| { | |
| "best_metric": 0.691815197467804, | |
| "best_model_checkpoint": "bert_base_lda_5_v1_rte/checkpoint-110", | |
| "epoch": 16.0, | |
| "eval_steps": 500, | |
| "global_step": 160, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.441200315952301, | |
| "learning_rate": 0.00098, | |
| "loss": 1.0506, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.7786834239959717, | |
| "eval_runtime": 0.1826, | |
| "eval_samples_per_second": 1517.252, | |
| "eval_steps_per_second": 10.955, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 2.4710278511047363, | |
| "learning_rate": 0.00096, | |
| "loss": 0.7306, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.8054137229919434, | |
| "eval_runtime": 0.1909, | |
| "eval_samples_per_second": 1450.764, | |
| "eval_steps_per_second": 10.475, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.8589054942131042, | |
| "learning_rate": 0.00094, | |
| "loss": 0.7253, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5234657039711191, | |
| "eval_loss": 0.6984149217605591, | |
| "eval_runtime": 0.1857, | |
| "eval_samples_per_second": 1492.0, | |
| "eval_steps_per_second": 10.773, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.399823933839798, | |
| "learning_rate": 0.00092, | |
| "loss": 0.7014, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.6923157572746277, | |
| "eval_runtime": 0.1846, | |
| "eval_samples_per_second": 1500.535, | |
| "eval_steps_per_second": 10.834, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.2132512331008911, | |
| "learning_rate": 0.0009000000000000001, | |
| "loss": 0.7016, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.51985559566787, | |
| "eval_loss": 0.7043237090110779, | |
| "eval_runtime": 0.1943, | |
| "eval_samples_per_second": 1425.341, | |
| "eval_steps_per_second": 10.291, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.2804683446884155, | |
| "learning_rate": 0.00088, | |
| "loss": 0.6998, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6968496441841125, | |
| "eval_runtime": 0.187, | |
| "eval_samples_per_second": 1481.54, | |
| "eval_steps_per_second": 10.697, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.1722436547279358, | |
| "learning_rate": 0.00086, | |
| "loss": 0.6943, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.51985559566787, | |
| "eval_loss": 0.6937753558158875, | |
| "eval_runtime": 0.1875, | |
| "eval_samples_per_second": 1477.581, | |
| "eval_steps_per_second": 10.668, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.15857921540737152, | |
| "learning_rate": 0.00084, | |
| "loss": 0.6937, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.695806086063385, | |
| "eval_runtime": 0.1858, | |
| "eval_samples_per_second": 1490.666, | |
| "eval_steps_per_second": 10.763, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.1902921497821808, | |
| "learning_rate": 0.00082, | |
| "loss": 0.6962, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.48014440433212996, | |
| "eval_loss": 0.6922100782394409, | |
| "eval_runtime": 0.1902, | |
| "eval_samples_per_second": 1456.246, | |
| "eval_steps_per_second": 10.514, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.2601146101951599, | |
| "learning_rate": 0.0008, | |
| "loss": 0.6955, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6962996125221252, | |
| "eval_runtime": 0.1893, | |
| "eval_samples_per_second": 1463.542, | |
| "eval_steps_per_second": 10.567, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.403894305229187, | |
| "learning_rate": 0.0007800000000000001, | |
| "loss": 0.6948, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.5270758122743683, | |
| "eval_loss": 0.691815197467804, | |
| "eval_runtime": 0.1877, | |
| "eval_samples_per_second": 1475.946, | |
| "eval_steps_per_second": 10.657, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.07746425271034241, | |
| "learning_rate": 0.00076, | |
| "loss": 0.6931, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.51985559566787, | |
| "eval_loss": 0.6937612891197205, | |
| "eval_runtime": 0.1885, | |
| "eval_samples_per_second": 1469.253, | |
| "eval_steps_per_second": 10.608, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.0822499617934227, | |
| "learning_rate": 0.00074, | |
| "loss": 0.6932, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6948471069335938, | |
| "eval_runtime": 0.1857, | |
| "eval_samples_per_second": 1491.974, | |
| "eval_steps_per_second": 10.772, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.06322963535785675, | |
| "learning_rate": 0.0007199999999999999, | |
| "loss": 0.6934, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6921818852424622, | |
| "eval_runtime": 0.1873, | |
| "eval_samples_per_second": 1478.54, | |
| "eval_steps_per_second": 10.675, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.08335640281438828, | |
| "learning_rate": 0.0007, | |
| "loss": 0.6933, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.4729241877256318, | |
| "eval_loss": 0.6936061382293701, | |
| "eval_runtime": 0.1851, | |
| "eval_samples_per_second": 1496.787, | |
| "eval_steps_per_second": 10.807, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.07828615605831146, | |
| "learning_rate": 0.00068, | |
| "loss": 0.6933, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.51985559566787, | |
| "eval_loss": 0.6943253874778748, | |
| "eval_runtime": 0.1872, | |
| "eval_samples_per_second": 1479.887, | |
| "eval_steps_per_second": 10.685, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "step": 160, | |
| "total_flos": 5241172222771200.0, | |
| "train_loss": 0.7218841880559921, | |
| "train_runtime": 103.9484, | |
| "train_samples_per_second": 1197.71, | |
| "train_steps_per_second": 4.81 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 500, | |
| "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": 5241172222771200.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |