m2m100_418M_swa_en_rel / train_results.json
Davlan's picture
add MT model
62b6276
{
"epoch": 3.0,
"train_loss": 0.9910313752351659,
"train_runtime": 64882.6806,
"train_samples": 869508,
"train_samples_per_second": 40.204,
"train_steps_per_second": 4.02
}