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

16 lines
434 B
JSON
Raw Normal View History

{
"ifbench": {
"pass@1": {
"num_prompts": 294,
"num_instructions": 335,
"average_score": 15.096710325921414,
"prompt_strict_accuracy": 12.244897959183673,
"instruction_strict_accuracy": 14.029850746268657,
"prompt_loose_accuracy": 15.306122448979592,
"instruction_loose_accuracy": 18.80597014925373,
"num_entries": 294,
"avg_tokens": 429,
"gen_seconds": 22
}
}
}