gpt_16_4_3e-5_lp5_nb5 / all_results.json
54data's picture
End of training
f01d634
raw
history blame contribute delete
194 Bytes
{
"epoch": 4.0,
"train_loss": 2.831424217569864,
"train_runtime": 8708.1966,
"train_samples": 42367,
"train_samples_per_second": 19.461,
"train_steps_per_second": 1.216
}