Qwen2.5-1.5B-Open-R1-Distill / all_results.json
DeeLearning's picture
End of training
2ea3562 verified
raw
history blame
396 Bytes
{
"epoch": 1.0,
"eval_loss": 0.7578943967819214,
"eval_runtime": 1.9314,
"eval_samples": 100,
"eval_samples_per_second": 66.273,
"eval_steps_per_second": 2.071,
"total_flos": 76916824473600.0,
"train_loss": 0.7337305870281874,
"train_runtime": 1235.0855,
"train_samples": 16610,
"train_samples_per_second": 17.497,
"train_steps_per_second": 0.137
}