05, | |
| "loss": 0.3335, | |
| "step": 161 | |
| }, | |
| { | |
| "epoch": 2.0736, | |
| "grad_norm": 0.53858698258229, | |
| "learning_rate": 2.9055877207392752e-05, | |
| "loss": 0.3509, | |
| "step": 162 | |
| }, | |
| { | |
| "epoch": 2.0864, | |
| "grad_norm": 0.6690528315686712, | |
| "learning_rate": 2.8895910234918828e-05, | |
| "loss": 0.3611, | |
| "step": 163 | |
| }, | |
| { | |
| "epoch": 2.0992, | |
| "grad_norm": 0.4572975534608181, | |
| "learning_rate": 2.873523061783426e-05, | |
| "loss": 0.3298, | |
| "step": 164 | |
| }, | |
| { | |
| "epoch": 2.112, | |
| "grad_norm": 0.6702050161260139, | |
| "learning_rate": 2.8573851228061084e-05, | |
| "loss": 0.352, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 2.1248, | |
| "grad_norm": 0.5086664233528277, | |
| "learning_rate": 2.8411784993579633e-05, | |
| "loss": 0.3387, | |
| "step": 166 | |
| }, | |
| { | |
| "epoch": 2.1376, | |
| "grad_norm": 0.5015646311040831, | |
| "learning_rate": 2.8249044897392814e-05, | |
| "loss": 0.356, | |
| "step": 167 | |
| }, | |
| { | |
| "epoch": 2.1504, | |
| "grad_norm": 0.5620733230159155, | |
| "learning_rate": 2.80856439764861e-05, | |
| "loss": 0.3416, | |
| "step": 168 | |
| }, | |
| { | |
| "epoch": 2.1632, | |
| "grad_norm": 0.6991859060247274, | |
| "learning_rate": 2.792159532078314e-05, | |
| "loss": 0.355, | |
| "step": 169 | |
| }, | |
| { | |
| "epoch": 2.176, | |
| "grad_norm": 0.4696751016229829, | |
| "learning_rate": 2.77569120720971e-05, | |
| "loss": 0.3317, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.1888, | |
| "grad_norm": 0.5372659436353339, | |
| "learning_rate": 2.7591607423077932e-05, | |
| "loss": 0.348, | |
| "step": 171 | |
| }, | |
| { | |
| "epoch": 2.2016, | |
| "grad_norm": 0.5256097178479281, | |
| "learning_rate": 2.7425694616155474e-05, | |
| "loss": 0.3587, | |
| "step": 172 | |
| }, | |
| { | |
| "epoch": 2.2144, | |
| "grad_norm": 0.506855230825987, | |
| "learning_rate": 2.7259186942478656e-05, | |
| "loss": 0.3529, | |
| "step": 173 | |
| }, | |
| { | |
| "epoch": 2.2272, | |
| "grad_norm": 0.5899083543684003, | |
| "learning_rate": 2.7092097740850712e-05, | |
| "loss": 0.3446, | |
| "step": 174 | |
| }, | |
| { | |
| "epoch": 2.24, | |
| "grad_norm": 0.4559291904509429, | |
| "learning_rate": 2.692444039666066e-05, | |
| "loss": 0.3452, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.2528, | |
| "grad_norm": 0.4856685575860094, | |
| "learning_rate": 2.6756228340810946e-05, | |
| "loss": 0.3515, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2656, | |
| "grad_norm": 0.4718487921134573, | |
| "learning_rate": 2.6587475048641596e-05, | |
| "loss": 0.344, | |
| "step": 177 | |
| }, | |
| { | |
| "epoch": 2.2784, | |
| "grad_norm": 0.45849572726166243, | |
| "learning_rate": 2.6418194038850634e-05, | |
| "loss": 0.3463, | |
| "step": 178 | |
| }, | |
| { | |
| "epoch": 2.2912, | |
| "grad_norm": 0.44131864184006275, | |
| "learning_rate": 2.624839887241115e-05, | |
| "loss": 0.3192, | |
| "step": 179 | |
| }, | |
| { | |
| "epoch": 2.304, | |
| "grad_norm": 0.5046794244143699, | |
| "learning_rate": 2.607810315148494e-05, | |
| "loss": 0.3869, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.3168, | |
| "grad_norm": 0.43143674929840947, | |
| "learning_rate": 2.5907320518332827e-05, | |
| "loss": 0.3308, | |
| "step": 181 | |
| }, | |
| { | |
| "epoch": 2.3296, | |
| "grad_norm": 0.4749972691797251, | |
| "learning_rate": 2.5736064654221808e-05, | |
| "loss": 0.3586, | |
| "step": 182 | |
| }, | |
| { | |
| "epoch": 2.3424, | |
| "grad_norm": 0.4670742366756154, | |
| "learning_rate": 2.5564349278329056e-05, | |
| "loss": 0.3224, | |
| "step": 183 | |
| }, | |
| { | |
| "epoch": 2.3552, | |
| "grad_norm": 0.6044112762030008, | |
| "learning_rate": 2.539218814664288e-05, | |
| "loss": 0.3505, | |
| "step": 184 | |
| }, | |
| { | |
| "epoch": 2.368, | |
| "grad_norm": 0.46526139315629983, | |
| "learning_rate": 2.521959505086075e-05, | |
| "loss": 0.3438, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 2.3808, | |
| "grad_norm": 0.6674006694022379, | |
| "learning_rate": 2.5046583817284437e-05, | |
| "loss": 0.3474, | |
| "step": 186 | |
| }, | |
| { | |
| "epoch": 2.3936, | |
| "grad_norm": 0.4883758901319177, | |
| "learning_rate": 2.487316830571244e-05, | |
| "loss": 0.357, | |
| "step": 187 | |
| }, | |
| { | |
| "epoch": 2.4064, | |
| "grad_norm": 0.5154832215345102, | |
| "learning_rate": 2.4699362408329646e-05, | |
| "loss": 0.3448, | |
| "step": 188 | |
| }, | |
| { | |
| "epoch": 2.4192, | |
| "grad_norm": 0.47391662694102077, | |
| "learning_rate": 2.4525180048594452e-05, | |
| "loss": 0.3743, | |
| "step": 189 | |
| }, | |
| { | |
| "epoch": 2.432, | |
| "grad_norm": 0.4454005957458192, | |
| "learning_rate": 2.435063518012335e-05, | |
| "loss": 0.3045, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.4448, | |
| "grad_norm": 0.4734146891850675, | |
| "learning_rate": 2.4175741785573177e-05, | |
| "loss": 0.3239, | |
| "step": 191 | |
| }, | |
| { | |
| "epoch": 2.4576000000000002, | |
| "grad_norm": 0.5040934564063406, | |
| "learning_rate": 2.4000513875520892e-05, | |
| "loss": 0.3558, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.4704, | |
| "grad_norm": 0.46554566394351465, | |
| "learning_rate": 2.3824965487341247e-05, | |
| "loss": 0.3545, | |
| "step": 193 | |
| }, | |
| { | |
| "epoch": 2.4832, | |
| "grad_norm": 0.43616190816442785, | |
| "learning_rate": 2.3649110684082258e-05, | |
| "loss": 0.334, | |
| "step": 194 | |
| }, | |
| { | |
| "epoch": 2.496, | |
| "grad_norm": 0.4915934784941002, | |
| "learning_rate": 2.3472963553338614e-05, | |
| "loss": 0.3422, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 2.5088, | |
| "grad_norm": 0.4331620977243421, | |
| "learning_rate": 2.3296538206123134e-05, | |
| "loss": 0.3355, | |
| "step": 196 | |
| }, | |
| { | |
| "epoch": 2.5216, | |
| "grad_norm": 0.46327084842411387, | |
| "learning_rate": 2.311984877573636e-05, | |
| "loss": 0.3429, | |
| "step": 197 | |
| }, | |
| { | |
| "epoch": 2.5343999999999998, | |
| "grad_norm": 0.4450625231886823, | |
| "learning_rate": 2.2942909416634326e-05, | |
| "loss": 0.3379, | |
| "step": 198 | |
| }, | |
| { | |
| "epoch": 2.5472, | |
| "grad_norm": 0.45692682784745403, | |
| "learning_rate": 2.2765734303294666e-05, | |
| "loss": 0.3697, | |
| "step": 199 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 0.4187222658971047, | |
| "learning_rate": 2.2588337629081107e-05, | |
| "loss": 0.3402, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5728, | |
| "grad_norm": 0.43639119580011876, | |
| "learning_rate": 2.2410733605106462e-05, | |
| "loss": 0.3729, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 2.5856, | |
| "grad_norm": 0.45184728777177147, | |
| "learning_rate": 2.2232936459094158e-05, | |
| "loss": 0.3623, | |
| "step": 202 | |
| }, | |
| { | |
| "epoch": 2.5984, | |
| "grad_norm": 0.4295344377689084, | |
| "learning_rate": 2.205496043423849e-05, | |
| "loss": 0.3312, | |
| "step": 203 | |
| }, | |
| { | |
| "epoch": 2.6112, | |
| "grad_norm": 0.3922866346818689, | |
| "learning_rate": 2.1876819788063586e-05, | |
| "loss": 0.3188, | |
| "step": 204 | |
| }, | |
| { | |
| "epoch": 2.624, | |
| "grad_norm": 0.4593243293098376, | |
| "learning_rate": 2.16985287912813e-05, | |
| "loss": 0.3524, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 2.6368, | |
| "grad_norm": 0.42017595638051924, | |
| "learning_rate": 2.1520101726647922e-05, | |
| "loss": 0.3406, | |
| "step": 206 | |
| }, | |
| { | |
| "epoch": 2.6496, | |
| "grad_norm": 0.4732877117660662, | |
| "learning_rate": 2.1341552887820048e-05, | |
| "loss": 0.3739, | |
| "step": 207 | |
| }, | |
| { | |
| "epoch": 2.6624, | |
| "grad_norm": 0.42820051638886547, | |
| "learning_rate": 2.1162896578209517e-05, | |
| "loss": 0.3144, | |
| "step": 208 | |
| }, | |
| { | |
| "epoch": 2.6752000000000002, | |
| "grad_norm": 0.5254867856060992, | |
| "learning_rate": 2.0984147109837564e-05, | |
| "loss": 0.3448, | |
| "step": 209 | |
| }, | |
| { | |
| "epoch": 2.6879999999999997, | |
| "grad_norm": 0.4386987640161933, | |
| "learning_rate": 2.0805318802188307e-05, | |
| "loss": 0.3357, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.7008, | |
| "grad_norm": 0.48510491913807946, | |
| "learning_rate": 2.0626425981061608e-05, | |
| "loss": 0.3336, | |
| "step": 211 | |
| }, | |
| { | |
| "epoch": 2.7136, | |
| "grad_norm": 0.44401009792064033, | |
| "learning_rate": 2.0447482977425465e-05, | |
| "loss": 0.3515, | |
| "step": 212 | |
| }, | |
| { | |
| "epoch": 2.7264, | |
| "grad_norm": 0.5179385151108092, | |
| "learning_rate": 2.0268504126267952e-05, | |
| "loss": 0.3383, | |
| "step": 213 | |
| }, | |
| { | |
| "epoch": 2.7392, | |
| "grad_norm": 0.41670446628081986, | |
| "learning_rate": 2.008950376544887e-05, | |
| "loss": 0.3424, | |
| "step": 214 | |
| }, | |
| { | |
| "epoch": 2.752, | |
| "grad_norm": 0.4788730740853163, | |
| "learning_rate": 1.9910496234551132e-05, | |
| "loss": 0.3475, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 2.7648, | |
| "grad_norm": 0.38646470432956626, | |
| "learning_rate": 1.9731495873732055e-05, | |
| "loss": 0.3266, | |
| "step": 216 | |
| }, | |
| { | |
| "epoch": 2.7776, | |
| "grad_norm": 0.4547668672029085, | |
| "learning_rate": 1.9552517022574542e-05, | |
| "loss": 0.3747, | |
| "step": 217 | |
| }, | |
| { | |
| "epoch": 2.7904, | |
| "grad_norm": 0.4144335813604013, | |
| "learning_rate": 1.93735740189384e-05, | |
| "loss": 0.3526, | |
| "step": 218 | |
| }, | |
| { | |
| "epoch": 2.8032, | |
| "grad_norm": 0.40049454542784196, | |
| "learning_rate": 1.9194681197811703e-05, | |
| "loss": 0.3571, | |
| "step": 219 | |
| }, | |
| { | |
| "epoch": 2.816, | |
| "grad_norm": 0.3946430999248786, | |
| "learning_rate": 1.901585289016244e-05, | |
| "loss": 0.3294, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.8288, | |
| "grad_norm": 0.430523698406278, | |
| "learning_rate": 1.8837103421790486e-05, | |
| "loss": 0.3581, | |
| "step": 221 | |
| }, | |
| { | |
| "epoch": 2.8416, | |
| "grad_norm": 0.3871686995470584, | |
| "learning_rate": 1.8658447112179952e-05, | |
| "loss": 0.3317, | |
| "step": 222 | |
| }, | |
| { | |
| "epoch": 2.8544, | |
| "grad_norm": 0.389874872526401, | |
| "learning_rate": 1.8479898273352084e-05, | |
| "loss": 0.3402, | |
| "step": 223 | |
| }, | |
| { | |
| "epoch": 2.8672, | |
| "grad_norm": 0.43151258638279233, | |
| "learning_rate": 1.83014712087187e-05, | |
| "loss": 0.3916, | |
| "step": 224 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 0.3705732365946505, | |
| "learning_rate": 1.8123180211936417e-05, | |
| "loss": 0.3207, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.8928000000000003, | |
| "grad_norm": 0.3875606484587239, | |
| "learning_rate": 1.794503956576152e-05, | |
| "loss": 0.3468, | |
| "step": 226 | |
| }, | |
| { | |
| "epoch": 2.9055999999999997, | |
| "grad_norm": 0.3941091498937557, | |
| "learning_rate": 1.776706354090585e-05, | |
| "loss": 0.35, | |
| "step": 227 | |
| }, | |
| { | |
| "epoch": 2.9184, | |
| "grad_norm": 0.38083604261343795, | |
| "learning_rate": 1.758926639489354e-05, | |
| "loss": 0.3206, | |
| "step": 228 | |
| }, | |
| { | |
| "epoch": 2.9312, | |
| "grad_norm": 0.36431116757268894, | |
| "learning_rate": 1.7411662370918893e-05, | |
| "loss": 0.3162, | |
| "step": 229 | |
| }, | |
| { | |
| "epoch": 2.944, | |
| "grad_norm": 0.4091406451868388, | |
| "learning_rate": 1.7234265696705344e-05, | |
| "loss": 0.3622, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.9568, | |
| "grad_norm": 0.41350631174485525, | |
| "learning_rate": 1.7057090583365678e-05, | |
| "loss": 0.3511, | |
| "step": 231 | |
| }, | |
| { | |
| "epoch": 2.9696, | |
| "grad_norm": 0.3858002904365008, | |
| "learning_rate": 1.6880151224263646e-05, | |
| "loss": 0.322, | |
| "step": 232 | |
| }, | |
| { | |
| "epoch": 2.9824, | |
| "grad_norm": 0.49033193075740544, | |
| "learning_rate": 1.6703461793876876e-05, | |
| "loss": 0.3846, | |
| "step": 233 | |
| }, | |
| { | |
| "epoch": 2.9952, | |
| "grad_norm": 0.3703699798680849, | |
| "learning_rate": 1.6527036446661396e-05, | |
| "loss": 0.3332, | |
| "step": 234 | |
| }, | |
| { | |
| "epoch": 3.008, | |
| "grad_norm": 1.230903914741192, | |
| "learning_rate": 1.635088931591775e-05, | |
| "loss": 0.5347, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 3.0208, | |
| "grad_norm": 0.5708009670453652, | |
| "learning_rate": 1.6175034512658753e-05, | |
| "loss": 0.2228, | |
| "step": 236 | |
| }, | |
| { | |
| "epoch": 3.0336, | |
| "grad_norm": 0.44431486716717483, | |
| "learning_rate": 1.5999486124479115e-05, | |
| "loss": 0.218, | |
| "step": 237 | |
| }, | |
| { | |
| "epoch": 3.0464, | |
| "grad_norm": 0.5796185847980379, | |
| "learning_rate": 1.5824258214426833e-05, | |
| "loss": 0.2172, | |
| "step": 238 | |
| }, | |
| { | |
| "epoch": 3.0592, | |
| "grad_norm": 0.62430183565173, | |
| "learning_rate": 1.5649364819876655e-05, | |
| "loss": 0.209, | |
| "step": 239 | |
| }, | |
| { | |
| "epoch": 3.072, | |
| "grad_norm": 0.4923092942734261, | |
| "learning_rate": 1.547481995140556e-05, | |
| "loss": 0.2202, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 3.0848, | |
| "grad_norm": 0.45971455029237823, | |
| "learning_rate": 1.5300637591670357e-05, | |
| "loss": 0.2304, | |
| "step": 241 | |
| }, | |
| { | |
| "epoch": 3.0976, | |
| "grad_norm": 0.4491662017980254, | |
| "learning_rate": 1.5126831694287564e-05, | |
| "loss": 0.2288, | |
| "step": 242 | |
| }, | |
| { | |
| "epoch": 3.1104, | |
| "grad_norm": 0.49585254832463616, | |
| "learning_rate": 1.4953416182715566e-05, | |
| "loss": 0.2244, | |
| "step": 243 | |
| }, | |
| { | |
| "epoch": 3.1232, | |
| "grad_norm": 0.41567876376912144, | |
| "learning_rate": 1.478040494913926e-05, | |
| "loss": 0.2192, | |
| "step": 244 | |
| }, | |
| { | |
| "epoch": 3.136, | |
| "grad_norm": 0.4168299784501378, | |
| "learning_rate": 1.460781185335713e-05, | |
| "loss": 0.2367, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 3.1488, | |
| "grad_norm": 0.46653084554853613, | |
| "learning_rate": 1.443565072167095e-05, | |
| "loss": 0.2252, | |
| "step": 246 | |
| }, | |
| { | |
| "epoch": 3.1616, | |
| "grad_norm": 0.44581700205788616, | |
| "learning_rate": 1.4263935345778202e-05, | |
| "loss": 0.2064, | |
| "step": 247 | |
| }, | |
| { | |
| "epoch": 3.1744, | |
| "grad_norm": 0.410954617171449, | |
| "learning_rate": 1.409267948166718e-05, | |
| "loss": 0.2067, | |
| "step": 248 | |
| }, | |
| { | |
| "epoch": 3.1872, | |
| "grad_norm": 0.4311205575718867, | |
| "learning_rate": 1.3921896848515064e-05, | |
| "loss": 0.2213, | |
| "step": 249 | |
| }, | |
| { | |
| "epoch": 3.2, | |
| "grad_norm": 0.42482138835075084, | |
| "learning_rate": 1.3751601127588849e-05, | |
| "loss": 0.2284, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.2128, | |
| "grad_norm": 0.3915313315421235, | |
| "learning_rate": 1.3581805961149371e-05, | |
| "loss": 0.2148, | |
| "step": 251 | |
| }, | |
| { | |
| "epoch": 3.2256, | |
| "grad_norm": 0.40951274571927976, | |
| "learning_rate": 1.341252495135841e-05, | |
| "loss": 0.2054, | |
| "step": 252 | |
| }, | |
| { | |
| "epoch": 3.2384, | |
| "grad_norm": 0.3855252216794962, | |
| "learning_rate": 1.324377165918906e-05, | |
| "loss": 0.2126, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 3.2512, | |
| "grad_norm": 0.36944390789400045, | |
| "learning_rate": 1.3075559603339354e-05, | |
| "loss": 0.2166, | |
| "step": 254 | |
| }, | |
| { | |
| "epoch": 3.2640000000000002, | |
| "grad_norm": 0.43439744560007926, | |
| "learning_rate": 1.2907902259149287e-05, | |
| "loss": 0.2255, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 3.2768, | |
| "grad_norm": 0.40550398110812774, | |
| "learning_rate": 1.274081305752135e-05, | |
| "loss": 0.2192, | |
| "step": 256 | |
| }, | |
| { | |
| "epoch": 3.2896, | |
| "grad_norm": 0.399356280744754, | |
| "learning_rate": 1.2574305383844528e-05, | |
| "loss": 0.2152, | |
| "step": 257 | |
| }, | |
| { | |
| "epoch": 3.3024, | |
| "grad_norm": 0.40095201848528533, | |
| "learning_rate": 1.2408392576922075e-05, | |
| "loss": 0.2157, | |
| "step": 258 | |
| }, | |
| { | |
| "epoch": 3.3152, | |
| "grad_norm": 0.37897554333695616, | |
| "learning_rate": 1.2243087927902905e-05, | |
| "loss": 0.2238, | |
| "step": 259 | |
| }, | |
| { | |
| "epoch": 3.328, | |
| "grad_norm": 0.34753706953786245, | |
| "learning_rate": 1.2078404679216864e-05, | |
| "loss": 0.2131, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.3407999999999998, | |
| "grad_norm": 0.39751115390233716, | |
| "learning_rate": 1.1914356023513904e-05, | |
| "loss": 0.226, | |
| "step": 261 | |
| }, | |
| { | |
| "epoch": 3.3536, | |
| "grad_norm": 0.3453913053137756, | |
| "learning_rate": 1.1750955102607193e-05, | |
| "loss": 0.2094, | |
| "step": 262 | |
| }, | |
| { | |
| "epoch": 3.3664, | |
| "grad_norm": 0.37898648212385433, | |
| "learning_rate": 1.1588215006420374e-05, | |
| "loss": 0.2193, | |
| "step": 263 | |
| }, | |
| { | |
| "epoch": 3.3792, | |
| "grad_norm": 0.3436831780155342, | |
| "learning_rate": 1.1426148771938915e-05, | |
| "loss": 0.2059, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.392, | |
| "grad_norm": 0.3911962630737766, | |
| "learning_rate": 1.1264769382165748e-05, | |
| "loss": 0.2354, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 3.4048, | |
| "grad_norm": 0.38819584240069904, | |
| "learning_rate": 1.110408976508118e-05, | |
| "loss": 0.2256, | |
| "step": 266 | |
| }, | |
| { | |
| "epoch": 3.4176, | |
| "grad_norm": 0.33867993395377527, | |
| "learning_rate": 1.094412279260726e-05, | |
| "loss": 0.203, | |
| "step": 267 | |
| }, | |
| { | |
| "epoch": 3.4304, | |
| "grad_norm": 0.34479724518922344, | |
| "learning_rate": 1.0784881279576635e-05, | |
| "loss": 0.2075, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 3.4432, | |
| "grad_norm": 0.3621979626768828, | |
| "learning_rate": 1.0626377982705929e-05, | |
| "loss": 0.2386, | |
| "step": 269 | |
| }, | |
| { | |
| "epoch": 3.456, | |
| "grad_norm": 0.333590058012057, | |
| "learning_rate": 1.0468625599573842e-05, | |
| "loss": 0.2134, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.4688, | |
| "grad_norm": 0.36614435788477756, | |
| "learning_rate": 1.0311636767603952e-05, | |
| "loss": 0.23, | |
| "step": 271 | |
| }, | |
| { | |
| "epoch": 3.4816, | |
| "grad_norm": 0.35011977693548507, | |
| "learning_rate": 1.0155424063052306e-05, | |
| "loss": 0.2183, | |
| "step": 272 | |
| }, | |
| { | |
| "epoch": 3.4944, | |
| "grad_norm": 0.352337867931405, | |
| "learning_rate": 1.0000000000000006e-05, | |
| "loss": 0.227, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 3.5072, | |
| "grad_norm": 0.3470042336285642, | |
| "learning_rate": 9.84537702935065e-06, | |
| "loss": 0.2063, | |
| "step": 274 | |
| }, | |
| { | |
| "epoch": 3.52, | |
| "grad_norm": 0.3555636785757053, | |
| "learning_rate": 9.691567537832964e-06, | |
| "loss": 0.2344, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.5328, | |
| "grad_norm": 0.3318740203890626, | |
| "learning_rate": 9.538583847008452e-06, | |
| "loss": 0.2088, | |
| "step": 276 | |
| }, | |
| { | |
| "epoch": 3.5456, | |
| "grad_norm": 0.3512135129234243, | |
| "learning_rate": 9.386438212284372e-06, | |
| "loss": 0.1917, | |
| "step": 277 | |
| }, | |
| { | |
| "epoch": 3.5584, | |
| "grad_norm": 0.3678735366335007, | |
| "learning_rate": 9.235142821931928e-06, | |
| "loss": 0.2287, | |
| "step": 278 | |
| }, | |
| { | |
| "epoch": 3.5712, | |
| "grad_norm": 0.40908577940209195, | |
| "learning_rate": 9.084709796109907e-06, | |
| "loss": 0.2161, | |
| "step": 279 | |
| }, | |
| { | |
| "epoch": 3.584, | |
| "grad_norm": 0.32455849576712914, | |
| "learning_rate": 8.93515118589373e-06, | |
| "loss": 0.199, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.5968, | |
| "grad_norm": 0.3385736184642146, | |
| "learning_rate": 8.786478972310023e-06, | |
| "loss": 0.214, | |
| "step": 281 | |
| }, | |
| { | |
| "epoch": 3.6096, | |
| "grad_norm": 0.34041214788404117, | |
| "learning_rate": 8.638705065376887e-06, | |
| "loss": 0.2166, | |
| "step": 282 | |
| }, | |
| { | |
| "epoch": 3.6224, | |
| "grad_norm": 0.3345041927540392, | |
| "learning_rate": 8.491841303149728e-06, | |
| "loss": 0.2096, | |
| "step": 283 | |
| }, | |
| { | |
| "epoch": 3.6352, | |
| "grad_norm": 0.33940661686748747, | |
| "learning_rate": 8.345899450772975e-06, | |
| "loss": 0.2218, | |
| "step": 284 | |
| }, | |
| { | |
| "epoch": 3.648, | |
| "grad_norm": 0.3235611881978904, | |
| "learning_rate": 8.200891199537549e-06, | |
| "loss": 0.2109, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 3.6608, | |
| "grad_norm": 0.33666041025505844, | |
| "learning_rate": 8.056828165944282e-06, | |
| "loss": 0.2108, | |
| "step": 286 | |
| }, | |
| { | |
| "epoch": 3.6736, | |
| "grad_norm": 0.34236636835881373, | |
| "learning_rate": 7.913721890773354e-06, | |
| "loss": 0.217, | |
| "step": 287 | |
| }, | |
| { | |
| "epoch": 3.6864, | |
| "grad_norm": 0.36378089560932597, | |
| "learning_rate": 7.771583838159756e-06, | |
| "loss": 0.2353, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.6992000000000003, | |
| "grad_norm": 0.3217659166399569, | |
| "learning_rate": 7.630425394674903e-06, | |
| "loss": 0.1991, | |
| "step": 289 | |
| }, | |
| { | |
| "epoch": 3.7119999999999997, | |
| "grad_norm": 0.34169339094610307, | |
| "learning_rate": 7.49025786841445e-06, | |
| "loss": 0.2156, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.7248, | |
| "grad_norm": 0.3332009871859823, | |
| "learning_rate": 7.3510924880924575e-06, | |
| "loss": 0.2012, | |
| "step": 291 | |
| }, | |
| { | |
| "epoch": 3.7376, | |
| "grad_norm": 0.342376772616765, | |
| "learning_rate": 7.212940402141808e-06, | |
| "loss": 0.2127, | |
| "step": 292 | |
| }, | |
| { | |
| "epoch": 3.7504, | |
| "grad_norm": 0.3530503765773409, | |
| "learning_rate": 7.075812677821145e-06, | |
| "loss": 0.218, | |
| "step": 293 | |
| }, | |
| { | |
| "epoch": 3.7632, | |
| "grad_norm": 0.3495062555255379, | |
| "learning_rate": 6.939720300328303e-06, | |
| "loss": 0.2076, | |
| "step": 294 | |
| }, | |
| { | |
| "epoch": 3.776, | |
| "grad_norm": 0.3288280777916389, | |
| "learning_rate": 6.8046741719202385e-06, | |
| "loss": 0.2073, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 3.7888, | |
| "grad_norm": 0.3563623359468322, | |
| "learning_rate": 6.67068511103971e-06, | |
| "loss": 0.2157, | |
| "step": 296 | |
| }, | |
| { | |
| "epoch": 3.8016, | |
| "grad_norm": 0.34771198329587544, | |
| "learning_rate": 6.537763851448593e-06, | |
| "loss": 0.2207, | |
| "step": 297 | |
| }, | |
| { | |
| "epoch": 3.8144, | |
| "grad_norm": 0.3338059137994478, | |
| "learning_rate": 6.4059210413680175e-06, | |
| "loss": 0.2164, | |
| "step": 298 | |
| }, | |
| { | |
| "epoch": 3.8272, | |
| "grad_norm": 0.34068667171170985, | |
| "learning_rate": 6.275167242625331e-06, | |
| "loss": 0.2027, | |
| "step": 299 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 0.35747176314868045, | |
| "learning_rate": 6.145512929808013e-06, | |
| "loss": 0.2196, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.8528000000000002, | |
| "grad_norm": 0.33626944060283187, | |
| "learning_rate": 6.016968489424572e-06, | |
| "loss": 0.2069, | |
| "step": 301 | |
| }, | |
| { | |
| "epoch": 3.8656, | |
| "grad_norm": 0.3378151951185304, | |
| "learning_rate": 5.889544219072465e-06, | |
| "loss": 0.2184, | |
| "step": 302 | |
| }, | |
| { | |
| "epoch": 3.8784, | |
| "grad_norm": 0.33816821636992045, | |
| "learning_rate": 5.7632503266131925e-06, | |
| "loss": 0.1992, | |
| "step": 303 | |
| }, | |
| { | |
| "epoch": 3.8912, | |
| "grad_norm": 0.34822739687287807, | |
| "learning_rate": 5.638096929354522e-06, | |
| "loss": 0.204, | |
| "step": 304 | |
| }, | |
| { | |
| "epoch": 3.904, | |
| "grad_norm": 0.3380262582356008, | |
| "learning_rate": 5.514094053240035e-06, | |
| "loss": 0.2093, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 3.9168, | |
| "grad_norm": 0.3572728372003161, | |
| "learning_rate": 5.39125163204594e-06, | |
| "loss": 0.2134, | |
| "step": 306 | |
| }, | |
| { | |
| "epoch": 3.9295999999999998, | |
| "grad_norm": 0.3506635709403538, | |
| "learning_rate": 5.269579506585259e-06, | |
| "loss": 0.2158, | |
| "step": 307 | |
| }, | |
| { | |
| "epoch": 3.9424, | |
| "grad_norm": 0.3271672206718031, | |
| "learning_rate": 5.149087423919541e-06, | |
| "loss": 0.2207, | |
| "step": 308 | |
| }, | |
| { | |
| "epoch": 3.9552, | |
| "grad_norm": 0.3530251376147923, | |
| "learning_rate": 5.029785036577976e-06, | |
| "loss": 0.2396, | |
| "step": 309 | |
| }, | |
| { | |
| "epoch": 3.968, | |
| "grad_norm": 0.3063813092424033, | |
| "learning_rate": 4.911681901784198e-06, | |
| "loss": 0.1792, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.9808, | |
| "grad_norm": 0.34319703479602753, | |
| "learning_rate": 4.794787480690597e-06, | |
| "loss": 0.2021, | |
| "step": 311 | |
| }, | |
| { | |
| "epoch": 3.9936, | |
| "grad_norm": 0.3265711987586607, | |
| "learning_rate": 4.679111137620442e-06, | |
| "loss": 0.1974, | |
| "step": 312 | |
| }, | |
| { | |
| "epoch": 4.0064, | |
| "grad_norm": 1.1663848243182808, | |
| "learning_rate": 4.5646621393177e-06, | |
| "loss": 0.3568, | |
| "step": 313 | |
| }, | |
| { | |
| "epoch": 4.0192, | |
| "grad_norm": 0.5012108815811315, | |
| "learning_rate": 4.451449654204685e-06, | |
| "loss": 0.1514, | |
| "step": 314 | |
| }, | |
| { | |
| "epoch": 4.032, | |
| "grad_norm": 0.4627452408631312, | |
| "learning_rate": 4.339482751647557e-06, | |
| "loss": 0.1645, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 4.0448, | |
| "grad_norm": 0.3315257284629698, | |
| "learning_rate": 4.228770401229824e-06, | |
| "loss": 0.1356, | |
| "step": 316 | |
| }, | |
| { | |
| "epoch": 4.0576, | |
| "grad_norm": 0.29659518925052164, | |
| "learning_rate": 4.119321472033779e-06, | |
| "loss": 0.1304, | |
| "step": 317 | |
| }, | |
| { | |
| "epoch": 4.0704, | |
| "grad_norm": 0.39787830390863205, | |
| "learning_rate": 4.011144731929981e-06, | |
| "loss": 0.1601, | |
| "step": 318 | |
| }, | |
| { | |
| "epoch": 4.0832, | |
| "grad_norm": 0.40193785167713025, | |
| "learning_rate": 3.904248846874894e-06, | |
| "loss": 0.1288, | |
| "step": 319 | |
| }, | |
| { | |
| "epoch": 4.096, | |
| "grad_norm": 0.5559029348058031, | |
| "learning_rate": 3.7986423802166705e-06, | |
| "loss": 0.1643, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 4.1088, | |
| "grad_norm": 0.47273936738632644, | |
| "learning_rate": 3.694333792009115e-06, | |
| "loss": 0.1535, | |
| "step": 321 | |
| }, | |
| { | |
| "epoch": 4.1216, | |
| "grad_norm": 0.3925352619413244, | |
| "learning_rate": 3.5913314383339937e-06, | |
| "loss": 0.1396, | |
| "step": 322 | |
| }, | |
| { | |
| "epoch": 4.1344, | |
| "grad_norm": 0.35492502175186963, | |
| "learning_rate": 3.4896435706316e-06, | |
| "loss": 0.1487, | |
| "step": 323 | |
| }, | |
| { | |
| "epoch": 4.1472, | |
| "grad_norm": 0.3212797317279305, | |
| "learning_rate": 3.3892783350397675e-06, | |
| "loss": 0.1398, | |
| "step": 324 | |
| }, | |
| { | |
| "epoch": 4.16, | |
| "grad_norm": 0.3478601948338465, | |
| "learning_rate": 3.290243771741275e-06, | |
| "loss": 0.1511, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 4.1728, | |
| "grad_norm": 0.3442076788752203, | |
| "learning_rate": 3.1925478143197418e-06, | |
| "loss": 0.134, | |
| "step": 326 | |
| }, | |
| { | |
| "epoch": 4.1856, | |
| "grad_norm": 0.3436759382136139, | |
| "learning_rate": 3.0961982891241083e-06, | |
| "loss": 0.1352, | |
| "step": 327 | |
| }, | |
| { | |
| "epoch": 4.1984, | |
| "grad_norm": 0.3287448723541899, | |
| "learning_rate": 3.001202914641628e-06, | |
| "loss": 0.1447, | |
| "step": 328 | |
| }, | |
| { | |
| "epoch": 4.2112, | |
| "grad_norm": 0.3251611260369395, | |
| "learning_rate": 2.907569300879596e-06, | |
| "loss": 0.1454, | |
| "step": 329 | |
| }, | |
| { | |
| "epoch": 4.224, | |
| "grad_norm": 0.3033791451397303, | |
| "learning_rate": 2.815304948755664e-06, | |
| "loss": 0.1257, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 4.2368, | |
| "grad_norm": 0.34224127243350305, | |
| "learning_rate": 2.7244172494969978e-06, | |
| "loss": 0.1634, | |
| "step": 331 | |
| }, | |
| { | |
| "epoch": 4.2496, | |
| "grad_norm": 0.31550478367570267, | |
| "learning_rate": 2.6349134840481294e-06, | |
| "loss": 0.1336, | |
| "step": 332 | |
| }, | |
| { | |
| "epoch": 4.2624, | |
| "grad_norm": 0.3155964103820071, | |
| "learning_rate": 2.546800822487714e-06, | |
| "loss": 0.1465, | |
| "step": 333 | |
| }, | |
| { | |
| "epoch": 4.2752, | |
| "grad_norm": 0.3192649341062579, | |
| "learning_rate": 2.4600863234541338e-06, | |
| "loss": 0.1509, | |
| "step": 334 | |
| }, | |
| { | |
| "epoch": 4.288, | |
| "grad_norm": 0.3103609369701921, | |
| "learning_rate": 2.374776933580025e-06, | |
| "loss": 0.1345, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 4.3008, | |
| "grad_norm": 0.3045591981524303, | |
| "learning_rate": 2.2908794869358044e-06, | |
| "loss": 0.1357, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 4.3136, | |
| "grad_norm": 0.298679432334344, | |
| "learning_rate": 2.2084007044821764e-06, | |
| "loss": 0.1412, | |
| "step": 337 | |
| }, | |
| { | |
| "epoch": 4.3264, | |
| "grad_norm": 0.3051143086286374, | |
| "learning_rate": 2.127347193531757e-06, | |
| "loss": 0.132, | |
| "step": 338 | |
| }, | |
| { | |
| "epoch": 4.3392, | |
| "grad_norm": 0.3046516139989146, | |
| "learning_rate": 2.0477254472197237e-06, | |
| "loss": 0.1377, | |
| "step": 339 | |
| }, | |
| { | |
| "epoch": 4.352, | |
| "grad_norm": 0.3116439229496776, | |
| "learning_rate": 1.96954184398368e-06, | |
| "loss": 0.15, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.3648, | |
| "grad_norm": 0.3075735363203657, | |
| "learning_rate": 1.8928026470526917e-06, | |
| "loss": 0.1459, | |
| "step": 341 | |
| }, | |
| { | |
| "epoch": 4.3776, | |
| "grad_norm": 0.3080971044824554, | |
| "learning_rate": 1.817514003945524e-06, | |
| "loss": 0.1358, | |
| "step": 342 | |
| }, | |
| { | |
| "epoch": 4.3904, | |
| "grad_norm": 0.2993405159427207, | |
| "learning_rate": 1.743681945978184e-06, | |
| "loss": 0.1507, | |
| "step": 343 | |
| }, | |
| { | |
| "epoch": 4.4032, | |
| "grad_norm": 0.30329528322920335, | |
| "learning_rate": 1.6713123877807413e-06, | |
| "loss": 0.1494, | |
| "step": 344 | |
| }, | |
| { | |
| "epoch": 4.416, | |
| "grad_norm": 0.2725185367335165, | |
| "learning_rate": 1.6004111268235156e-06, | |
| "loss": 0.1124, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 4.4288, | |
| "grad_norm": 0.3141824767703153, | |
| "learning_rate": 1.5309838429526714e-06, | |
| "loss": 0.1573, | |
| "step": 346 | |
| }, | |
| { | |
| "epoch": 4.4416, | |
| "grad_norm": 0.31213841612769094, | |
| "learning_rate": 1.4630360979351644e-06, | |
| "loss": 0.1549, | |
| "step": 347 | |
| }, | |
| { | |
| "epoch": 4.4544, | |
| "grad_norm": 0.27542521478625676, | |
| "learning_rate": 1.396573335013236e-06, | |
| "loss": 0.1299, | |
| "step": 348 | |
| }, | |
| { | |
| "epoch": 4.4672, | |
| "grad_norm": 0.30358273878093844, | |
| "learning_rate": 1.3316008784683265e-06, | |
| "loss": 0.1486, | |
| "step": 349 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 0.28785416145568044, | |
| "learning_rate": 1.2681239331945695e-06, | |
| "loss": 0.137, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.4928, | |
| "grad_norm": 0.2846655326585656, | |
| "learning_rate": 1.2061475842818337e-06, | |
| "loss": 0.1348, | |
| "step": 351 | |
| }, | |
| { | |
| "epoch": 4.5056, | |
| "grad_norm": 0.3071227038877262, | |
| "learning_rate": 1.1456767966083393e-06, | |
| "loss": 0.1524, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.5184, | |
| "grad_norm": 0.3095295948514431, | |
| "learning_rate": 1.086716414442952e-06, | |
| "loss": 0.1433, | |
| "step": 353 | |
| }, | |
| { | |
| "epoch": 4.5312, | |
| "grad_norm": 0.3101783611114451, | |
| "learning_rate": 1.0292711610570904e-06, | |
| "loss": 0.1387, | |
| "step": 354 | |
| }, | |
| { | |
| "epoch": 4.5440000000000005, | |
| "grad_norm": 0.2943461022126049, | |
| "learning_rate": 9.733456383463658e-07, | |
| "loss": 0.1353, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 4.5568, | |
| "grad_norm": 0.28357983291035677, | |
| "learning_rate": 9.189443264619102e-07, | |
| "loss": 0.1284, | |
| "step": 356 | |
| }, | |
| { | |
| "epoch": 4.5696, | |
| "grad_norm": 0.29002308746331795, | |
| "learning_rate": 8.660715834514977e-07, | |
| "loss": 0.152, | |
| "step": 357 | |
| }, | |
| { | |
| "epoch": 4.5824, | |
| "grad_norm": 0.2958114628757621, | |
| "learning_rate": 8.147316449103959e-07, | |
| "loss": 0.1526, | |
| "step": 358 | |
| }, | |
| { | |
| "epoch": 4.5952, | |
| "grad_norm": 0.30413065887006424, | |
| "learning_rate": 7.649286236420806e-07, | |
| "loss": 0.1429, | |
| "step": 359 | |
| }, | |
| { | |
| "epoch": 4.608, | |
| "grad_norm": 0.30231816724465405, | |
| "learning_rate": 7.166665093287539e-07, | |
| "loss": 0.1467, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.6208, | |
| "grad_norm": 0.2933981463411811, | |
| "learning_rate": 6.69949168211721e-07, | |
| "loss": 0.152, | |
| "step": 361 | |
| }, | |
| { | |
| "epoch": 4.6336, | |
| "grad_norm": 0.2785997520502762, | |
| "learning_rate": 6.247803427816945e-07, | |
| "loss": 0.1175, | |
| "step": 362 | |
| }, | |
| { | |
| "epoch": 4.6464, | |
| "grad_norm": 0.2860088341971988, | |
| "learning_rate": 5.811636514789598e-07, | |
| "loss": 0.1379, | |
| "step": 363 | |
| }, | |
| { | |
| "epoch": 4.6592, | |
| "grad_norm": 0.28102886709918307, | |
| "learning_rate": 5.391025884035239e-07, | |
| "loss": 0.1487, | |
| "step": 364 | |
| }, | |
| { | |
| "epoch": 4.672, | |
| "grad_norm": 0.3009861091752357, | |
| "learning_rate": 4.986005230351954e-07, | |
| "loss": 0.1436, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 4.6848, | |
| "grad_norm": 0.3011602330302086, | |
| "learning_rate": 4.5966069996365993e-07, | |
| "loss": 0.1291, | |
| "step": 366 | |
| }, | |
| { | |
| "epoch": 4.6975999999999996, | |
| "grad_norm": 0.3180175753768788, | |
| "learning_rate": 4.22286238628562e-07, | |
| "loss": 0.1442, | |
| "step": 367 | |
| }, | |
| { | |
| "epoch": 4.7104, | |
| "grad_norm": 0.301252561581282, | |
| "learning_rate": 3.8648013306960664e-07, | |
| "loss": 0.1316, | |
| "step": 368 | |
| }, | |
| { | |
| "epoch": 4.7232, | |
| "grad_norm": 0.30470484171104517, | |
| "learning_rate": 3.522452516867048e-07, | |
| "loss": 0.1403, | |
| "step": 369 | |
| }, | |
| { | |
| "epoch": 4.736, | |
| "grad_norm": 0.30462722669203124, | |
| "learning_rate": 3.1958433701019697e-07, | |
| "loss": 0.1381, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.7488, | |
| "grad_norm": 0.2960420686249916, | |
| "learning_rate": 2.8850000548115155e-07, | |
| "loss": 0.1375, | |
| "step": 371 | |
| }, | |
| { | |
| "epoch": 4.7616, | |
| "grad_norm": 0.28269403580010416, | |
| "learning_rate": 2.5899474724174313e-07, | |
| "loss": 0.1489, | |
| "step": 372 | |
| }, | |
| { | |
| "epoch": 4.7744, | |
| "grad_norm": 0.28108135703491965, | |
| "learning_rate": 2.3107092593579905e-07, | |
| "loss": 0.1458, | |
| "step": 373 | |
| }, | |
| { | |
| "epoch": 4.7872, | |
| "grad_norm": 0.28814494926625883, | |
| "learning_rate": 2.0473077851942858e-07, | |
| "loss": 0.1606, | |
| "step": 374 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 0.27165621319411104, | |
| "learning_rate": 1.799764150818306e-07, | |
| "loss": 0.1226, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.8128, | |
| "grad_norm": 0.27701832507225804, | |
| "learning_rate": 1.5680981867625566e-07, | |
| "loss": 0.1275, | |
| "step": 376 | |
| }, | |
| { | |
| "epoch": 4.8256, | |
| "grad_norm": 0.30647991698742894, | |
| "learning_rate": 1.3523284516113955e-07, | |
| "loss": 0.1429, | |
| "step": 377 | |
| }, | |
| { | |
| "epoch": 4.8384, | |
| "grad_norm": 0.3080395862683008, | |
| "learning_rate": 1.1524722305144231e-07, | |
| "loss": 0.1433, | |
| "step": 378 | |
| }, | |
| { | |
| "epoch": 4.8512, | |
| "grad_norm": 0.2749633367189585, | |
| "learning_rate": 9.685455338016347e-08, | |
| "loss": 0.1247, | |
| "step": 379 | |
| }, | |
| { | |
| "epoch": 4.864, | |
| "grad_norm": 0.30041736648492207, | |
| "learning_rate": 8.005630957010014e-08, | |
| "loss": 0.156, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.8768, | |
| "grad_norm": 0.30848394096505344, | |
| "learning_rate": 6.485383731580142e-08, | |
| "loss": 0.1529, | |
| "step": 381 | |
| }, | |
| { | |
| "epoch": 4.8896, | |
| "grad_norm": 0.2765668001469762, | |
| "learning_rate": 5.1248354475768034e-08, | |
| "loss": 0.1245, | |
| "step": 382 | |
| }, | |
| { | |
| "epoch": 4.9024, | |
| "grad_norm": 0.30913637432581675, | |
| "learning_rate": 3.924095097489922e-08, | |
| "loss": 0.1647, | |
| "step": 383 | |
| }, | |
| { | |
| "epoch": 4.9152000000000005, | |
| "grad_norm": 0.27961246372761445, | |
| "learning_rate": 2.8832588717164766e-08, | |
| "loss": 0.13, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.928, | |
| "grad_norm": 0.30398567130794335, | |
| "learning_rate": 2.0024101508555604e-08, | |
| "loss": 0.1428, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 4.9408, | |
| "grad_norm": 0.3035472609850489, | |
| "learning_rate": 1.281619499029274e-08, | |
| "loss": 0.1388, | |
| "step": 386 | |
| }, | |
| { | |
| "epoch": 4.9536, | |
| "grad_norm": 0.27295459982586584, | |
| "learning_rate": 7.209446582292501e-09, | |
| "loss": 0.1249, | |
| "step": 387 | |
| }, | |
| { | |
| "epoch": 4.9664, | |
| "grad_norm": 0.2851259937602045, | |
| "learning_rate": 3.2043054369057523e-09, | |
| "loss": 0.1393, | |
| "step": 388 | |
| }, | |
| { | |
| "epoch": 4.9792, | |
| "grad_norm": 0.29349398820559053, | |
| "learning_rate": 8.010924029533406e-10, | |
| "loss": 0.1328, | |
| "step": 389 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "grad_norm": 0.2694413970215221, | |
| "learning_rate": 0.0, | |
| "loss": 0.1075, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.992, | |
| "step": 390, | |
| "total_flos": 6.781150607693578e+17, | |
| "train_loss": 0.38343327595637394, | |
| "train_runtime": 12509.8929, | |
| "train_samples_per_second": 3.997, | |
| "train_steps_per_second": 0.031 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 390, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6.781150607693578e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |