{ "best_global_step": 760, "best_metric": 2.077368974685669, "best_model_checkpoint": "./checkpoints\\checkpoint-760", "epoch": 10.0, "eval_steps": 500, "global_step": 760, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.13311148086522462, "grad_norm": 24.95787811279297, "learning_rate": 1.9763157894736843e-05, "loss": 23.22579345703125, "step": 10 }, { "epoch": 0.26622296173044924, "grad_norm": 22.520177841186523, "learning_rate": 1.95e-05, "loss": 22.365089416503906, "step": 20 }, { "epoch": 0.39933444259567386, "grad_norm": 22.618810653686523, "learning_rate": 1.923684210526316e-05, "loss": 21.68932647705078, "step": 30 }, { "epoch": 0.5324459234608985, "grad_norm": 26.469789505004883, "learning_rate": 1.8973684210526316e-05, "loss": 21.058622741699217, "step": 40 }, { "epoch": 0.6655574043261231, "grad_norm": 18.552188873291016, "learning_rate": 1.8710526315789476e-05, "loss": 20.97467803955078, "step": 50 }, { "epoch": 0.7986688851913477, "grad_norm": 16.745996475219727, "learning_rate": 1.8447368421052632e-05, "loss": 20.954002380371094, "step": 60 }, { "epoch": 0.9317803660565723, "grad_norm": 19.304048538208008, "learning_rate": 1.8184210526315792e-05, "loss": 20.92015838623047, "step": 70 }, { "epoch": 1.0, "eval_loss": 2.250842332839966, "eval_runtime": 0.3928, "eval_samples_per_second": 254.561, "eval_steps_per_second": 33.093, "step": 76 }, { "epoch": 1.0532445923460898, "grad_norm": 17.96588706970215, "learning_rate": 1.792105263157895e-05, "loss": 18.285963439941405, "step": 80 }, { "epoch": 1.1863560732113145, "grad_norm": 31.747936248779297, "learning_rate": 1.765789473684211e-05, "loss": 20.70150146484375, "step": 90 }, { "epoch": 1.319467554076539, "grad_norm": 15.776130676269531, "learning_rate": 1.7394736842105265e-05, "loss": 19.400506591796876, "step": 100 }, { "epoch": 1.4525790349417638, "grad_norm": 22.212865829467773, "learning_rate": 1.7131578947368425e-05, "loss": 18.122430419921876, "step": 110 }, { "epoch": 1.5856905158069883, "grad_norm": 16.966033935546875, "learning_rate": 1.686842105263158e-05, "loss": 19.32366027832031, "step": 120 }, { "epoch": 1.718801996672213, "grad_norm": 23.452783584594727, "learning_rate": 1.6605263157894738e-05, "loss": 19.914886474609375, "step": 130 }, { "epoch": 1.8519134775374377, "grad_norm": 20.83370590209961, "learning_rate": 1.6342105263157898e-05, "loss": 19.02434387207031, "step": 140 }, { "epoch": 1.9850249584026622, "grad_norm": 15.527095794677734, "learning_rate": 1.6078947368421054e-05, "loss": 19.608319091796876, "step": 150 }, { "epoch": 2.0, "eval_loss": 2.1818811893463135, "eval_runtime": 0.415, "eval_samples_per_second": 240.96, "eval_steps_per_second": 31.325, "step": 152 }, { "epoch": 2.1064891846921796, "grad_norm": 17.349496841430664, "learning_rate": 1.581578947368421e-05, "loss": 17.56553497314453, "step": 160 }, { "epoch": 2.239600665557404, "grad_norm": 15.411664009094238, "learning_rate": 1.555263157894737e-05, "loss": 18.280393981933592, "step": 170 }, { "epoch": 2.372712146422629, "grad_norm": 23.818889617919922, "learning_rate": 1.5289473684210527e-05, "loss": 18.51408233642578, "step": 180 }, { "epoch": 2.5058236272878536, "grad_norm": 14.444835662841797, "learning_rate": 1.5026315789473684e-05, "loss": 19.563699340820314, "step": 190 }, { "epoch": 2.638935108153078, "grad_norm": 23.308435440063477, "learning_rate": 1.4763157894736844e-05, "loss": 18.859063720703126, "step": 200 }, { "epoch": 2.772046589018303, "grad_norm": 18.46608543395996, "learning_rate": 1.45e-05, "loss": 18.843687438964842, "step": 210 }, { "epoch": 2.9051580698835275, "grad_norm": 26.819814682006836, "learning_rate": 1.423684210526316e-05, "loss": 18.473309326171876, "step": 220 }, { "epoch": 3.0, "eval_loss": 2.146812915802002, "eval_runtime": 0.3967, "eval_samples_per_second": 252.107, "eval_steps_per_second": 32.774, "step": 228 }, { "epoch": 3.026622296173045, "grad_norm": 21.41319465637207, "learning_rate": 1.3973684210526316e-05, "loss": 17.324299621582032, "step": 230 }, { "epoch": 3.1597337770382694, "grad_norm": 15.841781616210938, "learning_rate": 1.3710526315789475e-05, "loss": 18.82971649169922, "step": 240 }, { "epoch": 3.2928452579034944, "grad_norm": 21.451080322265625, "learning_rate": 1.3447368421052633e-05, "loss": 18.335603332519533, "step": 250 }, { "epoch": 3.425956738768719, "grad_norm": 21.583675384521484, "learning_rate": 1.3184210526315791e-05, "loss": 18.502110290527344, "step": 260 }, { "epoch": 3.5590682196339434, "grad_norm": 21.023969650268555, "learning_rate": 1.2921052631578948e-05, "loss": 17.96020050048828, "step": 270 }, { "epoch": 3.6921797004991683, "grad_norm": 16.162145614624023, "learning_rate": 1.2657894736842107e-05, "loss": 17.875433349609374, "step": 280 }, { "epoch": 3.825291181364393, "grad_norm": 14.14526653289795, "learning_rate": 1.2394736842105264e-05, "loss": 17.802043151855468, "step": 290 }, { "epoch": 3.9584026622296173, "grad_norm": 18.50859260559082, "learning_rate": 1.2131578947368424e-05, "loss": 18.82775115966797, "step": 300 }, { "epoch": 4.0, "eval_loss": 2.1246590614318848, "eval_runtime": 0.3915, "eval_samples_per_second": 255.438, "eval_steps_per_second": 33.207, "step": 304 }, { "epoch": 4.079866888519135, "grad_norm": 16.61701011657715, "learning_rate": 1.186842105263158e-05, "loss": 15.94822998046875, "step": 310 }, { "epoch": 4.212978369384359, "grad_norm": 18.87012481689453, "learning_rate": 1.1605263157894737e-05, "loss": 18.40921630859375, "step": 320 }, { "epoch": 4.346089850249584, "grad_norm": 18.27983283996582, "learning_rate": 1.1342105263157897e-05, "loss": 17.69190216064453, "step": 330 }, { "epoch": 4.479201331114808, "grad_norm": 19.216075897216797, "learning_rate": 1.1078947368421053e-05, "loss": 18.44379119873047, "step": 340 }, { "epoch": 4.612312811980034, "grad_norm": 23.174020767211914, "learning_rate": 1.0815789473684211e-05, "loss": 17.72493896484375, "step": 350 }, { "epoch": 4.745424292845258, "grad_norm": 16.41116714477539, "learning_rate": 1.055263157894737e-05, "loss": 17.35713348388672, "step": 360 }, { "epoch": 4.878535773710483, "grad_norm": 17.488855361938477, "learning_rate": 1.0289473684210528e-05, "loss": 18.663365173339844, "step": 370 }, { "epoch": 5.0, "grad_norm": 17.824460983276367, "learning_rate": 1.0026315789473684e-05, "loss": 16.387326049804688, "step": 380 }, { "epoch": 5.0, "eval_loss": 2.1079440116882324, "eval_runtime": 0.3846, "eval_samples_per_second": 260.026, "eval_steps_per_second": 33.803, "step": 380 }, { "epoch": 5.1331114808652245, "grad_norm": 18.288578033447266, "learning_rate": 9.763157894736844e-06, "loss": 17.112588500976564, "step": 390 }, { "epoch": 5.266222961730449, "grad_norm": 16.713876724243164, "learning_rate": 9.5e-06, "loss": 17.891526794433595, "step": 400 }, { "epoch": 5.3993344425956735, "grad_norm": 18.55312156677246, "learning_rate": 9.236842105263159e-06, "loss": 16.945428466796876, "step": 410 }, { "epoch": 5.532445923460898, "grad_norm": 16.772281646728516, "learning_rate": 8.973684210526317e-06, "loss": 17.56818542480469, "step": 420 }, { "epoch": 5.665557404326123, "grad_norm": 15.288823127746582, "learning_rate": 8.710526315789475e-06, "loss": 17.746002197265625, "step": 430 }, { "epoch": 5.798668885191348, "grad_norm": 18.58148956298828, "learning_rate": 8.447368421052632e-06, "loss": 18.207766723632812, "step": 440 }, { "epoch": 5.931780366056572, "grad_norm": 28.35038948059082, "learning_rate": 8.18421052631579e-06, "loss": 17.60784454345703, "step": 450 }, { "epoch": 6.0, "eval_loss": 2.095176935195923, "eval_runtime": 0.3934, "eval_samples_per_second": 254.226, "eval_steps_per_second": 33.049, "step": 456 }, { "epoch": 6.05324459234609, "grad_norm": 21.211326599121094, "learning_rate": 7.921052631578948e-06, "loss": 15.379603576660156, "step": 460 }, { "epoch": 6.186356073211314, "grad_norm": 19.47175407409668, "learning_rate": 7.657894736842106e-06, "loss": 18.156336975097656, "step": 470 }, { "epoch": 6.319467554076539, "grad_norm": 18.88157081604004, "learning_rate": 7.3947368421052635e-06, "loss": 17.528657531738283, "step": 480 }, { "epoch": 6.452579034941763, "grad_norm": 23.712169647216797, "learning_rate": 7.131578947368422e-06, "loss": 17.638197326660155, "step": 490 }, { "epoch": 6.585690515806989, "grad_norm": 18.30068588256836, "learning_rate": 6.86842105263158e-06, "loss": 16.143788146972657, "step": 500 }, { "epoch": 6.718801996672213, "grad_norm": 37.50796890258789, "learning_rate": 6.605263157894738e-06, "loss": 17.233677673339844, "step": 510 }, { "epoch": 6.851913477537438, "grad_norm": 24.004518508911133, "learning_rate": 6.342105263157895e-06, "loss": 17.299192810058592, "step": 520 }, { "epoch": 6.985024958402662, "grad_norm": 16.603742599487305, "learning_rate": 6.0789473684210535e-06, "loss": 17.784269714355467, "step": 530 }, { "epoch": 7.0, "eval_loss": 2.088780164718628, "eval_runtime": 0.3911, "eval_samples_per_second": 255.71, "eval_steps_per_second": 33.242, "step": 532 }, { "epoch": 7.10648918469218, "grad_norm": 17.8812313079834, "learning_rate": 5.815789473684212e-06, "loss": 16.132447814941408, "step": 540 }, { "epoch": 7.239600665557404, "grad_norm": 13.379035949707031, "learning_rate": 5.552631578947368e-06, "loss": 17.281491088867188, "step": 550 }, { "epoch": 7.372712146422629, "grad_norm": 17.844518661499023, "learning_rate": 5.289473684210526e-06, "loss": 17.37604217529297, "step": 560 }, { "epoch": 7.505823627287853, "grad_norm": 14.862391471862793, "learning_rate": 5.026315789473685e-06, "loss": 16.679644775390624, "step": 570 }, { "epoch": 7.6389351081530785, "grad_norm": 24.299837112426758, "learning_rate": 4.763157894736842e-06, "loss": 18.08904266357422, "step": 580 }, { "epoch": 7.772046589018303, "grad_norm": 31.28373908996582, "learning_rate": 4.5e-06, "loss": 17.572297668457033, "step": 590 }, { "epoch": 7.9051580698835275, "grad_norm": 15.72098445892334, "learning_rate": 4.236842105263158e-06, "loss": 16.98594970703125, "step": 600 }, { "epoch": 8.0, "eval_loss": 2.0818235874176025, "eval_runtime": 0.3984, "eval_samples_per_second": 250.992, "eval_steps_per_second": 32.629, "step": 608 }, { "epoch": 8.026622296173045, "grad_norm": 24.72064781188965, "learning_rate": 3.9736842105263165e-06, "loss": 15.768360900878907, "step": 610 }, { "epoch": 8.15973377703827, "grad_norm": 13.181447982788086, "learning_rate": 3.710526315789474e-06, "loss": 17.512715148925782, "step": 620 }, { "epoch": 8.292845257903494, "grad_norm": 14.77968692779541, "learning_rate": 3.447368421052632e-06, "loss": 16.602384948730467, "step": 630 }, { "epoch": 8.425956738768718, "grad_norm": 19.155954360961914, "learning_rate": 3.1842105263157898e-06, "loss": 16.25328826904297, "step": 640 }, { "epoch": 8.559068219633943, "grad_norm": 21.890533447265625, "learning_rate": 2.921052631578948e-06, "loss": 17.177554321289062, "step": 650 }, { "epoch": 8.692179700499167, "grad_norm": 16.814281463623047, "learning_rate": 2.6578947368421053e-06, "loss": 18.03597412109375, "step": 660 }, { "epoch": 8.825291181364392, "grad_norm": 21.507701873779297, "learning_rate": 2.3947368421052635e-06, "loss": 17.011363220214843, "step": 670 }, { "epoch": 8.958402662229616, "grad_norm": 17.85552978515625, "learning_rate": 2.1315789473684212e-06, "loss": 17.358226013183593, "step": 680 }, { "epoch": 9.0, "eval_loss": 2.0784778594970703, "eval_runtime": 0.4032, "eval_samples_per_second": 247.996, "eval_steps_per_second": 32.239, "step": 684 }, { "epoch": 9.079866888519135, "grad_norm": 21.801002502441406, "learning_rate": 1.868421052631579e-06, "loss": 15.464262390136719, "step": 690 }, { "epoch": 9.21297836938436, "grad_norm": 14.75335693359375, "learning_rate": 1.605263157894737e-06, "loss": 16.6826171875, "step": 700 }, { "epoch": 9.346089850249584, "grad_norm": 17.793237686157227, "learning_rate": 1.342105263157895e-06, "loss": 17.95832061767578, "step": 710 }, { "epoch": 9.479201331114808, "grad_norm": 24.57212257385254, "learning_rate": 1.0789473684210527e-06, "loss": 17.297279357910156, "step": 720 }, { "epoch": 9.612312811980033, "grad_norm": 19.45013999938965, "learning_rate": 8.157894736842106e-07, "loss": 17.156161499023437, "step": 730 }, { "epoch": 9.745424292845257, "grad_norm": 17.178083419799805, "learning_rate": 5.526315789473684e-07, "loss": 17.450473022460937, "step": 740 }, { "epoch": 9.878535773710482, "grad_norm": 24.745452880859375, "learning_rate": 2.894736842105263e-07, "loss": 17.295619201660156, "step": 750 }, { "epoch": 10.0, "grad_norm": 20.33419418334961, "learning_rate": 2.631578947368421e-08, "loss": 15.296475219726563, "step": 760 }, { "epoch": 10.0, "eval_loss": 2.077368974685669, "eval_runtime": 0.388, "eval_samples_per_second": 257.713, "eval_steps_per_second": 33.503, "step": 760 } ], "logging_steps": 10, "max_steps": 760, "num_input_tokens_seen": 0, "num_train_epochs": 10, "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": 2723067341429760.0, "train_batch_size": 2, "trial_name": null, "trial_params": null }