distilled-mt5-small-0.4-2 / train_results.json
Lvxue's picture
End of training
ae40d23
{
"epoch": 5.0,
"train_loss": 139.736696875,
"train_runtime": 2715.2591,
"train_samples": 10000,
"train_samples_per_second": 18.414,
"train_steps_per_second": 4.604
}