mistral_sky_o1_5_full / all_results.json
CharlesLi's picture
End of training
23ed283 verified
raw
history blame contribute delete
392 Bytes
{
"epoch": 1.0,
"eval_loss": 0.5357503294944763,
"eval_runtime": 2.091,
"eval_samples": 20,
"eval_samples_per_second": 20.087,
"eval_steps_per_second": 1.435,
"total_flos": 125366068838400.0,
"train_loss": 0.5761321304437513,
"train_runtime": 6968.6131,
"train_samples": 15980,
"train_samples_per_second": 5.5,
"train_steps_per_second": 0.172
}