NuminaMath-7B-CoT / all_results.json
lewtun's picture
lewtun HF staff
Add AI-MO/deepseek-math-7b-sft-aimo_v31.24 checkpoint
4c9f488 verified
raw
history blame contribute delete
No virus
440 Bytes
{
"epoch": 3.0,
"eval_loss": 0.45375168323516846,
"eval_runtime": 0.5431,
"eval_samples": 100,
"eval_samples_per_second": 46.032,
"eval_steps_per_second": 1.841,
"perplexity": 1.5742070469361709,
"total_flos": 2049636776804352.0,
"train_loss": 0.42822988295175696,
"train_runtime": 30718.8165,
"train_samples": 863474,
"train_samples_per_second": 21.729,
"train_steps_per_second": 0.679
}