| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.204993429697766, | |
| "eval_steps": 500, | |
| "global_step": 25, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 0.14113833010196686, | |
| "learning_rate": 1.8090169943749477e-05, | |
| "loss": 1.0538, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "eval_loss": 1.0396900177001953, | |
| "eval_runtime": 438.5079, | |
| "eval_samples_per_second": 0.433, | |
| "eval_steps_per_second": 0.433, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 0.136638343334198, | |
| "learning_rate": 1.3090169943749475e-05, | |
| "loss": 1.0313, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "eval_loss": 1.0124541521072388, | |
| "eval_runtime": 437.8856, | |
| "eval_samples_per_second": 0.434, | |
| "eval_steps_per_second": 0.434, | |
| "step": 11 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 0.12238386273384094, | |
| "learning_rate": 6.909830056250527e-06, | |
| "loss": 1.0072, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 2.86, | |
| "eval_loss": 0.9969655871391296, | |
| "eval_runtime": 438.0615, | |
| "eval_samples_per_second": 0.434, | |
| "eval_steps_per_second": 0.434, | |
| "step": 17 | |
| }, | |
| { | |
| "epoch": 3.36, | |
| "grad_norm": 0.12007021903991699, | |
| "learning_rate": 1.9098300562505266e-06, | |
| "loss": 0.9954, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "eval_loss": 0.9925699234008789, | |
| "eval_runtime": 437.8318, | |
| "eval_samples_per_second": 0.434, | |
| "eval_steps_per_second": 0.434, | |
| "step": 23 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "grad_norm": 0.11785910278558731, | |
| "learning_rate": 0.0, | |
| "loss": 0.9927, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "eval_loss": 0.9924417734146118, | |
| "eval_runtime": 438.2372, | |
| "eval_samples_per_second": 0.434, | |
| "eval_steps_per_second": 0.434, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 4.2, | |
| "step": 25, | |
| "total_flos": 2.817465797050368e+17, | |
| "train_loss": 1.0161025619506836, | |
| "train_runtime": 23580.164, | |
| "train_samples_per_second": 0.161, | |
| "train_steps_per_second": 0.001 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 25, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 2.817465797050368e+17, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |