Qwen2.5-1.5B-hl-true-v3 / train_results.json
Muennighoff's picture
Model save
30ed8df verified
raw
history blame contribute delete
203 Bytes
{
"total_flos": 0.0,
"train_loss": -0.007817977162558236,
"train_runtime": 31764.4075,
"train_samples": 12000,
"train_samples_per_second": 45.132,
"train_steps_per_second": 0.05
}