byt5_en_zul_news / train_results.json
Davlan's picture
add MT model
1740c75
{
"epoch": 10.0,
"train_loss": 0.9517042933872768,
"train_runtime": 2753.4267,
"train_samples": 3500,
"train_samples_per_second": 12.711,
"train_steps_per_second": 1.271
}