distilled-mt5-small-0.6-0.25 / train_results.json
Lvxue's picture
End of training
03668a3
{
"epoch": 5.0,
"train_loss": 15.29005791015625,
"train_runtime": 2659.6394,
"train_samples": 10000,
"train_samples_per_second": 18.8,
"train_steps_per_second": 4.7
}