distilbert_rand_100_v2 / train_results.json
Hartunka's picture
End of training
1fe8a2b verified
{
"epoch": 25.0,
"total_flos": 7.69437063436032e+17,
"train_loss": 8.358330273888473,
"train_runtime": 17076.0076,
"train_samples": 228639,
"train_samples_per_second": 334.737,
"train_steps_per_second": 3.487
}