Qwen2.5-1.5B-hl-baseline-v2 / train_results.json
Muennighoff's picture
Model save
4644dca verified
{
"total_flos": 0.0,
"train_loss": 0.002493098353266703,
"train_runtime": 118997.1326,
"train_samples": 12000,
"train_samples_per_second": 12.047,
"train_steps_per_second": 0.013
}