| { | |
| "best_metric": 1.2710014581680298, | |
| "best_model_checkpoint": "/content/pubmedbert-fulltext-cord19/checkpoint-5000", | |
| "epoch": 3.0, | |
| "global_step": 56250, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 2.4970000000000003e-05, | |
| "loss": 1.7985, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "eval_accuracy": 0.7176016320015706, | |
| "eval_loss": 1.2710014581680298, | |
| "eval_runtime": 289.9232, | |
| "eval_samples_per_second": 86.23, | |
| "eval_steps_per_second": 5.391, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "learning_rate": 4.9965e-05, | |
| "loss": 1.7542, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "eval_accuracy": 0.7069832712721614, | |
| "eval_loss": 1.3359251022338867, | |
| "eval_runtime": 290.2058, | |
| "eval_samples_per_second": 86.146, | |
| "eval_steps_per_second": 5.386, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "learning_rate": 4.460432432432432e-05, | |
| "loss": 1.7462, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "eval_accuracy": 0.7034289815931059, | |
| "eval_loss": 1.3488777875900269, | |
| "eval_runtime": 290.0812, | |
| "eval_samples_per_second": 86.183, | |
| "eval_steps_per_second": 5.388, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 3.9202162162162166e-05, | |
| "loss": 1.8371, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "eval_accuracy": 0.6891450706903661, | |
| "eval_loss": 1.436103105545044, | |
| "eval_runtime": 289.9321, | |
| "eval_samples_per_second": 86.227, | |
| "eval_steps_per_second": 5.391, | |
| "step": 20000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "learning_rate": 3.379675675675676e-05, | |
| "loss": 1.7102, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "eval_accuracy": 0.7039453286069087, | |
| "eval_loss": 1.3501886129379272, | |
| "eval_runtime": 289.995, | |
| "eval_samples_per_second": 86.208, | |
| "eval_steps_per_second": 5.39, | |
| "step": 25000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "learning_rate": 2.8393513513513515e-05, | |
| "loss": 1.6596, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "eval_accuracy": 0.70652987764824, | |
| "eval_loss": 1.334069013595581, | |
| "eval_runtime": 290.2224, | |
| "eval_samples_per_second": 86.141, | |
| "eval_steps_per_second": 5.386, | |
| "step": 30000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "learning_rate": 2.2990270270270274e-05, | |
| "loss": 1.6265, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "eval_accuracy": 0.708723656830319, | |
| "eval_loss": 1.322791576385498, | |
| "eval_runtime": 291.0022, | |
| "eval_samples_per_second": 85.91, | |
| "eval_steps_per_second": 5.371, | |
| "step": 35000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "learning_rate": 1.7584864864864868e-05, | |
| "loss": 1.605, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "eval_accuracy": 0.7099296658889697, | |
| "eval_loss": 1.307939052581787, | |
| "eval_runtime": 290.1974, | |
| "eval_samples_per_second": 86.148, | |
| "eval_steps_per_second": 5.386, | |
| "step": 40000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "learning_rate": 1.2181621621621623e-05, | |
| "loss": 1.5731, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "eval_accuracy": 0.7121370440536863, | |
| "eval_loss": 1.2985501289367676, | |
| "eval_runtime": 290.9835, | |
| "eval_samples_per_second": 85.916, | |
| "eval_steps_per_second": 5.371, | |
| "step": 45000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "learning_rate": 6.778378378378379e-06, | |
| "loss": 1.5602, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "eval_accuracy": 0.7135782251340592, | |
| "eval_loss": 1.2928680181503296, | |
| "eval_runtime": 290.2984, | |
| "eval_samples_per_second": 86.118, | |
| "eval_steps_per_second": 5.384, | |
| "step": 50000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "learning_rate": 1.3762162162162162e-06, | |
| "loss": 1.5447, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "eval_accuracy": 0.7142735090199965, | |
| "eval_loss": 1.2874841690063477, | |
| "eval_runtime": 290.4326, | |
| "eval_samples_per_second": 86.078, | |
| "eval_steps_per_second": 5.382, | |
| "step": 55000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 56250, | |
| "total_flos": 1.156661775e+17, | |
| "train_loss": 1.6709515473090277, | |
| "train_runtime": 32029.5992, | |
| "train_samples_per_second": 28.099, | |
| "train_steps_per_second": 1.756 | |
| } | |
| ], | |
| "max_steps": 56250, | |
| "num_train_epochs": 3, | |
| "total_flos": 1.156661775e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |