| { | |
| "best_metric": 3.8532800674438477, | |
| "best_model_checkpoint": "shawgpt-ft-trial-0/checkpoint-51", | |
| "epoch": 16.32, | |
| "eval_steps": 500, | |
| "global_step": 51, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 1.182029128074646, | |
| "learning_rate": 1.4241804300364774e-05, | |
| "loss": 4.4787, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "eval_loss": 4.28471040725708, | |
| "eval_runtime": 4.3912, | |
| "eval_samples_per_second": 2.05, | |
| "eval_steps_per_second": 1.139, | |
| "step": 3 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 1.1892027854919434, | |
| "learning_rate": 1.3351691531591974e-05, | |
| "loss": 4.4208, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "eval_loss": 4.241559982299805, | |
| "eval_runtime": 4.5104, | |
| "eval_samples_per_second": 1.995, | |
| "eval_steps_per_second": 1.109, | |
| "step": 6 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "grad_norm": 1.1556155681610107, | |
| "learning_rate": 1.2461578762819176e-05, | |
| "loss": 4.3607, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 2.88, | |
| "eval_loss": 4.19962215423584, | |
| "eval_runtime": 4.4573, | |
| "eval_samples_per_second": 2.019, | |
| "eval_steps_per_second": 1.122, | |
| "step": 9 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "grad_norm": 1.2566847801208496, | |
| "learning_rate": 1.1571465994046378e-05, | |
| "loss": 4.3267, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 3.84, | |
| "eval_loss": 4.157830238342285, | |
| "eval_runtime": 4.4814, | |
| "eval_samples_per_second": 2.008, | |
| "eval_steps_per_second": 1.116, | |
| "step": 12 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "grad_norm": 1.3846135139465332, | |
| "learning_rate": 1.0681353225273581e-05, | |
| "loss": 4.2753, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 4.8, | |
| "eval_loss": 4.116264343261719, | |
| "eval_runtime": 4.4672, | |
| "eval_samples_per_second": 2.015, | |
| "eval_steps_per_second": 1.119, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 1.3648666143417358, | |
| "learning_rate": 9.791240456500782e-06, | |
| "loss": 4.2226, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "eval_loss": 4.075719833374023, | |
| "eval_runtime": 4.4752, | |
| "eval_samples_per_second": 2.011, | |
| "eval_steps_per_second": 1.117, | |
| "step": 18 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "grad_norm": 1.3280531167984009, | |
| "learning_rate": 8.901127687727984e-06, | |
| "loss": 4.1794, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 6.72, | |
| "eval_loss": 4.038341999053955, | |
| "eval_runtime": 4.4754, | |
| "eval_samples_per_second": 2.011, | |
| "eval_steps_per_second": 1.117, | |
| "step": 21 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "grad_norm": 1.3101249933242798, | |
| "learning_rate": 7.714310662697586e-06, | |
| "loss": 3.0992, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_loss": 3.9939630031585693, | |
| "eval_runtime": 4.4736, | |
| "eval_samples_per_second": 2.012, | |
| "eval_steps_per_second": 1.118, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "grad_norm": 1.2454874515533447, | |
| "learning_rate": 6.824197893924788e-06, | |
| "loss": 4.0964, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "eval_loss": 3.9650166034698486, | |
| "eval_runtime": 4.4761, | |
| "eval_samples_per_second": 2.011, | |
| "eval_steps_per_second": 1.117, | |
| "step": 28 | |
| }, | |
| { | |
| "epoch": 9.92, | |
| "grad_norm": 1.375960350036621, | |
| "learning_rate": 5.934085125151989e-06, | |
| "loss": 4.0783, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 9.92, | |
| "eval_loss": 3.939103364944458, | |
| "eval_runtime": 4.477, | |
| "eval_samples_per_second": 2.01, | |
| "eval_steps_per_second": 1.117, | |
| "step": 31 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "grad_norm": 1.290183424949646, | |
| "learning_rate": 5.04397235637919e-06, | |
| "loss": 4.0368, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 10.88, | |
| "eval_loss": 3.916520118713379, | |
| "eval_runtime": 4.472, | |
| "eval_samples_per_second": 2.013, | |
| "eval_steps_per_second": 1.118, | |
| "step": 34 | |
| }, | |
| { | |
| "epoch": 11.84, | |
| "grad_norm": 1.2861074209213257, | |
| "learning_rate": 4.153859587606393e-06, | |
| "loss": 4.0198, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 11.84, | |
| "eval_loss": 3.8970413208007812, | |
| "eval_runtime": 4.4703, | |
| "eval_samples_per_second": 2.013, | |
| "eval_steps_per_second": 1.119, | |
| "step": 37 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "grad_norm": 1.274073600769043, | |
| "learning_rate": 3.263746818833594e-06, | |
| "loss": 3.9997, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 12.8, | |
| "eval_loss": 3.8809890747070312, | |
| "eval_runtime": 4.4616, | |
| "eval_samples_per_second": 2.017, | |
| "eval_steps_per_second": 1.121, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 13.76, | |
| "grad_norm": 1.2961071729660034, | |
| "learning_rate": 2.3736340500607955e-06, | |
| "loss": 3.9693, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 13.76, | |
| "eval_loss": 3.868441581726074, | |
| "eval_runtime": 4.4563, | |
| "eval_samples_per_second": 2.02, | |
| "eval_steps_per_second": 1.122, | |
| "step": 43 | |
| }, | |
| { | |
| "epoch": 14.72, | |
| "grad_norm": 1.2889803647994995, | |
| "learning_rate": 1.4835212812879972e-06, | |
| "loss": 3.972, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 14.72, | |
| "eval_loss": 3.859734535217285, | |
| "eval_runtime": 4.4718, | |
| "eval_samples_per_second": 2.013, | |
| "eval_steps_per_second": 1.118, | |
| "step": 46 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "grad_norm": 1.3282195329666138, | |
| "learning_rate": 2.9670425625759943e-07, | |
| "loss": 2.9677, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 16.0, | |
| "eval_loss": 3.8539750576019287, | |
| "eval_runtime": 4.4757, | |
| "eval_samples_per_second": 2.011, | |
| "eval_steps_per_second": 1.117, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 16.32, | |
| "grad_norm": 1.3120520114898682, | |
| "learning_rate": 0.0, | |
| "loss": 3.8464, | |
| "step": 51 | |
| }, | |
| { | |
| "epoch": 16.32, | |
| "eval_loss": 3.8532800674438477, | |
| "eval_runtime": 4.4788, | |
| "eval_samples_per_second": 2.009, | |
| "eval_steps_per_second": 1.116, | |
| "step": 51 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 51, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 17, | |
| "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": 159625721610240.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |