| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.9767441860465116, | |
| "eval_steps": 500, | |
| "global_step": 192, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.15503875968992248, | |
| "grad_norm": 27.14871976634167, | |
| "learning_rate": 5e-06, | |
| "loss": 0.6931, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31007751937984496, | |
| "grad_norm": 14.215725662461802, | |
| "learning_rate": 1e-05, | |
| "loss": 0.6074, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.46511627906976744, | |
| "grad_norm": 8.789668411329039, | |
| "learning_rate": 9.91682838414733e-06, | |
| "loss": 0.563, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.6201550387596899, | |
| "grad_norm": 8.656020918037868, | |
| "learning_rate": 9.670080543662742e-06, | |
| "loss": 0.6109, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.7751937984496124, | |
| "grad_norm": 9.508510861756253, | |
| "learning_rate": 9.267965445186733e-06, | |
| "loss": 0.598, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.9302325581395349, | |
| "grad_norm": 7.438616074276716, | |
| "learning_rate": 8.72386091371891e-06, | |
| "loss": 0.5954, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.0852713178294573, | |
| "grad_norm": 11.43708438878541, | |
| "learning_rate": 8.055868570489247e-06, | |
| "loss": 0.3699, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.2403100775193798, | |
| "grad_norm": 17.52083315465849, | |
| "learning_rate": 7.286211616523193e-06, | |
| "loss": 0.1977, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.3953488372093024, | |
| "grad_norm": 13.876316165599672, | |
| "learning_rate": 6.440495496826189e-06, | |
| "loss": 0.2065, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.550387596899225, | |
| "grad_norm": 8.884685401439247, | |
| "learning_rate": 5.546856041889374e-06, | |
| "loss": 0.167, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.7054263565891472, | |
| "grad_norm": 14.293184788635473, | |
| "learning_rate": 4.635023426695462e-06, | |
| "loss": 0.1753, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.8604651162790697, | |
| "grad_norm": 11.602070326403544, | |
| "learning_rate": 3.7353330880415963e-06, | |
| "loss": 0.1473, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.0155038759689923, | |
| "grad_norm": 5.538956396622884, | |
| "learning_rate": 2.8777165056209256e-06, | |
| "loss": 0.1551, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.1705426356589146, | |
| "grad_norm": 4.859110219903441, | |
| "learning_rate": 2.0907054222102367e-06, | |
| "loss": 0.0216, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 2.3255813953488373, | |
| "grad_norm": 2.695580515171099, | |
| "learning_rate": 1.4004826312100218e-06, | |
| "loss": 0.0013, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 2.4806201550387597, | |
| "grad_norm": 13.977813199581801, | |
| "learning_rate": 8.30010910550611e-07, | |
| "loss": 0.0114, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 2.6356589147286824, | |
| "grad_norm": 0.13568716951643228, | |
| "learning_rate": 3.9826908215420344e-07, | |
| "loss": 0.0088, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.7906976744186047, | |
| "grad_norm": 13.78710482551732, | |
| "learning_rate": 1.196206122203647e-07, | |
| "loss": 0.0018, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.945736434108527, | |
| "grad_norm": 6.559094463464456, | |
| "learning_rate": 3.3357581488030476e-09, | |
| "loss": 0.0141, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.9767441860465116, | |
| "step": 192, | |
| "total_flos": 0.0, | |
| "train_loss": 0.26802204532820423, | |
| "train_runtime": 6969.4008, | |
| "train_samples_per_second": 3.109, | |
| "train_steps_per_second": 0.028 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 192, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "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": 0.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |