Files
rlvrmulti-qwen2.5-1.5b/eval-results/hendrycks_math/metrics.json

11 lines
185 B
JSON
Raw Normal View History

{
"hendrycks_math": {
"pass@1": {
"num_entries": 5000,
"avg_tokens": 634,
"gen_seconds": 125,
"symbolic_correct": 54.56,
"no_answer": 5.9
}
}
}