ALM-AHME's picture
End of training
bf5ddc8
raw
history blame
209 Bytes
{
"epoch": 4.96,
"total_flos": 8.138660625246413e+18,
"train_loss": 0.10703883119408161,
"train_runtime": 7147.9833,
"train_samples_per_second": 4.186,
"train_steps_per_second": 0.065
}