lpodina's picture
End of training
86f2b58
raw
history blame contribute delete
209 Bytes
{
"epoch": 3.0,
"total_flos": 1.8124066505760768e+18,
"train_loss": 0.3348683936554089,
"train_runtime": 1125.7736,
"train_samples_per_second": 64.755,
"train_steps_per_second": 0.506
}