| { | |
| "best_metric": 0.6863997578620911, | |
| "best_model_checkpoint": "bert_base_lda_wnli/checkpoint-39", | |
| "epoch": 18.0, | |
| "eval_steps": 500, | |
| "global_step": 54, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 18.621891021728516, | |
| "learning_rate": 0.00098, | |
| "loss": 1.1638, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 1.4607518911361694, | |
| "eval_runtime": 0.056, | |
| "eval_samples_per_second": 1268.435, | |
| "eval_steps_per_second": 17.865, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 6.636824131011963, | |
| "learning_rate": 0.00096, | |
| "loss": 1.0198, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 1.109704852104187, | |
| "eval_runtime": 0.0609, | |
| "eval_samples_per_second": 1164.911, | |
| "eval_steps_per_second": 16.407, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 2.0930373668670654, | |
| "learning_rate": 0.00094, | |
| "loss": 1.1474, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.8995377421379089, | |
| "eval_runtime": 0.0517, | |
| "eval_samples_per_second": 1372.362, | |
| "eval_steps_per_second": 19.329, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 3.016080617904663, | |
| "learning_rate": 0.00092, | |
| "loss": 0.8846, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.43661971830985913, | |
| "eval_loss": 0.820120096206665, | |
| "eval_runtime": 0.0532, | |
| "eval_samples_per_second": 1335.808, | |
| "eval_steps_per_second": 18.814, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 2.1296448707580566, | |
| "learning_rate": 0.0009000000000000001, | |
| "loss": 0.7886, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.43661971830985913, | |
| "eval_loss": 0.6994389295578003, | |
| "eval_runtime": 0.0527, | |
| "eval_samples_per_second": 1348.431, | |
| "eval_steps_per_second": 18.992, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 2.4899051189422607, | |
| "learning_rate": 0.00088, | |
| "loss": 0.738, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.708736777305603, | |
| "eval_runtime": 0.0518, | |
| "eval_samples_per_second": 1369.427, | |
| "eval_steps_per_second": 19.288, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.7382462024688721, | |
| "learning_rate": 0.00086, | |
| "loss": 0.7195, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.43661971830985913, | |
| "eval_loss": 0.7214460372924805, | |
| "eval_runtime": 0.0517, | |
| "eval_samples_per_second": 1372.273, | |
| "eval_steps_per_second": 19.328, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 0.9765424132347107, | |
| "learning_rate": 0.00084, | |
| "loss": 0.7036, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.693111777305603, | |
| "eval_runtime": 0.0536, | |
| "eval_samples_per_second": 1325.279, | |
| "eval_steps_per_second": 18.666, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.2757832407951355, | |
| "learning_rate": 0.00082, | |
| "loss": 0.6935, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.689645528793335, | |
| "eval_runtime": 0.0517, | |
| "eval_samples_per_second": 1372.33, | |
| "eval_steps_per_second": 19.329, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.41503846645355225, | |
| "learning_rate": 0.0008, | |
| "loss": 0.6941, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6926164627075195, | |
| "eval_runtime": 0.0542, | |
| "eval_samples_per_second": 1310.789, | |
| "eval_steps_per_second": 18.462, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 0.1713160276412964, | |
| "learning_rate": 0.0007800000000000001, | |
| "loss": 0.6949, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.43661971830985913, | |
| "eval_loss": 0.693606972694397, | |
| "eval_runtime": 0.0536, | |
| "eval_samples_per_second": 1325.279, | |
| "eval_steps_per_second": 18.666, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 0.817528247833252, | |
| "learning_rate": 0.00076, | |
| "loss": 0.6959, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6911312341690063, | |
| "eval_runtime": 0.0527, | |
| "eval_samples_per_second": 1346.784, | |
| "eval_steps_per_second": 18.969, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 0.3534504175186157, | |
| "learning_rate": 0.00074, | |
| "loss": 0.6927, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6863997578620911, | |
| "eval_runtime": 0.0533, | |
| "eval_samples_per_second": 1330.984, | |
| "eval_steps_per_second": 18.746, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "grad_norm": 0.8052321076393127, | |
| "learning_rate": 0.0007199999999999999, | |
| "loss": 0.6928, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 14.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6893155574798584, | |
| "eval_runtime": 0.0524, | |
| "eval_samples_per_second": 1353.85, | |
| "eval_steps_per_second": 19.068, | |
| "step": 42 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "grad_norm": 0.29218751192092896, | |
| "learning_rate": 0.0007, | |
| "loss": 0.6958, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 15.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.689645528793335, | |
| "eval_runtime": 0.0518, | |
| "eval_samples_per_second": 1371.091, | |
| "eval_steps_per_second": 19.311, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 0.19813981652259827, | |
| "learning_rate": 0.00068, | |
| "loss": 0.6936, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6911312341690063, | |
| "eval_runtime": 0.0523, | |
| "eval_samples_per_second": 1358.408, | |
| "eval_steps_per_second": 19.133, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "grad_norm": 0.6377934217453003, | |
| "learning_rate": 0.00066, | |
| "loss": 0.6955, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 17.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6911312341690063, | |
| "eval_runtime": 0.0549, | |
| "eval_samples_per_second": 1292.678, | |
| "eval_steps_per_second": 18.207, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "grad_norm": 0.3209940493106842, | |
| "learning_rate": 0.00064, | |
| "loss": 0.6939, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6906359791755676, | |
| "eval_runtime": 0.1325, | |
| "eval_samples_per_second": 535.893, | |
| "eval_steps_per_second": 7.548, | |
| "step": 54 | |
| }, | |
| { | |
| "epoch": 18.0, | |
| "step": 54, | |
| "total_flos": 1503679681382400.0, | |
| "train_loss": 0.7837636250036734, | |
| "train_runtime": 61.4069, | |
| "train_samples_per_second": 517.043, | |
| "train_steps_per_second": 2.443 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 150, | |
| "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": 1503679681382400.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |