llm_test_bpe / train_results.json
RefalMachine's picture
load model
8b5e245
raw
history blame contribute delete
199 Bytes
{
"epoch": 1.0,
"train_loss": 3.081914688561298,
"train_runtime": 169290.0352,
"train_samples": 28691269,
"train_samples_per_second": 169.48,
"train_steps_per_second": 0.706
}