| { | |
| "best_metric": 0.69140625, | |
| "best_model_checkpoint": "bert_base_lda_100_v1_qnli/checkpoint-1230", | |
| "epoch": 8.0, | |
| "eval_steps": 500, | |
| "global_step": 3280, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.23485292494297028, | |
| "learning_rate": 0.00098, | |
| "loss": 0.7139, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.4946000366099213, | |
| "eval_loss": 0.6933804750442505, | |
| "eval_runtime": 3.5205, | |
| "eval_samples_per_second": 1551.757, | |
| "eval_steps_per_second": 6.249, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.09086626023054123, | |
| "learning_rate": 0.00096, | |
| "loss": 0.6933, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.4946000366099213, | |
| "eval_loss": 0.6933804750442505, | |
| "eval_runtime": 3.45, | |
| "eval_samples_per_second": 1583.494, | |
| "eval_steps_per_second": 6.377, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.12760137021541595, | |
| "learning_rate": 0.00094, | |
| "loss": 0.6933, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5053999633900788, | |
| "eval_loss": 0.69140625, | |
| "eval_runtime": 3.4972, | |
| "eval_samples_per_second": 1562.095, | |
| "eval_steps_per_second": 6.291, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.01812651753425598, | |
| "learning_rate": 0.00092, | |
| "loss": 0.6931, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5053999633900788, | |
| "eval_loss": 0.6933382749557495, | |
| "eval_runtime": 3.4925, | |
| "eval_samples_per_second": 1564.198, | |
| "eval_steps_per_second": 6.299, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.051437895745038986, | |
| "learning_rate": 0.0009000000000000001, | |
| "loss": 0.6932, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.4946000366099213, | |
| "eval_loss": 0.6934648156166077, | |
| "eval_runtime": 3.4809, | |
| "eval_samples_per_second": 1569.414, | |
| "eval_steps_per_second": 6.32, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.05455218255519867, | |
| "learning_rate": 0.00088, | |
| "loss": 0.6932, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5053999633900788, | |
| "eval_loss": 0.6933382749557495, | |
| "eval_runtime": 3.5969, | |
| "eval_samples_per_second": 1518.792, | |
| "eval_steps_per_second": 6.116, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.06181074306368828, | |
| "learning_rate": 0.00086, | |
| "loss": 0.6933, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5053999633900788, | |
| "eval_loss": 0.6932539343833923, | |
| "eval_runtime": 3.6153, | |
| "eval_samples_per_second": 1511.066, | |
| "eval_steps_per_second": 6.085, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.13051092624664307, | |
| "learning_rate": 0.00084, | |
| "loss": 0.6934, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.4946000366099213, | |
| "eval_loss": 0.6934226751327515, | |
| "eval_runtime": 3.4839, | |
| "eval_samples_per_second": 1568.091, | |
| "eval_steps_per_second": 6.315, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "step": 3280, | |
| "total_flos": 1.1023616508628992e+17, | |
| "train_loss": 0.6958369836574647, | |
| "train_runtime": 1397.9533, | |
| "train_samples_per_second": 3746.298, | |
| "train_steps_per_second": 14.664 | |
| } | |
| ], | |
| "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": 1.1023616508628992e+17, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |