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

{
  "ifeval": {
    "pass@1": {
      "num_prompts": 541,
      "num_instructions": 834,
      "average_score": 53.42729513247074,
      "prompt_strict_accuracy": 47.874306839186694,
      "instruction_strict_accuracy": 57.31414868105516,
      "prompt_loose_accuracy": 49.16820702402958,
      "instruction_loose_accuracy": 59.352517985611506,
      "num_entries": 541,
      "avg_tokens": 494,
      "gen_seconds": 26
    }
  }
}