distilled-mt5-small-b0.5 / train_results.json
Lvxue's picture
End of training
3b22fd7
{
"epoch": 5.0,
"train_loss": 3.97135716796875,
"train_runtime": 2537.2035,
"train_samples": 10000,
"train_samples_per_second": 19.707,
"train_steps_per_second": 4.927
}