| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 3.0, | |
| "eval_steps": 500, | |
| "global_step": 18, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.16666666666666666, | |
| "grad_norm": 236.14656987198234, | |
| "learning_rate": 5e-06, | |
| "loss": 2.5604, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 201.89433604340425, | |
| "learning_rate": 1e-05, | |
| "loss": 2.1833, | |
| "step": 2 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 43.15775775201472, | |
| "learning_rate": 9.903926402016153e-06, | |
| "loss": 0.9756, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 13.315262915709834, | |
| "learning_rate": 9.619397662556434e-06, | |
| "loss": 0.7236, | |
| "step": 4 | |
| }, | |
| { | |
| "epoch": 0.8333333333333334, | |
| "grad_norm": 8.220944289845372, | |
| "learning_rate": 9.157348061512728e-06, | |
| "loss": 0.4679, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 16.58577267851266, | |
| "learning_rate": 8.535533905932739e-06, | |
| "loss": 0.402, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 0.32322201132774353, | |
| "eval_runtime": 1.0011, | |
| "eval_samples_per_second": 8.99, | |
| "eval_steps_per_second": 2.997, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.1666666666666667, | |
| "grad_norm": 2.76249506004443, | |
| "learning_rate": 7.777851165098012e-06, | |
| "loss": 0.2663, | |
| "step": 7 | |
| }, | |
| { | |
| "epoch": 1.3333333333333333, | |
| "grad_norm": 2.763402651988573, | |
| "learning_rate": 6.913417161825449e-06, | |
| "loss": 0.2429, | |
| "step": 8 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 3.883639510915858, | |
| "learning_rate": 5.975451610080643e-06, | |
| "loss": 0.2455, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 1.6666666666666665, | |
| "grad_norm": 3.161760936759259, | |
| "learning_rate": 5e-06, | |
| "loss": 0.2269, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.8333333333333335, | |
| "grad_norm": 2.335011386155214, | |
| "learning_rate": 4.02454838991936e-06, | |
| "loss": 0.2229, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 1.716063572976715, | |
| "learning_rate": 3.0865828381745515e-06, | |
| "loss": 0.1921, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_loss": 0.24893158674240112, | |
| "eval_runtime": 0.9909, | |
| "eval_samples_per_second": 9.083, | |
| "eval_steps_per_second": 3.028, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 2.1666666666666665, | |
| "grad_norm": 0.8996504564335516, | |
| "learning_rate": 2.2221488349019903e-06, | |
| "loss": 0.1879, | |
| "step": 13 | |
| }, | |
| { | |
| "epoch": 2.3333333333333335, | |
| "grad_norm": 1.3172641910170275, | |
| "learning_rate": 1.4644660940672628e-06, | |
| "loss": 0.1706, | |
| "step": 14 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 1.0444149673130907, | |
| "learning_rate": 8.426519384872733e-07, | |
| "loss": 0.1959, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.6666666666666665, | |
| "grad_norm": 0.6948355165030632, | |
| "learning_rate": 3.8060233744356634e-07, | |
| "loss": 0.2043, | |
| "step": 16 | |
| }, | |
| { | |
| "epoch": 2.8333333333333335, | |
| "grad_norm": 0.6328551118621334, | |
| "learning_rate": 9.607359798384785e-08, | |
| "loss": 0.1693, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.7570021616382325, | |
| "learning_rate": 0.0, | |
| "loss": 0.1928, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_loss": 0.23970693349838257, | |
| "eval_runtime": 0.9997, | |
| "eval_samples_per_second": 9.002, | |
| "eval_steps_per_second": 3.001, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 18, | |
| "total_flos": 3783747502080.0, | |
| "train_loss": 0.5461173463198874, | |
| "train_runtime": 901.4481, | |
| "train_samples_per_second": 0.562, | |
| "train_steps_per_second": 0.02 | |
| } | |
| ], | |
| "logging_steps": 1, | |
| "max_steps": 18, | |
| "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": 3783747502080.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |