openwebtext-49K-base / train_results.json
gartland's picture
Model save
dbb426f verified
raw
history blame contribute delete
232 Bytes
{
"epoch": 1.0,
"total_flos": 810453211545600.0,
"train_loss": 3.407503606151048,
"train_runtime": 8384.1282,
"train_samples": 3433221,
"train_samples_per_second": 409.491,
"train_steps_per_second": 1.6
}