modernisa-v2-byt5-base-lr0.0001 / train_results.json
versae's picture
End of training
adaada8
{
"epoch": 5.0,
"train_loss": 0.11780740710349318,
"train_runtime": 33856.2876,
"train_samples": 183774,
"train_samples_per_second": 27.14,
"train_steps_per_second": 1.696
}