m2m100_418M_wol_fr_rel_ft / train_results.json
Davlan's picture
add MT model
73afafa
raw
history blame contribute delete
193 Bytes
{
"epoch": 3.0,
"train_loss": 2.2616551414368646,
"train_runtime": 352.1749,
"train_samples": 3360,
"train_samples_per_second": 28.622,
"train_steps_per_second": 2.862
}