Qwen2.5-1.5B-hl-baseline-v3 / train_results.json
Muennighoff's picture
Model save
f255b7f verified
raw
history blame contribute delete
201 Bytes
{
"total_flos": 0.0,
"train_loss": 21.70611201519991,
"train_runtime": 127568.9117,
"train_samples": 12000,
"train_samples_per_second": 11.238,
"train_steps_per_second": 0.013
}