lixiqi's picture
End of training
89de6ac
raw
history blame contribute delete
205 Bytes
{
"epoch": 9.97,
"total_flos": 1.427863337526786e+18,
"train_loss": 0.991901319367545,
"train_runtime": 910.8419,
"train_samples_per_second": 20.3,
"train_steps_per_second": 0.154
}