| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 2.999725902238465, | |
| "eval_steps": 2000, | |
| "global_step": 16416, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.09, | |
| "learning_rate": 4.8477095516569206e-05, | |
| "loss": 1.6137, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "learning_rate": 4.69541910331384e-05, | |
| "loss": 1.5403, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "learning_rate": 4.54312865497076e-05, | |
| "loss": 1.5165, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "learning_rate": 4.39083820662768e-05, | |
| "loss": 1.5119, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "eval_loss": 1.3551265001296997, | |
| "eval_runtime": 1.6428, | |
| "eval_samples_per_second": 60.873, | |
| "eval_steps_per_second": 7.914, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "learning_rate": 4.238547758284601e-05, | |
| "loss": 1.5005, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "learning_rate": 4.086257309941521e-05, | |
| "loss": 1.498, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "learning_rate": 3.933966861598441e-05, | |
| "loss": 1.4876, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "learning_rate": 3.7816764132553604e-05, | |
| "loss": 1.4753, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "eval_loss": 1.338403344154358, | |
| "eval_runtime": 1.6495, | |
| "eval_samples_per_second": 60.624, | |
| "eval_steps_per_second": 7.881, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "learning_rate": 3.629385964912281e-05, | |
| "loss": 1.4879, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "learning_rate": 3.477095516569201e-05, | |
| "loss": 1.4864, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "learning_rate": 3.3248050682261215e-05, | |
| "loss": 1.481, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "learning_rate": 3.172514619883041e-05, | |
| "loss": 1.4226, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "eval_loss": 1.3284498453140259, | |
| "eval_runtime": 1.654, | |
| "eval_samples_per_second": 60.461, | |
| "eval_steps_per_second": 7.86, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "learning_rate": 3.0202241715399608e-05, | |
| "loss": 1.4117, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "learning_rate": 2.8679337231968812e-05, | |
| "loss": 1.4192, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "learning_rate": 2.7156432748538012e-05, | |
| "loss": 1.4175, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "learning_rate": 2.5633528265107216e-05, | |
| "loss": 1.4219, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "eval_loss": 1.3220161199569702, | |
| "eval_runtime": 1.6475, | |
| "eval_samples_per_second": 60.697, | |
| "eval_steps_per_second": 7.891, | |
| "step": 8000 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "learning_rate": 2.4110623781676416e-05, | |
| "loss": 1.4303, | |
| "step": 8500 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "learning_rate": 2.2587719298245616e-05, | |
| "loss": 1.4119, | |
| "step": 9000 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "learning_rate": 2.1064814814814816e-05, | |
| "loss": 1.4227, | |
| "step": 9500 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "learning_rate": 1.9541910331384016e-05, | |
| "loss": 1.4149, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "eval_loss": 1.3163000345230103, | |
| "eval_runtime": 1.6516, | |
| "eval_samples_per_second": 60.547, | |
| "eval_steps_per_second": 7.871, | |
| "step": 10000 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "learning_rate": 1.8019005847953216e-05, | |
| "loss": 1.4099, | |
| "step": 10500 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "learning_rate": 1.6496101364522417e-05, | |
| "loss": 1.404, | |
| "step": 11000 | |
| }, | |
| { | |
| "epoch": 2.1, | |
| "learning_rate": 1.497319688109162e-05, | |
| "loss": 1.3837, | |
| "step": 11500 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "learning_rate": 1.3450292397660819e-05, | |
| "loss": 1.3833, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "eval_loss": 1.316001534461975, | |
| "eval_runtime": 1.6487, | |
| "eval_samples_per_second": 60.654, | |
| "eval_steps_per_second": 7.885, | |
| "step": 12000 | |
| }, | |
| { | |
| "epoch": 2.28, | |
| "learning_rate": 1.192738791423002e-05, | |
| "loss": 1.3632, | |
| "step": 12500 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "learning_rate": 1.040448343079922e-05, | |
| "loss": 1.3841, | |
| "step": 13000 | |
| }, | |
| { | |
| "epoch": 2.47, | |
| "learning_rate": 8.881578947368421e-06, | |
| "loss": 1.3682, | |
| "step": 13500 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "learning_rate": 7.358674463937622e-06, | |
| "loss": 1.3714, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "eval_loss": 1.312637448310852, | |
| "eval_runtime": 1.6609, | |
| "eval_samples_per_second": 60.209, | |
| "eval_steps_per_second": 7.827, | |
| "step": 14000 | |
| }, | |
| { | |
| "epoch": 2.65, | |
| "learning_rate": 5.835769980506823e-06, | |
| "loss": 1.3754, | |
| "step": 14500 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "learning_rate": 4.312865497076023e-06, | |
| "loss": 1.3899, | |
| "step": 15000 | |
| }, | |
| { | |
| "epoch": 2.83, | |
| "learning_rate": 2.7899610136452243e-06, | |
| "loss": 1.3758, | |
| "step": 15500 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "learning_rate": 1.2670565302144249e-06, | |
| "loss": 1.3784, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "eval_loss": 1.3115426301956177, | |
| "eval_runtime": 1.6466, | |
| "eval_samples_per_second": 60.73, | |
| "eval_steps_per_second": 7.895, | |
| "step": 16000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "step": 16416, | |
| "total_flos": 2.077563543720837e+17, | |
| "train_loss": 1.4346119358293261, | |
| "train_runtime": 12475.7117, | |
| "train_samples_per_second": 21.055, | |
| "train_steps_per_second": 1.316 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 16416, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 2.077563543720837e+17, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |