Files
rlvrmathif-qwen2.5-1.5b/eval-results/ifeval/metrics.json

16 lines
429 B
JSON
Raw Normal View History

{
"ifeval": {
"pass@1": {
"num_prompts": 541,
"num_instructions": 834,
"average_score": 69.29170157404575,
"prompt_strict_accuracy": 63.21626617375231,
"instruction_strict_accuracy": 72.78177458033574,
"prompt_loose_accuracy": 65.98890942698706,
"instruction_loose_accuracy": 75.17985611510791,
"num_entries": 541,
"avg_tokens": 494,
"gen_seconds": 60
}
}
}