djbp's picture
End of training
4ff2747 verified
raw
history blame
222 Bytes
{
"epoch": 9.836065573770492,
"total_flos": 1.9365344998357156e+18,
"train_loss": 0.1960654123624166,
"train_runtime": 8023.8475,
"train_samples_per_second": 9.716,
"train_steps_per_second": 0.037
}