t1msan's picture
End of training
13f7d81 verified
raw
history blame
208 Bytes
{
"epoch": 8.42,
"total_flos": 7.88469671840514e+17,
"train_loss": 0.18382581882178783,
"train_runtime": 367.9939,
"train_samples_per_second": 32.446,
"train_steps_per_second": 0.109
}