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

11 lines
184 B
JSON
Raw Normal View History

{
"hendrycks_math": {
"pass@1": {
"num_entries": 5000,
"avg_tokens": 622,
"gen_seconds": 126,
"symbolic_correct": 52.5,
"no_answer": 5.1
}
}
}