Qwen2.5-1.5B-hl-baseline-v7 / train_results.json
Muennighoff's picture
Model save
7917f98 verified
{
"total_flos": 0.0,
"train_loss": -0.016174088740447498,
"train_runtime": 46045.8921,
"train_samples": 12000,
"train_samples_per_second": 31.134,
"train_steps_per_second": 0.035
}