| { | |
| "best_metric": 72.0587, | |
| "best_model_checkpoint": "bin/indosum-base-1/checkpoint-4460", | |
| "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": 0.9205507636070251, | |
| "learning_rate": 0.0008, | |
| "loss": 1.1904, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_gen_len": 95.85981308411215, | |
| "eval_loss": 0.8053300976753235, | |
| "eval_rouge1": 65.8257, | |
| "eval_rouge2": 57.6167, | |
| "eval_rougeL": 62.6222, | |
| "eval_rougeLsum": 65.0027, | |
| "eval_runtime": 190.0984, | |
| "eval_samples_per_second": 3.94, | |
| "eval_steps_per_second": 0.126, | |
| "step": 892 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "grad_norm": 0.9236537218093872, | |
| "learning_rate": 0.0006, | |
| "loss": 0.6851, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_gen_len": 96.29105473965286, | |
| "eval_loss": 0.6778599619865417, | |
| "eval_rouge1": 67.8889, | |
| "eval_rouge2": 60.0878, | |
| "eval_rougeL": 64.5868, | |
| "eval_rougeLsum": 66.9914, | |
| "eval_runtime": 191.4753, | |
| "eval_samples_per_second": 3.912, | |
| "eval_steps_per_second": 0.125, | |
| "step": 1784 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.7036951184272766, | |
| "learning_rate": 0.0004, | |
| "loss": 0.4856, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_gen_len": 96.92122830440587, | |
| "eval_loss": 0.6460219621658325, | |
| "eval_rouge1": 70.9241, | |
| "eval_rouge2": 63.6363, | |
| "eval_rougeL": 67.8555, | |
| "eval_rougeLsum": 70.153, | |
| "eval_runtime": 191.9289, | |
| "eval_samples_per_second": 3.902, | |
| "eval_steps_per_second": 0.125, | |
| "step": 2676 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "grad_norm": 0.5336605906486511, | |
| "learning_rate": 0.0002, | |
| "loss": 0.3358, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_gen_len": 101.87449933244326, | |
| "eval_loss": 0.6564512848854065, | |
| "eval_rouge1": 69.9002, | |
| "eval_rouge2": 62.4, | |
| "eval_rougeL": 66.5928, | |
| "eval_rougeLsum": 69.0347, | |
| "eval_runtime": 192.0018, | |
| "eval_samples_per_second": 3.901, | |
| "eval_steps_per_second": 0.125, | |
| "step": 3568 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.46569395065307617, | |
| "learning_rate": 0.0, | |
| "loss": 0.1973, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_gen_len": 99.37650200267022, | |
| "eval_loss": 0.7477750778198242, | |
| "eval_rouge1": 72.0587, | |
| "eval_rouge2": 64.7973, | |
| "eval_rougeL": 68.9279, | |
| "eval_rougeLsum": 71.3028, | |
| "eval_runtime": 192.0514, | |
| "eval_samples_per_second": 3.9, | |
| "eval_steps_per_second": 0.125, | |
| "step": 4460 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "step": 4460, | |
| "total_flos": 4.883347058982912e+16, | |
| "train_loss": 0.5788277185551254, | |
| "train_runtime": 5011.2822, | |
| "train_samples_per_second": 14.231, | |
| "train_steps_per_second": 0.89 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 4460, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 4.883347058982912e+16, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |