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

16 lines
430 B
JSON
Raw Normal View History

{
"ifeval": {
"pass@1": {
"num_prompts": 541,
"num_instructions": 834,
"average_score": 54.65005075422103,
"prompt_strict_accuracy": 47.874306839186694,
"instruction_strict_accuracy": 58.39328537170264,
"prompt_loose_accuracy": 50.46210720887245,
"instruction_loose_accuracy": 61.87050359712231,
"num_entries": 541,
"avg_tokens": 472,
"gen_seconds": 30
}
}
}