Bisher's picture
End of training
e784ba7 verified
raw
history blame contribute delete
227 Bytes
{
"epoch": 0.48959608323133413,
"total_flos": 1.6503006591246067e+18,
"train_loss": 0.018675562590360642,
"train_runtime": 8627.3162,
"train_samples_per_second": 43.148,
"train_steps_per_second": 0.071
}