| { | |
| "best_metric": 0.6814756989479065, | |
| "best_model_checkpoint": "bert_base_lda_20_v1_wnli/checkpoint-24", | |
| "epoch": 13.0, | |
| "eval_steps": 500, | |
| "global_step": 39, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 2.379216432571411, | |
| "learning_rate": 4.9e-05, | |
| "loss": 0.8433, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.43661971830985913, | |
| "eval_loss": 0.733467161655426, | |
| "eval_runtime": 0.0534, | |
| "eval_samples_per_second": 1329.611, | |
| "eval_steps_per_second": 18.727, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 5.9057841300964355, | |
| "learning_rate": 4.8e-05, | |
| "loss": 0.7419, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.43661971830985913, | |
| "eval_loss": 0.7789943218231201, | |
| "eval_runtime": 0.0547, | |
| "eval_samples_per_second": 1298.937, | |
| "eval_steps_per_second": 18.295, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 3.7641913890838623, | |
| "learning_rate": 4.7e-05, | |
| "loss": 0.7237, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6985859870910645, | |
| "eval_runtime": 0.0566, | |
| "eval_samples_per_second": 1254.04, | |
| "eval_steps_per_second": 17.663, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 1.4714839458465576, | |
| "learning_rate": 4.600000000000001e-05, | |
| "loss": 0.7156, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.4507042253521127, | |
| "eval_loss": 0.7405645847320557, | |
| "eval_runtime": 0.0626, | |
| "eval_samples_per_second": 1134.494, | |
| "eval_steps_per_second": 15.979, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.7439093589782715, | |
| "learning_rate": 4.5e-05, | |
| "loss": 0.7143, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6858494281768799, | |
| "eval_runtime": 0.0533, | |
| "eval_samples_per_second": 1330.936, | |
| "eval_steps_per_second": 18.746, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.680374264717102, | |
| "learning_rate": 4.4000000000000006e-05, | |
| "loss": 0.6989, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6855193376541138, | |
| "eval_runtime": 0.0533, | |
| "eval_samples_per_second": 1332.037, | |
| "eval_steps_per_second": 18.761, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "grad_norm": 0.9396877884864807, | |
| "learning_rate": 4.3e-05, | |
| "loss": 0.6949, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_accuracy": 0.5070422535211268, | |
| "eval_loss": 0.693166971206665, | |
| "eval_runtime": 0.0529, | |
| "eval_samples_per_second": 1343.406, | |
| "eval_steps_per_second": 18.921, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 2.6080543994903564, | |
| "learning_rate": 4.2e-05, | |
| "loss": 0.6949, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_accuracy": 0.5633802816901409, | |
| "eval_loss": 0.6814756989479065, | |
| "eval_runtime": 0.0551, | |
| "eval_samples_per_second": 1288.171, | |
| "eval_steps_per_second": 18.143, | |
| "step": 24 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "grad_norm": 0.7388402223587036, | |
| "learning_rate": 4.1e-05, | |
| "loss": 0.6963, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_accuracy": 0.49295774647887325, | |
| "eval_loss": 0.696027934551239, | |
| "eval_runtime": 0.0535, | |
| "eval_samples_per_second": 1327.027, | |
| "eval_steps_per_second": 18.691, | |
| "step": 27 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.8557111620903015, | |
| "learning_rate": 4e-05, | |
| "loss": 0.6932, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_accuracy": 0.5352112676056338, | |
| "eval_loss": 0.6830435991287231, | |
| "eval_runtime": 0.0551, | |
| "eval_samples_per_second": 1288.116, | |
| "eval_steps_per_second": 18.142, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "grad_norm": 1.1222957372665405, | |
| "learning_rate": 3.9000000000000006e-05, | |
| "loss": 0.6989, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 11.0, | |
| "eval_accuracy": 0.5352112676056338, | |
| "eval_loss": 0.6837586760520935, | |
| "eval_runtime": 0.0534, | |
| "eval_samples_per_second": 1330.448, | |
| "eval_steps_per_second": 18.739, | |
| "step": 33 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "grad_norm": 1.9935102462768555, | |
| "learning_rate": 3.8e-05, | |
| "loss": 0.6969, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 12.0, | |
| "eval_accuracy": 0.49295774647887325, | |
| "eval_loss": 0.7047203779220581, | |
| "eval_runtime": 0.0537, | |
| "eval_samples_per_second": 1321.152, | |
| "eval_steps_per_second": 18.608, | |
| "step": 36 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "grad_norm": 1.0254613161087036, | |
| "learning_rate": 3.7e-05, | |
| "loss": 0.7032, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "eval_accuracy": 0.5070422535211268, | |
| "eval_loss": 0.6938820481300354, | |
| "eval_runtime": 0.0538, | |
| "eval_samples_per_second": 1319.104, | |
| "eval_steps_per_second": 18.579, | |
| "step": 39 | |
| }, | |
| { | |
| "epoch": 13.0, | |
| "step": 39, | |
| "total_flos": 1085990880998400.0, | |
| "train_loss": 0.7166132437877166, | |
| "train_runtime": 48.6385, | |
| "train_samples_per_second": 652.775, | |
| "train_steps_per_second": 3.084 | |
| } | |
| ], | |
| "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": 1085990880998400.0, | |
| "train_batch_size": 256, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |