| { | |
| "best_metric": 0.541095890410959, | |
| "best_model_checkpoint": "tiny-bert-sst2-distilled/run-5/checkpoint-402", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 670, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.22246789932250977, | |
| "learning_rate": 1.5679734473172616e-05, | |
| "loss": 0.2458, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.1889893114566803, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 28.3905, | |
| "eval_samples_per_second": 35.998, | |
| "eval_steps_per_second": 1.127, | |
| "step": 67 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.2654605507850647, | |
| "learning_rate": 1.3937541753931215e-05, | |
| "loss": 0.1861, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5, | |
| "eval_f1": 0.0, | |
| "eval_loss": 0.17826829850673676, | |
| "eval_precision": 0.0, | |
| "eval_recall": 0.0, | |
| "eval_runtime": 29.9866, | |
| "eval_samples_per_second": 34.082, | |
| "eval_steps_per_second": 1.067, | |
| "step": 134 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.32397571206092834, | |
| "learning_rate": 1.2195349034689811e-05, | |
| "loss": 0.1759, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5107632093933464, | |
| "eval_f1": 0.07063197026022304, | |
| "eval_loss": 0.16691070795059204, | |
| "eval_precision": 0.7037037037037037, | |
| "eval_recall": 0.03718199608610567, | |
| "eval_runtime": 28.5351, | |
| "eval_samples_per_second": 35.816, | |
| "eval_steps_per_second": 1.121, | |
| "step": 201 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.6136437058448792, | |
| "learning_rate": 1.045315631544841e-05, | |
| "loss": 0.168, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5362035225048923, | |
| "eval_f1": 0.18556701030927833, | |
| "eval_loss": 0.160459965467453, | |
| "eval_precision": 0.7605633802816901, | |
| "eval_recall": 0.10567514677103718, | |
| "eval_runtime": 28.6402, | |
| "eval_samples_per_second": 35.684, | |
| "eval_steps_per_second": 1.117, | |
| "step": 268 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.3216162919998169, | |
| "learning_rate": 8.710963596207009e-06, | |
| "loss": 0.1647, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5401174168297456, | |
| "eval_f1": 0.20068027210884357, | |
| "eval_loss": 0.1576094627380371, | |
| "eval_precision": 0.7662337662337663, | |
| "eval_recall": 0.11545988258317025, | |
| "eval_runtime": 28.3765, | |
| "eval_samples_per_second": 36.016, | |
| "eval_steps_per_second": 1.128, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.45282211899757385, | |
| "learning_rate": 6.968770876965607e-06, | |
| "loss": 0.1625, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.15556302666664124, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 28.8112, | |
| "eval_samples_per_second": 35.472, | |
| "eval_steps_per_second": 1.111, | |
| "step": 402 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.3418940007686615, | |
| "learning_rate": 5.226578157724205e-06, | |
| "loss": 0.1611, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.15463578701019287, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 28.8835, | |
| "eval_samples_per_second": 35.384, | |
| "eval_steps_per_second": 1.108, | |
| "step": 469 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.3419853448867798, | |
| "learning_rate": 3.4843854384828036e-06, | |
| "loss": 0.1596, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.15391579270362854, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 28.6467, | |
| "eval_samples_per_second": 35.676, | |
| "eval_steps_per_second": 1.117, | |
| "step": 536 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.3018151521682739, | |
| "learning_rate": 1.7421927192414018e-06, | |
| "loss": 0.1589, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.15353241562843323, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 28.7082, | |
| "eval_samples_per_second": 35.6, | |
| "eval_steps_per_second": 1.115, | |
| "step": 603 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.4547845721244812, | |
| "learning_rate": 0.0, | |
| "loss": 0.1592, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.541095890410959, | |
| "eval_f1": 0.20373514431239387, | |
| "eval_loss": 0.15339058637619019, | |
| "eval_precision": 0.7692307692307693, | |
| "eval_recall": 0.11741682974559686, | |
| "eval_runtime": 28.3962, | |
| "eval_samples_per_second": 35.991, | |
| "eval_steps_per_second": 1.127, | |
| "step": 670 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 670, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 2356951972800.0, | |
| "train_batch_size": 46, | |
| "trial_name": null, | |
| "trial_params": { | |
| "alpha": 0.19075569878013487, | |
| "learning_rate": 1.7421927192414017e-05, | |
| "num_train_epochs": 10, | |
| "per_device_train_batch_size": 46, | |
| "temperature": 24 | |
| } | |
| } | |