|
{ |
|
"best_metric": null, |
|
"best_model_checkpoint": null, |
|
"epoch": 3.0, |
|
"eval_steps": 500, |
|
"global_step": 14214, |
|
"is_hyper_param_search": false, |
|
"is_local_process_zero": true, |
|
"is_world_process_zero": true, |
|
"log_history": [ |
|
{ |
|
"epoch": 0.11, |
|
"learning_rate": 4.8241170676797525e-05, |
|
"loss": 5.1993, |
|
"step": 500 |
|
}, |
|
{ |
|
"epoch": 0.21, |
|
"learning_rate": 4.648234135359505e-05, |
|
"loss": 2.7321, |
|
"step": 1000 |
|
}, |
|
{ |
|
"epoch": 0.32, |
|
"learning_rate": 4.472351203039257e-05, |
|
"loss": 2.4547, |
|
"step": 1500 |
|
}, |
|
{ |
|
"epoch": 0.42, |
|
"learning_rate": 4.2964682707190094e-05, |
|
"loss": 2.3454, |
|
"step": 2000 |
|
}, |
|
{ |
|
"epoch": 0.53, |
|
"learning_rate": 4.120585338398762e-05, |
|
"loss": 2.2375, |
|
"step": 2500 |
|
}, |
|
{ |
|
"epoch": 0.63, |
|
"learning_rate": 3.944702406078514e-05, |
|
"loss": 2.1503, |
|
"step": 3000 |
|
}, |
|
{ |
|
"epoch": 0.74, |
|
"learning_rate": 3.768819473758266e-05, |
|
"loss": 2.0775, |
|
"step": 3500 |
|
}, |
|
{ |
|
"epoch": 0.84, |
|
"learning_rate": 3.592936541438019e-05, |
|
"loss": 2.0802, |
|
"step": 4000 |
|
}, |
|
{ |
|
"epoch": 0.95, |
|
"learning_rate": 3.4170536091177716e-05, |
|
"loss": 2.0028, |
|
"step": 4500 |
|
}, |
|
{ |
|
"epoch": 1.06, |
|
"learning_rate": 3.241170676797524e-05, |
|
"loss": 1.9862, |
|
"step": 5000 |
|
}, |
|
{ |
|
"epoch": 1.16, |
|
"learning_rate": 3.065287744477276e-05, |
|
"loss": 1.95, |
|
"step": 5500 |
|
}, |
|
{ |
|
"epoch": 1.27, |
|
"learning_rate": 2.8894048121570285e-05, |
|
"loss": 1.9279, |
|
"step": 6000 |
|
}, |
|
{ |
|
"epoch": 1.37, |
|
"learning_rate": 2.7135218798367808e-05, |
|
"loss": 1.8896, |
|
"step": 6500 |
|
}, |
|
{ |
|
"epoch": 1.48, |
|
"learning_rate": 2.537638947516533e-05, |
|
"loss": 1.8719, |
|
"step": 7000 |
|
}, |
|
{ |
|
"epoch": 1.58, |
|
"learning_rate": 2.3617560151962857e-05, |
|
"loss": 1.8719, |
|
"step": 7500 |
|
}, |
|
{ |
|
"epoch": 1.69, |
|
"learning_rate": 2.185873082876038e-05, |
|
"loss": 1.8422, |
|
"step": 8000 |
|
}, |
|
{ |
|
"epoch": 1.79, |
|
"learning_rate": 2.0099901505557903e-05, |
|
"loss": 1.8601, |
|
"step": 8500 |
|
}, |
|
{ |
|
"epoch": 1.9, |
|
"learning_rate": 1.8341072182355426e-05, |
|
"loss": 1.8463, |
|
"step": 9000 |
|
}, |
|
{ |
|
"epoch": 2.01, |
|
"learning_rate": 1.658224285915295e-05, |
|
"loss": 1.8008, |
|
"step": 9500 |
|
}, |
|
{ |
|
"epoch": 2.11, |
|
"learning_rate": 1.4823413535950473e-05, |
|
"loss": 1.812, |
|
"step": 10000 |
|
}, |
|
{ |
|
"epoch": 2.22, |
|
"learning_rate": 1.3064584212747996e-05, |
|
"loss": 1.7943, |
|
"step": 10500 |
|
}, |
|
{ |
|
"epoch": 2.32, |
|
"learning_rate": 1.130575488954552e-05, |
|
"loss": 1.7935, |
|
"step": 11000 |
|
}, |
|
{ |
|
"epoch": 2.43, |
|
"learning_rate": 9.546925566343042e-06, |
|
"loss": 1.782, |
|
"step": 11500 |
|
}, |
|
{ |
|
"epoch": 2.53, |
|
"learning_rate": 7.788096243140567e-06, |
|
"loss": 1.7745, |
|
"step": 12000 |
|
}, |
|
{ |
|
"epoch": 2.64, |
|
"learning_rate": 6.02926691993809e-06, |
|
"loss": 1.7724, |
|
"step": 12500 |
|
}, |
|
{ |
|
"epoch": 2.74, |
|
"learning_rate": 4.270437596735613e-06, |
|
"loss": 1.7846, |
|
"step": 13000 |
|
}, |
|
{ |
|
"epoch": 2.85, |
|
"learning_rate": 2.5116082735331366e-06, |
|
"loss": 1.7903, |
|
"step": 13500 |
|
}, |
|
{ |
|
"epoch": 2.95, |
|
"learning_rate": 7.5277895033066e-07, |
|
"loss": 1.763, |
|
"step": 14000 |
|
}, |
|
{ |
|
"epoch": 3.0, |
|
"step": 14214, |
|
"total_flos": 6.011566735491072e+16, |
|
"train_loss": 2.0746695444824423, |
|
"train_runtime": 9365.9127, |
|
"train_samples_per_second": 6.07, |
|
"train_steps_per_second": 1.518 |
|
} |
|
], |
|
"logging_steps": 500, |
|
"max_steps": 14214, |
|
"num_input_tokens_seen": 0, |
|
"num_train_epochs": 3, |
|
"save_steps": 500, |
|
"total_flos": 6.011566735491072e+16, |
|
"trial_name": null, |
|
"trial_params": null |
|
} |
|
|