| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.0, | |
| "eval_steps": 500, | |
| "global_step": 8456, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.23651844843897823, | |
| "grad_norm": 1.046875, | |
| "learning_rate": 1.881740775780511e-05, | |
| "loss": 0.3019, | |
| "mean_token_accuracy": 0.9215967762470245, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.47303689687795647, | |
| "grad_norm": 1.234375, | |
| "learning_rate": 1.763481551561022e-05, | |
| "loss": 0.246, | |
| "mean_token_accuracy": 0.9323436969518661, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.7095553453169348, | |
| "grad_norm": 1.890625, | |
| "learning_rate": 1.6452223273415328e-05, | |
| "loss": 0.2338, | |
| "mean_token_accuracy": 0.9348658822774887, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.9460737937559129, | |
| "grad_norm": 1.2109375, | |
| "learning_rate": 1.5269631031220437e-05, | |
| "loss": 0.2312, | |
| "mean_token_accuracy": 0.9352898219823838, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.1825922421948911, | |
| "grad_norm": 1.5703125, | |
| "learning_rate": 1.4087038789025544e-05, | |
| "loss": 0.2174, | |
| "mean_token_accuracy": 0.9378906596899033, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 1.4191106906338695, | |
| "grad_norm": 1.5546875, | |
| "learning_rate": 1.2904446546830653e-05, | |
| "loss": 0.2107, | |
| "mean_token_accuracy": 0.9396841472387314, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 1.6556291390728477, | |
| "grad_norm": 0.875, | |
| "learning_rate": 1.1721854304635763e-05, | |
| "loss": 0.2129, | |
| "mean_token_accuracy": 0.9386629806756973, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 1.8921475875118259, | |
| "grad_norm": 1.625, | |
| "learning_rate": 1.0539262062440872e-05, | |
| "loss": 0.2041, | |
| "mean_token_accuracy": 0.9411084994077682, | |
| "step": 4000 | |
| }, | |
| { | |
| "epoch": 2.128666035950804, | |
| "grad_norm": 1.4375, | |
| "learning_rate": 9.35666982024598e-06, | |
| "loss": 0.2061, | |
| "mean_token_accuracy": 0.9402695513963699, | |
| "step": 4500 | |
| }, | |
| { | |
| "epoch": 2.3651844843897822, | |
| "grad_norm": 1.1953125, | |
| "learning_rate": 8.174077578051088e-06, | |
| "loss": 0.201, | |
| "mean_token_accuracy": 0.9417461235523223, | |
| "step": 5000 | |
| }, | |
| { | |
| "epoch": 2.6017029328287604, | |
| "grad_norm": 0.86328125, | |
| "learning_rate": 6.991485335856198e-06, | |
| "loss": 0.1962, | |
| "mean_token_accuracy": 0.9427423716783524, | |
| "step": 5500 | |
| }, | |
| { | |
| "epoch": 2.838221381267739, | |
| "grad_norm": 1.65625, | |
| "learning_rate": 5.808893093661306e-06, | |
| "loss": 0.1989, | |
| "mean_token_accuracy": 0.9420522735118866, | |
| "step": 6000 | |
| }, | |
| { | |
| "epoch": 3.074739829706717, | |
| "grad_norm": 1.453125, | |
| "learning_rate": 4.626300851466415e-06, | |
| "loss": 0.1967, | |
| "mean_token_accuracy": 0.9427533217668533, | |
| "step": 6500 | |
| }, | |
| { | |
| "epoch": 3.3112582781456954, | |
| "grad_norm": 1.6796875, | |
| "learning_rate": 3.443708609271523e-06, | |
| "loss": 0.1928, | |
| "mean_token_accuracy": 0.943574698805809, | |
| "step": 7000 | |
| }, | |
| { | |
| "epoch": 3.5477767265846736, | |
| "grad_norm": 1.2734375, | |
| "learning_rate": 2.2611163670766324e-06, | |
| "loss": 0.199, | |
| "mean_token_accuracy": 0.9419986149072647, | |
| "step": 7500 | |
| }, | |
| { | |
| "epoch": 3.7842951750236518, | |
| "grad_norm": 1.9765625, | |
| "learning_rate": 1.078524124881741e-06, | |
| "loss": 0.1973, | |
| "mean_token_accuracy": 0.9425687384605408, | |
| "step": 8000 | |
| } | |
| ], | |
| "logging_steps": 500, | |
| "max_steps": 8456, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 5.722176240290611e+17, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |