deberta-em-large / train_results.json
Kerem P
End of training
cb4b708
raw
history blame
195 Bytes
{
"epoch": 2.98,
"train_loss": 0.17317123087043437,
"train_runtime": 1075.8425,
"train_samples": 5010,
"train_samples_per_second": 13.97,
"train_steps_per_second": 0.218
}