mt5_base-qg-ap-oficial / all_results.json
tiagoblima's picture
End of training
e07105b verified
raw
history blame
No virus
363 Bytes
{
"epoch": 5.0,
"eval_loss": 1.3846333026885986,
"eval_runtime": 55.4739,
"eval_samples": 10891,
"eval_samples_per_second": 196.326,
"eval_steps_per_second": 3.083,
"train_loss": 1.524873844560914,
"train_runtime": 7822.2573,
"train_samples": 75722,
"train_samples_per_second": 48.402,
"train_steps_per_second": 1.513
}