List your company
ResearchFormerly WizWand

Multi-turn Dialogue Evaluation on MT-Bench-zh

6.34ScoreTaP-SFT (GPT-4)
2.50243.49874.4955.4913Jun 30, 2025
Updated 8mo ago

Evaluation results

MethodLinks
TaP-SFT (GPT-4)Base Model=Qwen2.5-14B, Evaluator=GPT-4o2025.066.34
TaP-SFT (DeepSeek-V2)Base Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.066.19
TaP-SFT (GPT-4)Base Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.066.18
TaP-SFT (GPT-4)Base Model=Qwen2.5-7B, Evaluator=GPT-4o2025.066.16
TaP-SFT (DeepSeek-V2)Base Model=Qwen2.5-14B, Evaluator=GPT-4o2025.066.12
TaP-SFT (DeepSeek-V2)Base Model=Qwen2.5-7B, Evaluator=GPT-4o2025.066.1
TaP-SFT (GPT-4)Base Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.065.99
TaP-SFT (DeepSeek-V2)Base Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.065.97
TaP-SFT (GPT-4)Base Model=Qwen2.5-3B, Evaluator=GPT-4o2025.065.79
TaP-SFT (DeepSeek-V2)Base Model=Qwen2.5-3B, Evaluator=GPT-4o2025.065.67
TaP-SFT (DeepSeek-V2)Base Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.065.6
TaP-SFT (GPT-4)Base Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.065.53
Alpaca-GPT-4-ZHBase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.065.44
TaP-SFT (GPT-4)Base Model=Gemma-2-9B, Evaluator=GPT-4o2025.065.37
Alpaca-GPT-4-ZHBase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.065.36
Infinity-InstructBase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.065.33
MOSS-SFTBase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.065.26
TaP-SFT (GPT-4)Base Model=Llama-3.1-8B, Evaluator=GPT-4o2025.065.24
TaP-SFT (DeepSeek-V2)Base Model=Llama-3.1-8B, Evaluator=GPT-4o2025.065.11
MOSS-SFTBase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.065.06
Infinity-InstructBase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.065.06
MOSS-SFTBase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.065.04
TaP-SFT (GPT-4)Base Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.065.03
Alpaca-GPT-4-ZHBase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.065.02
BELLE-SFTBase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.064.99
Infinity-InstructBase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.064.96
Infinity-InstructBase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.064.95
TaP-SFT (GPT-4)Base Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.064.95
BELLE-SFTBase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.064.93
COIG-CQIABase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.064.92
MOSS-SFTBase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.064.9
TaP-SFT (DeepSeek-V2)Base Model=Gemma-2-9B, Evaluator=GPT-4o2025.064.9
Infinity-InstructBase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.064.89
Alpaca-GPT-4-ZHBase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.064.89
TaP-SFT (DeepSeek-V2)Base Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.064.89
BELLE-SFTBase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.064.87
Alpaca-GPT-4-ZHBase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.064.83
BELLE-SFTBase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.064.83
BELLE-SFTBase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.064.83
FireflyBase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.064.79
Infinity-InstructBase Model=Gemma-2-9B, Evaluator=GPT-4o2025.064.78
COIG-CQIABase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.064.78
FireflyBase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.064.77
BELLE-SFTBase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.064.74
COIG-CQIABase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.064.73
Infinity-InstructBase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.064.71
BELLE-SFTBase Model=Gemma-2-9B, Evaluator=GPT-4o2025.064.69
TaP-SFT (DeepSeek-V2)Base Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.064.69
Alpaca-GPT-4-ZHBase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.064.68
Infinity-InstructBase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.064.68
MOSS-SFTBase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.064.66
BELLE-SFTBase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.064.65
BELLE-SFTBase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.064.63
Infinity-InstructBase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.064.6
COIG-CQIABase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.064.51
BELLE-SFTBase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.064.49
COIG-CQIABase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.064.47
FireflyBase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.064.47
FireflyBase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.064.45
Infinity-InstructBase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.064.45
FireflyBase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.064.43
FireflyBase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.064.41
Alpaca-GPT-4-ZHBase Model=Gemma-2-9B, Evaluator=GPT-4o2025.064.39
MOSS-SFTBase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.064.38
COIG-CQIABase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.064.35
Alpaca-GPT-4-ZHBase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.064.28
MOSS-SFTBase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.064.23
Alpaca-GPT-4-ZHBase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.064.21
MOSS-SFTBase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.064.21
MOSS-SFTBase Model=Gemma-2-9B, Evaluator=GPT-4o2025.064.14
Alpaca-GPT-4-ZHBase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.064.09
MOSS-SFTBase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.064.06
FireflyBase Model=Gemma-2-9B, Evaluator=GPT-4o2025.063.92
FireflyBase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.063.77
FireflyBase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.063.72
FireflyBase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.063.69
COIGBase Model=Qwen2.5-7B, Evaluator=DeepSeek-V32025.063.66
COIGBase Model=Qwen2.5-14B, Evaluator=GPT-4o2025.063.63
COIGBase Model=Qwen2.5-14B, Evaluator=DeepSeek-V32025.063.56
COIGBase Model=Qwen2.5-7B, Evaluator=GPT-4o2025.063.48
COIGBase Model=Qwen2.5-3B, Evaluator=DeepSeek-V32025.063.44
COIG-CQIABase Model=Gemma-2-9B, Evaluator=GPT-4o2025.063.36
COIGBase Model=Qwen2.5-3B, Evaluator=GPT-4o2025.063.35
COIG-CQIABase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.063.34
COIG-CQIABase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.063.32
COIG-CQIABase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.063.13
COIGBase Model=Llama-3.1-8B, Evaluator=DeepSeek-V32025.062.74
COIGBase Model=Gemma-2-9B, Evaluator=GPT-4o2025.062.69
COIGBase Model=Llama-3.1-8B, Evaluator=GPT-4o2025.062.68
COIGBase Model=Gemma-2-9B, Evaluator=DeepSeek-V32025.062.65