| { | |
| "best_metric": 0.5783476233482361, | |
| "best_model_checkpoint": "bert_tiny_lda_100_v1_qnli/checkpoint-1640", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 3690, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 1.6202223300933838, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.6638, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.6251144060040271, | |
| "eval_loss": 0.6427455544471741, | |
| "eval_runtime": 1.7242, | |
| "eval_samples_per_second": 3168.366, | |
| "eval_steps_per_second": 12.759, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.5754749774932861, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.6373, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.6340838367197511, | |
| "eval_loss": 0.6345319151878357, | |
| "eval_runtime": 1.7183, | |
| "eval_samples_per_second": 3179.304, | |
| "eval_steps_per_second": 12.803, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.2697596549987793, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.5861, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.6809445359692476, | |
| "eval_loss": 0.5894870162010193, | |
| "eval_runtime": 1.746, | |
| "eval_samples_per_second": 3128.855, | |
| "eval_steps_per_second": 12.6, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.8899729251861572, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.4866, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.727622185612301, | |
| "eval_loss": 0.5783476233482361, | |
| "eval_runtime": 1.7484, | |
| "eval_samples_per_second": 3124.518, | |
| "eval_steps_per_second": 12.583, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 4.109302997589111, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.4086, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.7212154493867838, | |
| "eval_loss": 0.5973373651504517, | |
| "eval_runtime": 1.7308, | |
| "eval_samples_per_second": 3156.341, | |
| "eval_steps_per_second": 12.711, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 3.982114791870117, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.3472, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.7375068643602416, | |
| "eval_loss": 0.5946151614189148, | |
| "eval_runtime": 1.7169, | |
| "eval_samples_per_second": 3181.859, | |
| "eval_steps_per_second": 12.814, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 6.715934753417969, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.2941, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.7173714076514736, | |
| "eval_loss": 0.7428101301193237, | |
| "eval_runtime": 1.7169, | |
| "eval_samples_per_second": 3181.973, | |
| "eval_steps_per_second": 12.814, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.238851547241211, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.2498, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.738239062786015, | |
| "eval_loss": 0.6951647400856018, | |
| "eval_runtime": 1.7348, | |
| "eval_samples_per_second": 3149.124, | |
| "eval_steps_per_second": 12.682, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 4.454581260681152, | |
| "learning_rate": 4.1e-05, | |
| "loss": 0.209, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.729269632070291, | |
| "eval_loss": 0.7844876646995544, | |
| "eval_runtime": 1.7105, | |
| "eval_samples_per_second": 3193.794, | |
| "eval_steps_per_second": 12.862, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "step": 3690, | |
| "total_flos": 2.4720626179666944e+16, | |
| "train_loss": 0.43138372529812946, | |
| "train_runtime": 534.734, | |
| "train_samples_per_second": 9793.934, | |
| "train_steps_per_second": 38.337 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 20500, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 50, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "EarlyStoppingCallback": { | |
| "args": { | |
| "early_stopping_patience": 5, | |
| "early_stopping_threshold": 0.0 | |
| }, | |
| "attributes": { | |
| "early_stopping_patience_counter": 5 | |
| } | |
| }, | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.4720626179666944e+16, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |