| { |
| "best_metric": 0.6732283464566929, |
| "best_model_checkpoint": "./llama_acc_reward_model_1e5-bz32/checkpoint-550", |
| "epoch": 3.8194444444444446, |
| "eval_steps": 50, |
| "global_step": 550, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.06944444444444445, |
| "grad_norm": 237.0, |
| "learning_rate": 1.3888888888888892e-06, |
| "loss": 1.5531, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.1388888888888889, |
| "grad_norm": 284.0, |
| "learning_rate": 2.7777777777777783e-06, |
| "loss": 1.2538, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.20833333333333334, |
| "grad_norm": 83.0, |
| "learning_rate": 4.166666666666667e-06, |
| "loss": 0.8203, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.2777777777777778, |
| "grad_norm": 171.0, |
| "learning_rate": 5.555555555555557e-06, |
| "loss": 0.7287, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.3472222222222222, |
| "grad_norm": 88.0, |
| "learning_rate": 6.944444444444445e-06, |
| "loss": 0.8232, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.3472222222222222, |
| "eval_accuracy": 0.6141732283464567, |
| "eval_auc": 0.6284406565656565, |
| "eval_f1": 0.7083333333333334, |
| "eval_loss": 0.8070767521858215, |
| "eval_precision": 0.6197916666666666, |
| "eval_recall": 0.8263888888888888, |
| "eval_runtime": 17.2507, |
| "eval_samples_per_second": 14.724, |
| "eval_steps_per_second": 0.232, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.4166666666666667, |
| "grad_norm": 184.0, |
| "learning_rate": 8.333333333333334e-06, |
| "loss": 0.922, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.4861111111111111, |
| "grad_norm": 45.5, |
| "learning_rate": 9.722222222222223e-06, |
| "loss": 0.838, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.5555555555555556, |
| "grad_norm": 91.5, |
| "learning_rate": 9.996239762521152e-06, |
| "loss": 0.7773, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.625, |
| "grad_norm": 50.5, |
| "learning_rate": 9.980973490458728e-06, |
| "loss": 0.744, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.6944444444444444, |
| "grad_norm": 22.625, |
| "learning_rate": 9.954002016824226e-06, |
| "loss": 0.9683, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.6944444444444444, |
| "eval_accuracy": 0.5787401574803149, |
| "eval_auc": 0.6018623737373737, |
| "eval_f1": 0.7002801120448179, |
| "eval_loss": 0.7812102437019348, |
| "eval_precision": 0.5868544600938967, |
| "eval_recall": 0.8680555555555556, |
| "eval_runtime": 17.3832, |
| "eval_samples_per_second": 14.612, |
| "eval_steps_per_second": 0.23, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.7638888888888888, |
| "grad_norm": 126.5, |
| "learning_rate": 9.915388724114301e-06, |
| "loss": 0.8441, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.8333333333333334, |
| "grad_norm": 44.0, |
| "learning_rate": 9.86522435289912e-06, |
| "loss": 0.7468, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.9027777777777778, |
| "grad_norm": 145.0, |
| "learning_rate": 9.803626788583603e-06, |
| "loss": 0.7598, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.9722222222222222, |
| "grad_norm": 100.0, |
| "learning_rate": 9.730740784378755e-06, |
| "loss": 0.7506, |
| "step": 140 |
| }, |
| { |
| "epoch": 1.0416666666666667, |
| "grad_norm": 18.375, |
| "learning_rate": 9.646737621134112e-06, |
| "loss": 0.669, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.0416666666666667, |
| "eval_accuracy": 0.6023622047244095, |
| "eval_auc": 0.6256628787878789, |
| "eval_f1": 0.6710097719869706, |
| "eval_loss": 0.7021759748458862, |
| "eval_precision": 0.6319018404907976, |
| "eval_recall": 0.7152777777777778, |
| "eval_runtime": 17.2473, |
| "eval_samples_per_second": 14.727, |
| "eval_steps_per_second": 0.232, |
| "step": 150 |
| }, |
| { |
| "epoch": 1.1111111111111112, |
| "grad_norm": 74.0, |
| "learning_rate": 9.551814704830734e-06, |
| "loss": 0.6399, |
| "step": 160 |
| }, |
| { |
| "epoch": 1.1805555555555556, |
| "grad_norm": 91.5, |
| "learning_rate": 9.446195102680531e-06, |
| "loss": 0.6868, |
| "step": 170 |
| }, |
| { |
| "epoch": 1.25, |
| "grad_norm": 23.875, |
| "learning_rate": 9.330127018922195e-06, |
| "loss": 0.6817, |
| "step": 180 |
| }, |
| { |
| "epoch": 1.3194444444444444, |
| "grad_norm": 80.5, |
| "learning_rate": 9.203883211545517e-06, |
| "loss": 0.7281, |
| "step": 190 |
| }, |
| { |
| "epoch": 1.3888888888888888, |
| "grad_norm": 50.25, |
| "learning_rate": 9.067760351314838e-06, |
| "loss": 0.606, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.3888888888888888, |
| "eval_accuracy": 0.5905511811023622, |
| "eval_auc": 0.6167929292929293, |
| "eval_f1": 0.695906432748538, |
| "eval_loss": 0.7337387204170227, |
| "eval_precision": 0.601010101010101, |
| "eval_recall": 0.8263888888888888, |
| "eval_runtime": 17.381, |
| "eval_samples_per_second": 14.614, |
| "eval_steps_per_second": 0.23, |
| "step": 200 |
| }, |
| { |
| "epoch": 1.4583333333333333, |
| "grad_norm": 44.5, |
| "learning_rate": 8.92207832459788e-06, |
| "loss": 0.5274, |
| "step": 210 |
| }, |
| { |
| "epoch": 1.5277777777777777, |
| "grad_norm": 61.75, |
| "learning_rate": 8.767179481638303e-06, |
| "loss": 0.6415, |
| "step": 220 |
| }, |
| { |
| "epoch": 1.5972222222222223, |
| "grad_norm": 60.75, |
| "learning_rate": 8.603427832038574e-06, |
| "loss": 0.7424, |
| "step": 230 |
| }, |
| { |
| "epoch": 1.6666666666666665, |
| "grad_norm": 67.5, |
| "learning_rate": 8.43120818934367e-06, |
| "loss": 0.5821, |
| "step": 240 |
| }, |
| { |
| "epoch": 1.7361111111111112, |
| "grad_norm": 118.0, |
| "learning_rate": 8.25092526673592e-06, |
| "loss": 0.6577, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.7361111111111112, |
| "eval_accuracy": 0.6338582677165354, |
| "eval_auc": 0.6801136363636364, |
| "eval_f1": 0.7319884726224783, |
| "eval_loss": 0.6989323496818542, |
| "eval_precision": 0.625615763546798, |
| "eval_recall": 0.8819444444444444, |
| "eval_runtime": 17.2572, |
| "eval_samples_per_second": 14.718, |
| "eval_steps_per_second": 0.232, |
| "step": 250 |
| }, |
| { |
| "epoch": 1.8055555555555556, |
| "grad_norm": 35.25, |
| "learning_rate": 8.063002725966014e-06, |
| "loss": 0.5856, |
| "step": 260 |
| }, |
| { |
| "epoch": 1.875, |
| "grad_norm": 19.125, |
| "learning_rate": 7.86788218175523e-06, |
| "loss": 0.5973, |
| "step": 270 |
| }, |
| { |
| "epoch": 1.9444444444444444, |
| "grad_norm": 14.375, |
| "learning_rate": 7.666022164008458e-06, |
| "loss": 0.5621, |
| "step": 280 |
| }, |
| { |
| "epoch": 2.013888888888889, |
| "grad_norm": 59.75, |
| "learning_rate": 7.457897040276853e-06, |
| "loss": 0.5428, |
| "step": 290 |
| }, |
| { |
| "epoch": 2.0833333333333335, |
| "grad_norm": 88.0, |
| "learning_rate": 7.243995901002312e-06, |
| "loss": 0.5459, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.0833333333333335, |
| "eval_accuracy": 0.6417322834645669, |
| "eval_auc": 0.7095012626262626, |
| "eval_f1": 0.7436619718309859, |
| "eval_loss": 0.704635500907898, |
| "eval_precision": 0.6255924170616114, |
| "eval_recall": 0.9166666666666666, |
| "eval_runtime": 17.2333, |
| "eval_samples_per_second": 14.739, |
| "eval_steps_per_second": 0.232, |
| "step": 300 |
| }, |
| { |
| "epoch": 2.1527777777777777, |
| "grad_norm": 32.0, |
| "learning_rate": 7.0248214101633685e-06, |
| "loss": 0.5229, |
| "step": 310 |
| }, |
| { |
| "epoch": 2.2222222222222223, |
| "grad_norm": 22.0, |
| "learning_rate": 6.800888624023552e-06, |
| "loss": 0.5532, |
| "step": 320 |
| }, |
| { |
| "epoch": 2.2916666666666665, |
| "grad_norm": 21.375, |
| "learning_rate": 6.572723780758069e-06, |
| "loss": 0.5129, |
| "step": 330 |
| }, |
| { |
| "epoch": 2.361111111111111, |
| "grad_norm": 46.0, |
| "learning_rate": 6.340863063803187e-06, |
| "loss": 0.4531, |
| "step": 340 |
| }, |
| { |
| "epoch": 2.4305555555555554, |
| "grad_norm": 24.125, |
| "learning_rate": 6.105851341834439e-06, |
| "loss": 0.4496, |
| "step": 350 |
| }, |
| { |
| "epoch": 2.4305555555555554, |
| "eval_accuracy": 0.6496062992125984, |
| "eval_auc": 0.6922032828282829, |
| "eval_f1": 0.6898954703832753, |
| "eval_loss": 0.6863601207733154, |
| "eval_precision": 0.6923076923076923, |
| "eval_recall": 0.6875, |
| "eval_runtime": 17.3556, |
| "eval_samples_per_second": 14.635, |
| "eval_steps_per_second": 0.23, |
| "step": 350 |
| }, |
| { |
| "epoch": 2.5, |
| "grad_norm": 66.0, |
| "learning_rate": 5.8682408883346535e-06, |
| "loss": 0.4314, |
| "step": 360 |
| }, |
| { |
| "epoch": 2.5694444444444446, |
| "grad_norm": 38.5, |
| "learning_rate": 5.628590083760815e-06, |
| "loss": 0.4262, |
| "step": 370 |
| }, |
| { |
| "epoch": 2.638888888888889, |
| "grad_norm": 28.75, |
| "learning_rate": 5.387462103359655e-06, |
| "loss": 0.461, |
| "step": 380 |
| }, |
| { |
| "epoch": 2.7083333333333335, |
| "grad_norm": 20.625, |
| "learning_rate": 5.145423593715558e-06, |
| "loss": 0.4406, |
| "step": 390 |
| }, |
| { |
| "epoch": 2.7777777777777777, |
| "grad_norm": 36.25, |
| "learning_rate": 4.903043341140879e-06, |
| "loss": 0.3986, |
| "step": 400 |
| }, |
| { |
| "epoch": 2.7777777777777777, |
| "eval_accuracy": 0.6535433070866141, |
| "eval_auc": 0.6909722222222223, |
| "eval_f1": 0.6691729323308271, |
| "eval_loss": 0.7401882410049438, |
| "eval_precision": 0.7295081967213115, |
| "eval_recall": 0.6180555555555556, |
| "eval_runtime": 17.2396, |
| "eval_samples_per_second": 14.734, |
| "eval_steps_per_second": 0.232, |
| "step": 400 |
| }, |
| { |
| "epoch": 2.8472222222222223, |
| "grad_norm": 35.0, |
| "learning_rate": 4.660890935037954e-06, |
| "loss": 0.3808, |
| "step": 410 |
| }, |
| { |
| "epoch": 2.9166666666666665, |
| "grad_norm": 42.25, |
| "learning_rate": 4.4195354293738484e-06, |
| "loss": 0.4725, |
| "step": 420 |
| }, |
| { |
| "epoch": 2.986111111111111, |
| "grad_norm": 26.125, |
| "learning_rate": 4.17954400541338e-06, |
| "loss": 0.4157, |
| "step": 430 |
| }, |
| { |
| "epoch": 3.0555555555555554, |
| "grad_norm": 20.875, |
| "learning_rate": 3.941480638852948e-06, |
| "loss": 0.2813, |
| "step": 440 |
| }, |
| { |
| "epoch": 3.125, |
| "grad_norm": 23.75, |
| "learning_rate": 3.705904774487396e-06, |
| "loss": 0.2891, |
| "step": 450 |
| }, |
| { |
| "epoch": 3.125, |
| "eval_accuracy": 0.6456692913385826, |
| "eval_auc": 0.6821338383838383, |
| "eval_f1": 0.6938775510204082, |
| "eval_loss": 0.7412765622138977, |
| "eval_precision": 0.68, |
| "eval_recall": 0.7083333333333334, |
| "eval_runtime": 17.2413, |
| "eval_samples_per_second": 14.732, |
| "eval_steps_per_second": 0.232, |
| "step": 450 |
| }, |
| { |
| "epoch": 3.1944444444444446, |
| "grad_norm": 18.0, |
| "learning_rate": 3.473370011524435e-06, |
| "loss": 0.1611, |
| "step": 460 |
| }, |
| { |
| "epoch": 3.263888888888889, |
| "grad_norm": 51.0, |
| "learning_rate": 3.244422802636057e-06, |
| "loss": 0.1808, |
| "step": 470 |
| }, |
| { |
| "epoch": 3.3333333333333335, |
| "grad_norm": 50.5, |
| "learning_rate": 3.019601169804216e-06, |
| "loss": 0.1928, |
| "step": 480 |
| }, |
| { |
| "epoch": 3.4027777777777777, |
| "grad_norm": 74.5, |
| "learning_rate": 2.7994334399784773e-06, |
| "loss": 0.2171, |
| "step": 490 |
| }, |
| { |
| "epoch": 3.4722222222222223, |
| "grad_norm": 55.25, |
| "learning_rate": 2.5844370035168077e-06, |
| "loss": 0.1798, |
| "step": 500 |
| }, |
| { |
| "epoch": 3.4722222222222223, |
| "eval_accuracy": 0.6535433070866141, |
| "eval_auc": 0.6843118686868688, |
| "eval_f1": 0.696551724137931, |
| "eval_loss": 0.8344177603721619, |
| "eval_precision": 0.6917808219178082, |
| "eval_recall": 0.7013888888888888, |
| "eval_runtime": 17.2629, |
| "eval_samples_per_second": 14.714, |
| "eval_steps_per_second": 0.232, |
| "step": 500 |
| }, |
| { |
| "epoch": 3.5416666666666665, |
| "grad_norm": 40.0, |
| "learning_rate": 2.3751170983272e-06, |
| "loss": 0.2129, |
| "step": 510 |
| }, |
| { |
| "epoch": 3.611111111111111, |
| "grad_norm": 38.5, |
| "learning_rate": 2.171965622567308e-06, |
| "loss": 0.1615, |
| "step": 520 |
| }, |
| { |
| "epoch": 3.6805555555555554, |
| "grad_norm": 50.0, |
| "learning_rate": 1.9754599786922913e-06, |
| "loss": 0.1841, |
| "step": 530 |
| }, |
| { |
| "epoch": 3.75, |
| "grad_norm": 33.5, |
| "learning_rate": 1.7860619515673034e-06, |
| "loss": 0.1303, |
| "step": 540 |
| }, |
| { |
| "epoch": 3.8194444444444446, |
| "grad_norm": 23.875, |
| "learning_rate": 1.6042166232810346e-06, |
| "loss": 0.1689, |
| "step": 550 |
| }, |
| { |
| "epoch": 3.8194444444444446, |
| "eval_accuracy": 0.6732283464566929, |
| "eval_auc": 0.6888888888888889, |
| "eval_f1": 0.7087719298245614, |
| "eval_loss": 0.8513283133506775, |
| "eval_precision": 0.7163120567375887, |
| "eval_recall": 0.7013888888888888, |
| "eval_runtime": 17.2698, |
| "eval_samples_per_second": 14.708, |
| "eval_steps_per_second": 0.232, |
| "step": 550 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 720, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 50, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.1871759346771968e+18, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
