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

16 lines
435 B
JSON
Raw Normal View History

{
"ifbench": {
"pass@1": {
"num_prompts": 294,
"num_instructions": 335,
"average_score": 24.357041323992284,
"prompt_strict_accuracy": 22.108843537414966,
"instruction_strict_accuracy": 23.582089552238806,
"prompt_loose_accuracy": 25.170068027210885,
"instruction_loose_accuracy": 26.56716417910448,
"num_entries": 294,
"avg_tokens": 1242,
"gen_seconds": 36
}
}
}