| { | |
| "best_global_step": 440, | |
| "best_metric": 0.7817979098493669, | |
| "best_model_checkpoint": "outputs/models/decision_model\\checkpoint-440", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 440, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.2840909090909091, | |
| "grad_norm": 2.0270469188690186, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 1.391720733642578, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.5681818181818182, | |
| "grad_norm": 2.6033480167388916, | |
| "learning_rate": 1.9747474747474747e-05, | |
| "loss": 1.3292826843261718, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.8522727272727273, | |
| "grad_norm": 3.2250678539276123, | |
| "learning_rate": 1.8484848484848487e-05, | |
| "loss": 0.9665608978271485, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6933333333333334, | |
| "eval_loss": 0.7237751483917236, | |
| "eval_macro_f1": 0.6931818181818181, | |
| "eval_macro_precision": 0.7383305573636584, | |
| "eval_macro_recall": 0.6907309258679122, | |
| "eval_runtime": 12.0023, | |
| "eval_samples_per_second": 24.995, | |
| "eval_steps_per_second": 0.833, | |
| "eval_weighted_f1": 0.696861471861472, | |
| "step": 88 | |
| }, | |
| { | |
| "epoch": 1.1363636363636362, | |
| "grad_norm": 5.046335697174072, | |
| "learning_rate": 1.7222222222222224e-05, | |
| "loss": 0.7137407684326171, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.4204545454545454, | |
| "grad_norm": 2.440394401550293, | |
| "learning_rate": 1.595959595959596e-05, | |
| "loss": 0.5940697860717773, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 1.7045454545454546, | |
| "grad_norm": 12.461901664733887, | |
| "learning_rate": 1.4696969696969699e-05, | |
| "loss": 0.5451191329956054, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.9886363636363638, | |
| "grad_norm": 6.133688449859619, | |
| "learning_rate": 1.3434343434343436e-05, | |
| "loss": 0.5883205032348633, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7466666666666667, | |
| "eval_loss": 0.6313713192939758, | |
| "eval_macro_f1": 0.7507497948674419, | |
| "eval_macro_precision": 0.7921349294988134, | |
| "eval_macro_recall": 0.7446015720673256, | |
| "eval_runtime": 11.3397, | |
| "eval_samples_per_second": 26.456, | |
| "eval_steps_per_second": 0.882, | |
| "eval_weighted_f1": 0.7538517764007959, | |
| "step": 176 | |
| }, | |
| { | |
| "epoch": 2.2727272727272725, | |
| "grad_norm": 5.935120582580566, | |
| "learning_rate": 1.2171717171717172e-05, | |
| "loss": 0.46342819213867187, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.5568181818181817, | |
| "grad_norm": 7.027043342590332, | |
| "learning_rate": 1.0909090909090909e-05, | |
| "loss": 0.4100628662109375, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 2.840909090909091, | |
| "grad_norm": 8.357532501220703, | |
| "learning_rate": 9.646464646464647e-06, | |
| "loss": 0.39318439483642575, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7666666666666667, | |
| "eval_loss": 0.5723195672035217, | |
| "eval_macro_f1": 0.7632203915397024, | |
| "eval_macro_precision": 0.7755475675118532, | |
| "eval_macro_recall": 0.7631792593436428, | |
| "eval_runtime": 9.3457, | |
| "eval_samples_per_second": 32.1, | |
| "eval_steps_per_second": 1.07, | |
| "eval_weighted_f1": 0.7663719158480559, | |
| "step": 264 | |
| }, | |
| { | |
| "epoch": 3.125, | |
| "grad_norm": 6.643225193023682, | |
| "learning_rate": 8.383838383838384e-06, | |
| "loss": 0.41530742645263674, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 3.409090909090909, | |
| "grad_norm": 3.8116602897644043, | |
| "learning_rate": 7.121212121212122e-06, | |
| "loss": 0.30169292449951174, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.6931818181818183, | |
| "grad_norm": 5.07578182220459, | |
| "learning_rate": 5.858585858585859e-06, | |
| "loss": 0.34119693756103514, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 3.9772727272727275, | |
| "grad_norm": 5.61444616317749, | |
| "learning_rate": 4.595959595959597e-06, | |
| "loss": 0.34068592071533205, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7533333333333333, | |
| "eval_loss": 0.6002340316772461, | |
| "eval_macro_f1": 0.7527508236810562, | |
| "eval_macro_precision": 0.7775673527557759, | |
| "eval_macro_recall": 0.7493573605217441, | |
| "eval_runtime": 9.4128, | |
| "eval_samples_per_second": 31.872, | |
| "eval_steps_per_second": 1.062, | |
| "eval_weighted_f1": 0.7562322082167043, | |
| "step": 352 | |
| }, | |
| { | |
| "epoch": 4.261363636363637, | |
| "grad_norm": 14.103662490844727, | |
| "learning_rate": 3.3333333333333333e-06, | |
| "loss": 0.36039134979248044, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 4.545454545454545, | |
| "grad_norm": 10.582133293151855, | |
| "learning_rate": 2.070707070707071e-06, | |
| "loss": 0.2752532386779785, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.829545454545455, | |
| "grad_norm": 10.563464164733887, | |
| "learning_rate": 8.080808080808082e-07, | |
| "loss": 0.21057830810546874, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7833333333333333, | |
| "eval_loss": 0.5563565492630005, | |
| "eval_macro_f1": 0.7817979098493669, | |
| "eval_macro_precision": 0.7985651345558467, | |
| "eval_macro_recall": 0.7799016033947541, | |
| "eval_runtime": 11.0713, | |
| "eval_samples_per_second": 27.097, | |
| "eval_steps_per_second": 0.903, | |
| "eval_weighted_f1": 0.7847290985647283, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 440, | |
| "total_flos": 180259372028736.0, | |
| "train_loss": 0.5567306355996565, | |
| "train_runtime": 1081.1026, | |
| "train_samples_per_second": 6.475, | |
| "train_steps_per_second": 0.407 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 440, | |
| "test_accuracy": 0.83, | |
| "test_loss": 0.4470374584197998, | |
| "test_macro_f1": 0.8242622940008544, | |
| "test_macro_precision": 0.8358748568155785, | |
| "test_macro_recall": 0.8268016594043991, | |
| "test_runtime": 10.3642, | |
| "test_samples_per_second": 28.946, | |
| "test_steps_per_second": 0.965, | |
| "test_weighted_f1": 0.8269116167241574 | |
| } | |
| ], | |
| "logging_steps": 25, | |
| "max_steps": 440, | |
| "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": 180259372028736.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |