| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 100.0, | |
| "eval_steps": 300, | |
| "global_step": 3000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 10.267565727233887, | |
| "learning_rate": 4.5e-05, | |
| "loss": 2.6587, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "eval_loss": 2.672071695327759, | |
| "eval_runtime": 2.1303, | |
| "eval_samples_per_second": 336.103, | |
| "eval_steps_per_second": 3.755, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "grad_norm": 5.43134069442749, | |
| "learning_rate": 4e-05, | |
| "loss": 0.5242, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 20.0, | |
| "eval_loss": 1.995083212852478, | |
| "eval_runtime": 2.183, | |
| "eval_samples_per_second": 327.995, | |
| "eval_steps_per_second": 3.665, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "grad_norm": 2.2661261558532715, | |
| "learning_rate": 3.5e-05, | |
| "loss": 0.1995, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 30.0, | |
| "eval_loss": 1.776659369468689, | |
| "eval_runtime": 2.105, | |
| "eval_samples_per_second": 340.146, | |
| "eval_steps_per_second": 3.801, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "grad_norm": 1.7297999858856201, | |
| "learning_rate": 3e-05, | |
| "loss": 0.1025, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 40.0, | |
| "eval_loss": 1.6002683639526367, | |
| "eval_runtime": 2.1049, | |
| "eval_samples_per_second": 340.159, | |
| "eval_steps_per_second": 3.801, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "grad_norm": 1.119903326034546, | |
| "learning_rate": 2.5e-05, | |
| "loss": 0.0609, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 50.0, | |
| "eval_loss": 1.5019861459732056, | |
| "eval_runtime": 2.0694, | |
| "eval_samples_per_second": 345.998, | |
| "eval_steps_per_second": 3.866, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "grad_norm": 0.4384348690509796, | |
| "learning_rate": 2e-05, | |
| "loss": 0.042, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 60.0, | |
| "eval_loss": 1.3371723890304565, | |
| "eval_runtime": 2.1069, | |
| "eval_samples_per_second": 339.838, | |
| "eval_steps_per_second": 3.797, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "grad_norm": 0.4751300811767578, | |
| "learning_rate": 1.5e-05, | |
| "loss": 0.0315, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 70.0, | |
| "eval_loss": 1.3104065656661987, | |
| "eval_runtime": 2.0197, | |
| "eval_samples_per_second": 354.506, | |
| "eval_steps_per_second": 3.961, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "grad_norm": 1.0900623798370361, | |
| "learning_rate": 1e-05, | |
| "loss": 0.0271, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 80.0, | |
| "eval_loss": 1.2714661359786987, | |
| "eval_runtime": 2.1021, | |
| "eval_samples_per_second": 340.608, | |
| "eval_steps_per_second": 3.806, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "grad_norm": 0.31860601902008057, | |
| "learning_rate": 5e-06, | |
| "loss": 0.0212, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 90.0, | |
| "eval_loss": 1.2446495294570923, | |
| "eval_runtime": 2.1073, | |
| "eval_samples_per_second": 339.765, | |
| "eval_steps_per_second": 3.796, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "grad_norm": 0.22674699127674103, | |
| "learning_rate": 0.0, | |
| "loss": 0.0202, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "eval_loss": 1.237874984741211, | |
| "eval_runtime": 2.1055, | |
| "eval_samples_per_second": 340.065, | |
| "eval_steps_per_second": 3.8, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 100.0, | |
| "step": 3000, | |
| "total_flos": 3.1668214733568e+16, | |
| "train_loss": 0.36876122029622393, | |
| "train_runtime": 1639.2452, | |
| "train_samples_per_second": 146.043, | |
| "train_steps_per_second": 1.83 | |
| } | |
| ], | |
| "logging_steps": 300, | |
| "max_steps": 3000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 100, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3.1668214733568e+16, | |
| "train_batch_size": 80, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |