| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.9100676482292083, | |
| "eval_steps": 200, | |
| "global_step": 4800, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.03979307600477517, | |
| "grad_norm": 4.196262359619141, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.7988, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.07958615200955034, | |
| "grad_norm": 1.3701090812683105, | |
| "learning_rate": 5e-05, | |
| "loss": 0.6726, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.07958615200955034, | |
| "eval_accuracy": 0.571913002806361, | |
| "eval_loss": 0.6743200421333313, | |
| "eval_runtime": 70.7503, | |
| "eval_samples_per_second": 120.876, | |
| "eval_steps_per_second": 1.894, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.1193792280143255, | |
| "grad_norm": 1.1724385023117065, | |
| "learning_rate": 4.994704807509721e-05, | |
| "loss": 0.6713, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.15917230401910068, | |
| "grad_norm": 2.7935872077941895, | |
| "learning_rate": 4.9788416612896896e-05, | |
| "loss": 0.665, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.15917230401910068, | |
| "eval_accuracy": 0.5989242282507016, | |
| "eval_loss": 0.6570606231689453, | |
| "eval_runtime": 70.7356, | |
| "eval_samples_per_second": 120.901, | |
| "eval_steps_per_second": 1.894, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.19896538002387584, | |
| "grad_norm": 2.3964438438415527, | |
| "learning_rate": 4.952477760070096e-05, | |
| "loss": 0.6614, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.238758456028651, | |
| "grad_norm": 2.5537328720092773, | |
| "learning_rate": 4.9157247853963424e-05, | |
| "loss": 0.6644, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.238758456028651, | |
| "eval_accuracy": 0.5892188961646398, | |
| "eval_loss": 0.6891445517539978, | |
| "eval_runtime": 70.7814, | |
| "eval_samples_per_second": 120.823, | |
| "eval_steps_per_second": 1.893, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.2785515320334262, | |
| "grad_norm": 6.364818572998047, | |
| "learning_rate": 4.8687384285288185e-05, | |
| "loss": 0.6698, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.31834460803820136, | |
| "grad_norm": 5.387712478637695, | |
| "learning_rate": 4.811717730910749e-05, | |
| "loss": 0.6665, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.31834460803820136, | |
| "eval_accuracy": 0.5938961646398503, | |
| "eval_loss": 0.66575688123703, | |
| "eval_runtime": 70.8533, | |
| "eval_samples_per_second": 120.7, | |
| "eval_steps_per_second": 1.891, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.3581376840429765, | |
| "grad_norm": 2.7228779792785645, | |
| "learning_rate": 4.744904240997987e-05, | |
| "loss": 0.674, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.3979307600477517, | |
| "grad_norm": 5.516107082366943, | |
| "learning_rate": 4.668580991022563e-05, | |
| "loss": 0.6545, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.3979307600477517, | |
| "eval_accuracy": 0.5989242282507016, | |
| "eval_loss": 0.663463294506073, | |
| "eval_runtime": 70.5961, | |
| "eval_samples_per_second": 121.14, | |
| "eval_steps_per_second": 1.898, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.43772383605252685, | |
| "grad_norm": 1.8297033309936523, | |
| "learning_rate": 4.5830712980245576e-05, | |
| "loss": 0.6549, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.477516912057302, | |
| "grad_norm": 1.0240168571472168, | |
| "learning_rate": 4.48873739423134e-05, | |
| "loss": 0.6797, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.477516912057302, | |
| "eval_accuracy": 0.5445509822263798, | |
| "eval_loss": 0.688525915145874, | |
| "eval_runtime": 70.7499, | |
| "eval_samples_per_second": 120.876, | |
| "eval_steps_per_second": 1.894, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.5173099880620772, | |
| "grad_norm": 1.1778732538223267, | |
| "learning_rate": 4.385978892586064e-05, | |
| "loss": 0.6805, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.5571030640668524, | |
| "grad_norm": 1.2296079397201538, | |
| "learning_rate": 4.275231093925711e-05, | |
| "loss": 0.6722, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5571030640668524, | |
| "eval_accuracy": 0.5915575304022451, | |
| "eval_loss": 0.6639552116394043, | |
| "eval_runtime": 71.55, | |
| "eval_samples_per_second": 119.525, | |
| "eval_steps_per_second": 1.873, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.5968961400716275, | |
| "grad_norm": 5.959015369415283, | |
| "learning_rate": 4.1569631429797054e-05, | |
| "loss": 0.6624, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.6366892160764027, | |
| "grad_norm": 4.830468654632568, | |
| "learning_rate": 4.031676041000599e-05, | |
| "loss": 0.6536, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6366892160764027, | |
| "eval_accuracy": 0.6013797942001871, | |
| "eval_loss": 0.6594148278236389, | |
| "eval_runtime": 71.4819, | |
| "eval_samples_per_second": 119.639, | |
| "eval_steps_per_second": 1.875, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.6764822920811778, | |
| "grad_norm": 5.347845077514648, | |
| "learning_rate": 3.899900523445614e-05, | |
| "loss": 0.6496, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.716275368085953, | |
| "grad_norm": 2.6406002044677734, | |
| "learning_rate": 3.76219481169952e-05, | |
| "loss": 0.6514, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.716275368085953, | |
| "eval_accuracy": 0.6093311506080449, | |
| "eval_loss": 0.6483238935470581, | |
| "eval_runtime": 71.7594, | |
| "eval_samples_per_second": 119.176, | |
| "eval_steps_per_second": 1.867, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.7560684440907283, | |
| "grad_norm": 2.9834582805633545, | |
| "learning_rate": 3.61914224836288e-05, | |
| "loss": 0.6433, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.7958615200955034, | |
| "grad_norm": 1.781412124633789, | |
| "learning_rate": 3.4713488261229724e-05, | |
| "loss": 0.6481, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.7958615200955034, | |
| "eval_accuracy": 0.6115528531337698, | |
| "eval_loss": 0.6509259939193726, | |
| "eval_runtime": 71.5666, | |
| "eval_samples_per_second": 119.497, | |
| "eval_steps_per_second": 1.872, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.8356545961002786, | |
| "grad_norm": 2.0181474685668945, | |
| "learning_rate": 3.319440620675442e-05, | |
| "loss": 0.6479, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.8754476721050537, | |
| "grad_norm": 1.780383586883545, | |
| "learning_rate": 3.164061138571247e-05, | |
| "loss": 0.6455, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.8754476721050537, | |
| "eval_accuracy": 0.6241814780168382, | |
| "eval_loss": 0.6406447887420654, | |
| "eval_runtime": 71.0764, | |
| "eval_samples_per_second": 120.321, | |
| "eval_steps_per_second": 1.885, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.9152407481098289, | |
| "grad_norm": 1.8478803634643555, | |
| "learning_rate": 3.005868591223814e-05, | |
| "loss": 0.6414, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.955033824114604, | |
| "grad_norm": 2.130547046661377, | |
| "learning_rate": 2.8455331066241263e-05, | |
| "loss": 0.6345, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.955033824114604, | |
| "eval_accuracy": 0.6162301216089804, | |
| "eval_loss": 0.6427032351493835, | |
| "eval_runtime": 71.4565, | |
| "eval_samples_per_second": 119.681, | |
| "eval_steps_per_second": 1.875, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.9948269001193792, | |
| "grad_norm": 1.3212426900863647, | |
| "learning_rate": 2.6837338905753685e-05, | |
| "loss": 0.6455, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.0346199761241544, | |
| "grad_norm": 2.934187412261963, | |
| "learning_rate": 2.5211563494725422e-05, | |
| "loss": 0.6022, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.0346199761241544, | |
| "eval_accuracy": 0.6224275023386342, | |
| "eval_loss": 0.6543847322463989, | |
| "eval_runtime": 71.1065, | |
| "eval_samples_per_second": 120.27, | |
| "eval_steps_per_second": 1.884, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 1.0744130521289295, | |
| "grad_norm": 2.43818736076355, | |
| "learning_rate": 2.3584891868154375e-05, | |
| "loss": 0.5924, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 1.1142061281337048, | |
| "grad_norm": 3.3583297729492188, | |
| "learning_rate": 2.196421485754547e-05, | |
| "loss": 0.5914, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.1142061281337048, | |
| "eval_accuracy": 0.6216089803554724, | |
| "eval_loss": 0.653032124042511, | |
| "eval_runtime": 71.6366, | |
| "eval_samples_per_second": 119.38, | |
| "eval_steps_per_second": 1.871, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 1.15399920413848, | |
| "grad_norm": 3.130229949951172, | |
| "learning_rate": 2.0356397900287307e-05, | |
| "loss": 0.5859, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 1.193792280143255, | |
| "grad_norm": 3.7583000659942627, | |
| "learning_rate": 1.8768251956602138e-05, | |
| "loss": 0.5798, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.193792280143255, | |
| "eval_accuracy": 0.6282740879326474, | |
| "eval_loss": 0.6515339612960815, | |
| "eval_runtime": 71.5158, | |
| "eval_samples_per_second": 119.582, | |
| "eval_steps_per_second": 1.874, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.2335853561480303, | |
| "grad_norm": 3.0673882961273193, | |
| "learning_rate": 1.7206504657269547e-05, | |
| "loss": 0.5836, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 1.2733784321528054, | |
| "grad_norm": 3.4926602840423584, | |
| "learning_rate": 1.5677771804346432e-05, | |
| "loss": 0.5757, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.2733784321528054, | |
| "eval_accuracy": 0.6352899906454631, | |
| "eval_loss": 0.6421129107475281, | |
| "eval_runtime": 71.3733, | |
| "eval_samples_per_second": 119.821, | |
| "eval_steps_per_second": 1.877, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 1.3131715081575805, | |
| "grad_norm": 4.403800964355469, | |
| "learning_rate": 1.4188529345610756e-05, | |
| "loss": 0.5776, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 1.3529645841623559, | |
| "grad_norm": 2.677273750305176, | |
| "learning_rate": 1.2745085941449484e-05, | |
| "loss": 0.5843, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.3529645841623559, | |
| "eval_accuracy": 0.6383302151543498, | |
| "eval_loss": 0.6391754150390625, | |
| "eval_runtime": 71.8797, | |
| "eval_samples_per_second": 118.977, | |
| "eval_steps_per_second": 1.864, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 1.392757660167131, | |
| "grad_norm": 2.8799660205841064, | |
| "learning_rate": 1.1353556240401692e-05, | |
| "loss": 0.5758, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.432550736171906, | |
| "grad_norm": 3.6387939453125, | |
| "learning_rate": 1.0019834976565779e-05, | |
| "loss": 0.5758, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.432550736171906, | |
| "eval_accuracy": 0.6308465855940131, | |
| "eval_loss": 0.642611563205719, | |
| "eval_runtime": 71.6295, | |
| "eval_samples_per_second": 119.392, | |
| "eval_steps_per_second": 1.871, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 1.4723438121766812, | |
| "grad_norm": 3.1025497913360596, | |
| "learning_rate": 8.749571998598052e-06, | |
| "loss": 0.5826, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 1.5121368881814563, | |
| "grad_norm": 3.0935654640197754, | |
| "learning_rate": 7.548148336083807e-06, | |
| "loss": 0.5739, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.5121368881814563, | |
| "eval_accuracy": 0.6361085126286249, | |
| "eval_loss": 0.6375930309295654, | |
| "eval_runtime": 71.3926, | |
| "eval_samples_per_second": 119.788, | |
| "eval_steps_per_second": 1.877, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 1.5519299641862316, | |
| "grad_norm": 2.953458786010742, | |
| "learning_rate": 6.420653404667356e-06, | |
| "loss": 0.5688, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 1.5917230401910067, | |
| "grad_norm": 3.1475229263305664, | |
| "learning_rate": 5.371863446503628e-06, | |
| "loss": 0.5717, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.5917230401910067, | |
| "eval_accuracy": 0.6387979420018709, | |
| "eval_loss": 0.6375735402107239, | |
| "eval_runtime": 71.6947, | |
| "eval_samples_per_second": 119.284, | |
| "eval_steps_per_second": 1.869, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.6315161161957819, | |
| "grad_norm": 3.241312265396118, | |
| "learning_rate": 4.406221297360902e-06, | |
| "loss": 0.564, | |
| "step": 4100 | |
| }, | |
| { | |
| "epoch": 1.6713091922005572, | |
| "grad_norm": 3.462087392807007, | |
| "learning_rate": 3.5278175660843267e-06, | |
| "loss": 0.5679, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.6713091922005572, | |
| "eval_accuracy": 0.6421889616463985, | |
| "eval_loss": 0.6346476674079895, | |
| "eval_runtime": 71.4517, | |
| "eval_samples_per_second": 119.689, | |
| "eval_steps_per_second": 1.875, | |
| "step": 4200 | |
| }, | |
| { | |
| "epoch": 1.7111022682053323, | |
| "grad_norm": 3.355526924133301, | |
| "learning_rate": 2.7403733061469454e-06, | |
| "loss": 0.5736, | |
| "step": 4300 | |
| }, | |
| { | |
| "epoch": 1.7508953442101074, | |
| "grad_norm": 4.239299297332764, | |
| "learning_rate": 2.047224252694127e-06, | |
| "loss": 0.5525, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.7508953442101074, | |
| "eval_accuracy": 0.6434752104770813, | |
| "eval_loss": 0.638582706451416, | |
| "eval_runtime": 71.4737, | |
| "eval_samples_per_second": 119.652, | |
| "eval_steps_per_second": 1.875, | |
| "step": 4400 | |
| }, | |
| { | |
| "epoch": 1.7906884202148827, | |
| "grad_norm": 4.165896415710449, | |
| "learning_rate": 1.4513066918558722e-06, | |
| "loss": 0.5634, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 1.8304814962196578, | |
| "grad_norm": 3.2473928928375244, | |
| "learning_rate": 9.551450221865599e-07, | |
| "loss": 0.5702, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.8304814962196578, | |
| "eval_accuracy": 0.6438260056127222, | |
| "eval_loss": 0.6338381767272949, | |
| "eval_runtime": 71.8808, | |
| "eval_samples_per_second": 118.975, | |
| "eval_steps_per_second": 1.864, | |
| "step": 4600 | |
| }, | |
| { | |
| "epoch": 1.870274572224433, | |
| "grad_norm": 3.846665620803833, | |
| "learning_rate": 5.608410609239501e-07, | |
| "loss": 0.5714, | |
| "step": 4700 | |
| }, | |
| { | |
| "epoch": 1.9100676482292083, | |
| "grad_norm": 3.0735666751861572, | |
| "learning_rate": 2.7006514036768957e-07, | |
| "loss": 0.5652, | |
| "step": 4800 | |
| }, | |
| { | |
| "epoch": 1.9100676482292083, | |
| "eval_accuracy": 0.6452291861552854, | |
| "eval_loss": 0.6332165002822876, | |
| "eval_runtime": 71.8255, | |
| "eval_samples_per_second": 119.066, | |
| "eval_steps_per_second": 1.866, | |
| "step": 4800 | |
| } | |
| ], | |
| "logging_steps": 100, | |
| "max_steps": 5026, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 2, | |
| "save_steps": 400, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 0.0, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |