distilled-mt5-small-0.03-0.25 / train_results.json
Lvxue's picture
End of training
9b2f919
{
"epoch": 5.0,
"train_loss": 3.96877859375,
"train_runtime": 3283.259,
"train_samples": 10000,
"train_samples_per_second": 15.229,
"train_steps_per_second": 3.807
}