| { |
| "best_metric": 72.8512, |
| "best_model_checkpoint": "bin/indosum-lora-3/checkpoint-3568", |
| "epoch": 5.0, |
| "eval_steps": 500, |
| "global_step": 4460, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 1.0, |
| "grad_norm": 0.6515501141548157, |
| "learning_rate": 0.0008, |
| "loss": 0.7809, |
| "step": 892 |
| }, |
| { |
| "epoch": 1.0, |
| "eval_gen_len": 100.42266666666667, |
| "eval_loss": 0.5510576963424683, |
| "eval_rouge1": 70.7387, |
| "eval_rouge2": 63.5003, |
| "eval_rougeL": 67.571, |
| "eval_rougeLsum": 69.7824, |
| "eval_runtime": 218.1822, |
| "eval_samples_per_second": 3.437, |
| "eval_steps_per_second": 0.11, |
| "step": 892 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 0.4482235908508301, |
| "learning_rate": 0.0006, |
| "loss": 0.5981, |
| "step": 1784 |
| }, |
| { |
| "epoch": 2.0, |
| "eval_gen_len": 104.784, |
| "eval_loss": 0.5352391600608826, |
| "eval_rouge1": 71.407, |
| "eval_rouge2": 64.2122, |
| "eval_rougeL": 68.3064, |
| "eval_rougeLsum": 70.4861, |
| "eval_runtime": 218.5017, |
| "eval_samples_per_second": 3.432, |
| "eval_steps_per_second": 0.11, |
| "step": 1784 |
| }, |
| { |
| "epoch": 3.0, |
| "grad_norm": 0.627132773399353, |
| "learning_rate": 0.0004, |
| "loss": 0.5542, |
| "step": 2676 |
| }, |
| { |
| "epoch": 3.0, |
| "eval_gen_len": 102.44266666666667, |
| "eval_loss": 0.5363225340843201, |
| "eval_rouge1": 72.3351, |
| "eval_rouge2": 65.2788, |
| "eval_rougeL": 69.2683, |
| "eval_rougeLsum": 71.4248, |
| "eval_runtime": 217.0288, |
| "eval_samples_per_second": 3.456, |
| "eval_steps_per_second": 0.111, |
| "step": 2676 |
| }, |
| { |
| "epoch": 4.0, |
| "grad_norm": 0.4163409471511841, |
| "learning_rate": 0.0002, |
| "loss": 0.5238, |
| "step": 3568 |
| }, |
| { |
| "epoch": 4.0, |
| "eval_gen_len": 102.41733333333333, |
| "eval_loss": 0.5148317217826843, |
| "eval_rouge1": 72.8512, |
| "eval_rouge2": 65.9167, |
| "eval_rougeL": 69.8532, |
| "eval_rougeLsum": 71.9591, |
| "eval_runtime": 217.8801, |
| "eval_samples_per_second": 3.442, |
| "eval_steps_per_second": 0.11, |
| "step": 3568 |
| }, |
| { |
| "epoch": 5.0, |
| "grad_norm": 0.39186742901802063, |
| "learning_rate": 0.0, |
| "loss": 0.5034, |
| "step": 4460 |
| }, |
| { |
| "epoch": 5.0, |
| "eval_gen_len": 102.25333333333333, |
| "eval_loss": 0.5104866027832031, |
| "eval_rouge1": 72.6979, |
| "eval_rouge2": 65.6731, |
| "eval_rougeL": 69.6184, |
| "eval_rougeLsum": 71.7604, |
| "eval_runtime": 218.1186, |
| "eval_samples_per_second": 3.438, |
| "eval_steps_per_second": 0.11, |
| "step": 4460 |
| }, |
| { |
| "epoch": 5.0, |
| "step": 4460, |
| "total_flos": 4.925218464202752e+16, |
| "train_loss": 0.5920752076290113, |
| "train_runtime": 4819.5172, |
| "train_samples_per_second": 14.806, |
| "train_steps_per_second": 0.925 |
| } |
| ], |
| "logging_steps": 500, |
| "max_steps": 4460, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 5, |
| "save_steps": 500, |
| "total_flos": 4.925218464202752e+16, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
