| { | |
| "best_metric": 0.7028411284769034, | |
| "best_model_checkpoint": "./nlu_finetuned_models/cola/roberta-large_lr1e-05/checkpoint-3848", | |
| "epoch": 10.0, | |
| "eval_steps": 500, | |
| "global_step": 4810, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.3714715838432312, | |
| "eval_matthews_correlation": 0.6178566076184573, | |
| "eval_runtime": 0.9181, | |
| "eval_samples_per_second": 932.378, | |
| "eval_steps_per_second": 58.818, | |
| "step": 481 | |
| }, | |
| { | |
| "epoch": 1.0395010395010396, | |
| "grad_norm": 31.42021942138672, | |
| "learning_rate": 9.533289095332891e-06, | |
| "loss": 0.5165, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.36615189909935, | |
| "eval_matthews_correlation": 0.672732946200059, | |
| "eval_runtime": 0.8969, | |
| "eval_samples_per_second": 954.433, | |
| "eval_steps_per_second": 60.21, | |
| "step": 962 | |
| }, | |
| { | |
| "epoch": 2.079002079002079, | |
| "grad_norm": 6.862401962280273, | |
| "learning_rate": 8.427339084273391e-06, | |
| "loss": 0.3468, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.39393267035484314, | |
| "eval_matthews_correlation": 0.6436994142086258, | |
| "eval_runtime": 0.8844, | |
| "eval_samples_per_second": 967.838, | |
| "eval_steps_per_second": 61.055, | |
| "step": 1443 | |
| }, | |
| { | |
| "epoch": 3.1185031185031185, | |
| "grad_norm": 8.505253791809082, | |
| "learning_rate": 7.3213890732138915e-06, | |
| "loss": 0.2448, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_loss": 0.5278908014297485, | |
| "eval_matthews_correlation": 0.7005174840601028, | |
| "eval_runtime": 0.8814, | |
| "eval_samples_per_second": 971.171, | |
| "eval_steps_per_second": 61.265, | |
| "step": 1924 | |
| }, | |
| { | |
| "epoch": 4.158004158004158, | |
| "grad_norm": 61.827049255371094, | |
| "learning_rate": 6.2154390621543915e-06, | |
| "loss": 0.1829, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_loss": 0.5846863389015198, | |
| "eval_matthews_correlation": 0.6887827880453191, | |
| "eval_runtime": 0.8956, | |
| "eval_samples_per_second": 955.76, | |
| "eval_steps_per_second": 60.293, | |
| "step": 2405 | |
| }, | |
| { | |
| "epoch": 5.197505197505198, | |
| "grad_norm": 0.20127160847187042, | |
| "learning_rate": 5.1094890510948916e-06, | |
| "loss": 0.149, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_loss": 0.7285173535346985, | |
| "eval_matthews_correlation": 0.6970512081270855, | |
| "eval_runtime": 0.88, | |
| "eval_samples_per_second": 972.692, | |
| "eval_steps_per_second": 61.361, | |
| "step": 2886 | |
| }, | |
| { | |
| "epoch": 6.237006237006237, | |
| "grad_norm": 0.9558518528938293, | |
| "learning_rate": 4.003539040035391e-06, | |
| "loss": 0.1009, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_loss": 0.8499833345413208, | |
| "eval_matthews_correlation": 0.6829154400379273, | |
| "eval_runtime": 0.8827, | |
| "eval_samples_per_second": 969.703, | |
| "eval_steps_per_second": 61.173, | |
| "step": 3367 | |
| }, | |
| { | |
| "epoch": 7.276507276507276, | |
| "grad_norm": 9.234942436218262, | |
| "learning_rate": 2.897589028975891e-06, | |
| "loss": 0.083, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 0.7806959748268127, | |
| "eval_matthews_correlation": 0.7028411284769034, | |
| "eval_runtime": 0.8826, | |
| "eval_samples_per_second": 969.849, | |
| "eval_steps_per_second": 61.182, | |
| "step": 3848 | |
| }, | |
| { | |
| "epoch": 8.316008316008316, | |
| "grad_norm": 0.022788817062973976, | |
| "learning_rate": 1.7916390179163902e-06, | |
| "loss": 0.056, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_loss": 0.8418498039245605, | |
| "eval_matthews_correlation": 0.7012056940021191, | |
| "eval_runtime": 0.8786, | |
| "eval_samples_per_second": 974.326, | |
| "eval_steps_per_second": 61.464, | |
| "step": 4329 | |
| }, | |
| { | |
| "epoch": 9.355509355509355, | |
| "grad_norm": 0.3841009736061096, | |
| "learning_rate": 6.856890068568902e-07, | |
| "loss": 0.051, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 0.8874779939651489, | |
| "eval_matthews_correlation": 0.6993251302364928, | |
| "eval_runtime": 0.8567, | |
| "eval_samples_per_second": 999.194, | |
| "eval_steps_per_second": 63.033, | |
| "step": 4810 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "step": 4810, | |
| "total_flos": 2856669958043580.0, | |
| "train_loss": 0.18191755675476454, | |
| "train_runtime": 542.0493, | |
| "train_samples_per_second": 141.961, | |
| "train_steps_per_second": 8.874 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4810, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "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": 2856669958043580.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |