light_r1 / all_results.json
sedrickkeh's picture
End of training
c4fa14b verified
raw
history blame contribute delete
222 Bytes
{
"epoch": 4.993958920660491,
"total_flos": 6.310856434702221e+18,
"train_loss": 0.2839509548799646,
"train_runtime": 73731.2238,
"train_samples_per_second": 5.387,
"train_steps_per_second": 0.042
}