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

16 lines
432 B
JSON
Raw Normal View History

{
"ifbench": {
"pass@1": {
"num_prompts": 294,
"num_instructions": 335,
"average_score": 13.37166209767489,
"prompt_strict_accuracy": 10.54421768707483,
"instruction_strict_accuracy": 11.641791044776118,
"prompt_loose_accuracy": 14.285714285714285,
"instruction_loose_accuracy": 17.01492537313433,
"num_entries": 294,
"avg_tokens": 429,
"gen_seconds": 27
}
}
}