List your company
ResearchFormerly WizWand

Conversational Ability on Vicuna Eval (test)

8.91Vicuna Eval GPT-4 ScoreFAA
7.28767.70888.138.5512Dec 26, 2025
Updated 8mo ago

Evaluation results

MethodLinks
FAABackbone Model=Qwen2 7B, Trainable Parameters=7.30M2025.128.91
LoRABackbone Model=Qwen2 7B, Trainable Parameters=5.40M2025.128.89
FFBackbone Model=Qwen2 7B, Trainable Parameters=7.07B2025.128.88
FourierFTBackbone Model=Qwen2 7B, Trainable Parameters=0.85M2025.128.85
LoRABackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.21M2025.128.85
FFBackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.5B2025.128.83
AdapterHBackbone Model=Qwen2 7B, Trainable Parameters=7.29M2025.128.82
FourierFTBackbone Model=Deepseek R1-1.5B, Trainable Parameters=0.15M2025.128.82
FAABackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.63M2025.128.82
AdapterHBackbone Model=Deepseek R1-1.5B, Trainable Parameters=1.63M2025.128.79
LoRABackbone Model=LLaMA3 8B, Trainable Parameters=6.47M2025.128.25
FourierFTBackbone Model=LLaMA3 8B, Trainable Parameters=0.91M2025.128.23
AdapterHBackbone Model=LLaMA3 8B, Trainable Parameters=8.73M2025.128.21
FAABackbone Model=LLaMA3 8B, Trainable Parameters=8.73M2025.128.19
FFBackbone Model=LLaMA3 8B, Trainable Parameters=8.03B2025.128.14
LoRABackbone Model=LLaMA2 7B, Trainable Parameters=5.37M2025.127.45
FourierFTBackbone Model=LLaMA2 7B, Trainable Parameters=0.82M2025.127.42
FAABackbone Model=LLaMA2 7B, Trainable Parameters=7.27M2025.127.4
FFBackbone Model=LLaMA2 7B, Trainable Parameters=6.94B2025.127.39
AdapterHBackbone Model=LLaMA2 7B, Trainable Parameters=7.27M2025.127.35