distilled-mt5-small-0.4-0.5 / train_results.json
Lvxue's picture
End of training
7cf35f4
{
"epoch": 5.0,
"train_loss": 24.4316719140625,
"train_runtime": 2691.0647,
"train_samples": 10000,
"train_samples_per_second": 18.58,
"train_steps_per_second": 4.645
}