Qwen2.5-1.5B-hl-baseline-v10 / all_results.json
Muennighoff's picture
Model save
9e6f069 verified
{
"total_flos": 0.0,
"train_loss": -0.0037147246307250095,
"train_runtime": 83704.5789,
"train_samples": 12000,
"train_samples_per_second": 17.127,
"train_steps_per_second": 0.019
}