| { | |
| "best_metric": 0.8228962818003914, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-0/checkpoint-768", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 768, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.8004393577575684, | |
| "learning_rate": 0.0007697157586205755, | |
| "loss": 0.4935, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.7436399217221135, | |
| "eval_f1": 0.7689594356261024, | |
| "eval_loss": 0.448881596326828, | |
| "eval_precision": 0.6998394863563403, | |
| "eval_recall": 0.8532289628180039, | |
| "eval_runtime": 28.6476, | |
| "eval_samples_per_second": 35.675, | |
| "eval_steps_per_second": 1.117, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 4.9879961013793945, | |
| "learning_rate": 0.0006841917854405115, | |
| "loss": 0.4571, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.7896281800391389, | |
| "eval_f1": 0.8057813911472448, | |
| "eval_loss": 0.42557916045188904, | |
| "eval_precision": 0.7483221476510067, | |
| "eval_recall": 0.87279843444227, | |
| "eval_runtime": 28.772, | |
| "eval_samples_per_second": 35.521, | |
| "eval_steps_per_second": 1.112, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.279979705810547, | |
| "learning_rate": 0.0005986678122604475, | |
| "loss": 0.4337, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.7945205479452054, | |
| "eval_f1": 0.8059149722735675, | |
| "eval_loss": 0.4328659176826477, | |
| "eval_precision": 0.7635726795096323, | |
| "eval_recall": 0.8532289628180039, | |
| "eval_runtime": 28.2776, | |
| "eval_samples_per_second": 36.142, | |
| "eval_steps_per_second": 1.132, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 2.231189727783203, | |
| "learning_rate": 0.0005131438390803836, | |
| "loss": 0.4235, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.7690802348336595, | |
| "eval_f1": 0.7704280155642023, | |
| "eval_loss": 0.45082738995552063, | |
| "eval_precision": 0.7659574468085106, | |
| "eval_recall": 0.7749510763209393, | |
| "eval_runtime": 29.4798, | |
| "eval_samples_per_second": 34.668, | |
| "eval_steps_per_second": 1.085, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.6898621320724487, | |
| "learning_rate": 0.00042761986590031967, | |
| "loss": 0.4136, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.8033268101761253, | |
| "eval_f1": 0.814404432132964, | |
| "eval_loss": 0.41852009296417236, | |
| "eval_precision": 0.7709790209790209, | |
| "eval_recall": 0.863013698630137, | |
| "eval_runtime": 29.6952, | |
| "eval_samples_per_second": 34.416, | |
| "eval_steps_per_second": 1.078, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 1.360769510269165, | |
| "learning_rate": 0.00034209589272025573, | |
| "loss": 0.405, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.8043052837573386, | |
| "eval_f1": 0.8204667863554758, | |
| "eval_loss": 0.4140284061431885, | |
| "eval_precision": 0.75787728026534, | |
| "eval_recall": 0.8943248532289628, | |
| "eval_runtime": 28.8106, | |
| "eval_samples_per_second": 35.473, | |
| "eval_steps_per_second": 1.111, | |
| "step": 576 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 1.9378271102905273, | |
| "learning_rate": 0.0002565719195401918, | |
| "loss": 0.3961, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.815068493150685, | |
| "eval_f1": 0.8301886792452831, | |
| "eval_loss": 0.4103885591030121, | |
| "eval_precision": 0.7674418604651163, | |
| "eval_recall": 0.9041095890410958, | |
| "eval_runtime": 29.2823, | |
| "eval_samples_per_second": 34.902, | |
| "eval_steps_per_second": 1.093, | |
| "step": 672 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.5738866329193115, | |
| "learning_rate": 0.00017104794636012787, | |
| "loss": 0.3931, | |
| "step": 768 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.8228962818003914, | |
| "eval_f1": 0.8340971585701191, | |
| "eval_loss": 0.39944925904273987, | |
| "eval_precision": 0.7844827586206896, | |
| "eval_recall": 0.8904109589041096, | |
| "eval_runtime": 29.1614, | |
| "eval_samples_per_second": 35.046, | |
| "eval_steps_per_second": 1.097, | |
| "step": 768 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 960, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 1885561578240.0, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.735211359802956, | |
| "learning_rate": 0.0008552397318006393, | |
| "num_train_epochs": 10, | |
| "temperature": 25 | |
| } | |
| } | |