| { | |
| "best_metric": 0.8405963302752294, | |
| "best_model_checkpoint": "microsoft/Multilingual-MiniLM-L12-H384-finetuned-sst2/run-0/checkpoint-1684", | |
| "epoch": 5.0, | |
| "eval_steps": 500, | |
| "global_step": 2105, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8119266055045872, | |
| "eval_loss": 0.4295983910560608, | |
| "eval_runtime": 1.0076, | |
| "eval_samples_per_second": 865.424, | |
| "eval_steps_per_second": 54.585, | |
| "step": 421 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 8.123950004577637, | |
| "learning_rate": 1.3357804670815412e-05, | |
| "loss": 0.4991, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.823394495412844, | |
| "eval_loss": 0.39807233214378357, | |
| "eval_runtime": 1.0042, | |
| "eval_samples_per_second": 868.388, | |
| "eval_steps_per_second": 54.772, | |
| "step": 842 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 5.427116870880127, | |
| "learning_rate": 9.196494804517776e-06, | |
| "loss": 0.3654, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 0.8337155963302753, | |
| "eval_loss": 0.40886953473091125, | |
| "eval_runtime": 0.9135, | |
| "eval_samples_per_second": 954.617, | |
| "eval_steps_per_second": 60.211, | |
| "step": 1263 | |
| }, | |
| { | |
| "epoch": 3.56, | |
| "grad_norm": 1.0015597343444824, | |
| "learning_rate": 5.03518493822014e-06, | |
| "loss": 0.2923, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 0.8405963302752294, | |
| "eval_loss": 0.43070515990257263, | |
| "eval_runtime": 0.931, | |
| "eval_samples_per_second": 936.614, | |
| "eval_steps_per_second": 59.075, | |
| "step": 1684 | |
| }, | |
| { | |
| "epoch": 4.75, | |
| "grad_norm": 18.156158447265625, | |
| "learning_rate": 8.738750719225037e-07, | |
| "loss": 0.2538, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_accuracy": 0.838302752293578, | |
| "eval_loss": 0.4588472545146942, | |
| "eval_runtime": 1.0148, | |
| "eval_samples_per_second": 859.266, | |
| "eval_steps_per_second": 54.197, | |
| "step": 2105 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 2105, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 170928927395856.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 1.751911453711305e-05, | |
| "num_train_epochs": 5, | |
| "per_device_train_batch_size": 16, | |
| "seed": 24 | |
| } | |
| } | |