distilled-mt5-small-b0.04 / train_results.json
Lvxue's picture
End of training
2f63853
{
"epoch": 5.0,
"train_loss": 3.76287099609375,
"train_runtime": 2865.4475,
"train_samples": 10000,
"train_samples_per_second": 17.449,
"train_steps_per_second": 4.362
}