distilled-mt5-small-0.005-0.5 / train_results.json
Lvxue's picture
End of training
3333937
{
"epoch": 5.0,
"train_loss": 3.500236611328125,
"train_runtime": 2808.5008,
"train_samples": 10000,
"train_samples_per_second": 17.803,
"train_steps_per_second": 4.451
}