ALM-AHME's picture
End of training
e58b3a5
{
"epoch": 4.96,
"total_flos": 8.138660625246413e+18,
"train_loss": 0.22687274961061374,
"train_runtime": 10415.6562,
"train_samples_per_second": 2.873,
"train_steps_per_second": 0.045
}