distilled-mt5-small-0.07-0.5 / train_results.json
Lvxue's picture
End of training
7d6af08
{
"epoch": 5.0,
"train_loss": 7.20621490234375,
"train_runtime": 2655.9783,
"train_samples": 10000,
"train_samples_per_second": 18.825,
"train_steps_per_second": 4.706
}