ALM-AHME's picture
End of training
2dc96a6
raw
history blame contribute delete
211 Bytes
{
"epoch": 4.97,
"total_flos": 1.2267729254916882e+19,
"train_loss": 0.10130466574936041,
"train_runtime": 12554.5867,
"train_samples_per_second": 3.584,
"train_steps_per_second": 0.056
}