Files
sft-qwen2.5-1.5b/eval-results/ifbench/metrics.json
ModelHub XC d765952c57 初始化项目,由ModelHub XC社区提供模型
Model: seopbo/sft-qwen2.5-1.5b
Source: Original Platform
2026-06-13 17:41:06 +08:00

16 lines
431 B
JSON

{
"ifbench": {
"pass@1": {
"num_prompts": 294,
"num_instructions": 335,
"average_score": 13.7447964260331,
"prompt_strict_accuracy": 10.54421768707483,
"instruction_strict_accuracy": 12.53731343283582,
"prompt_loose_accuracy": 14.285714285714285,
"instruction_loose_accuracy": 17.611940298507463,
"num_entries": 294,
"avg_tokens": 457,
"gen_seconds": 21
}
}
}