m2m100_418M_swa_en_rel_ft / train_results.json
Davlan's picture
add MT model
eae203e
{
"epoch": 3.0,
"train_loss": 1.4262250119942639,
"train_runtime": 2645.2092,
"train_samples": 30782,
"train_samples_per_second": 34.911,
"train_steps_per_second": 3.492
}