gpt_train_2_768_new / train_results.json
gokulsrinivasagan's picture
End of training
bc548b6 verified
raw
history blame contribute delete
252 Bytes
{
"epoch": 9.321127579192096,
"total_flos": 5.363892569191219e+17,
"train_loss": 4.763968430426513,
"train_runtime": 93599.1484,
"train_samples": 660643,
"train_samples_per_second": 705.822,
"train_steps_per_second": 7.353
}