| { | |
| "best_metric": 12.0444, | |
| "best_model_checkpoint": "./tst-translation-output5/checkpoint-6000", | |
| "epoch": 2.571795970852979, | |
| "eval_steps": 2000, | |
| "global_step": 6000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.21, | |
| "learning_rate": 1.6566666666666665e-05, | |
| "loss": 2.2544, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "learning_rate": 3.323333333333333e-05, | |
| "loss": 1.5518, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 4.99e-05, | |
| "loss": 1.4716, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "learning_rate": 4.9729361794815946e-05, | |
| "loss": 1.3822, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "eval_bleu": 8.4562, | |
| "eval_gen_len": 15.8491, | |
| "eval_loss": 1.369325041770935, | |
| "eval_runtime": 299.4199, | |
| "eval_samples_per_second": 15.58, | |
| "eval_steps_per_second": 0.975, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "learning_rate": 4.9457089958614686e-05, | |
| "loss": 1.2265, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "learning_rate": 4.918481812241342e-05, | |
| "loss": 1.0251, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "learning_rate": 4.891254628621216e-05, | |
| "loss": 1.0351, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "learning_rate": 4.864027445001089e-05, | |
| "loss": 1.0081, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "eval_bleu": 11.3956, | |
| "eval_gen_len": 15.239, | |
| "eval_loss": 1.2806029319763184, | |
| "eval_runtime": 274.3132, | |
| "eval_samples_per_second": 17.006, | |
| "eval_steps_per_second": 1.064, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "learning_rate": 4.836854715748203e-05, | |
| "loss": 1.0022, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.14, | |
| "learning_rate": 4.809627532128077e-05, | |
| "loss": 0.774, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.36, | |
| "learning_rate": 4.78240034850795e-05, | |
| "loss": 0.673, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "learning_rate": 4.755227619255064e-05, | |
| "loss": 0.7007, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 2.57, | |
| "eval_bleu": 12.0444, | |
| "eval_gen_len": 15.34, | |
| "eval_loss": 1.3149274587631226, | |
| "eval_runtime": 275.621, | |
| "eval_samples_per_second": 16.925, | |
| "eval_steps_per_second": 1.059, | |
| "step": 6000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 93320, | |
| "num_train_epochs": 40, | |
| "save_steps": 2000, | |
| "total_flos": 2.08044457721856e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |