distilled-mt5-small-0.2-2 / train_results.json
Lvxue's picture
End of training
f218c45
{
"epoch": 5.0,
"train_loss": 72.67423921875,
"train_runtime": 2632.3566,
"train_samples": 10000,
"train_samples_per_second": 18.994,
"train_steps_per_second": 4.749
}