SmolLM2-360M-sft / train_results.json
wassname's picture
End of training
f0a6d5f verified
raw
history blame
237 Bytes
{
"epoch": 3.0,
"total_flos": 7.533943292711404e+17,
"train_loss": 1.4212341141200129,
"train_runtime": 13229.3679,
"train_samples": 117772,
"train_samples_per_second": 14.73,
"train_steps_per_second": 0.058
}