ruDialoGPT-small / train_results.json
d.tsimerman
squash
d4a38f8
{
"epoch": 1.0,
"train_loss": 3.773873895684371,
"train_runtime": 142283.9744,
"train_samples": 8570756,
"train_samples_per_second": 60.237,
"train_steps_per_second": 5.02
}