Open-ended Question Answering on CrossAlpaca-Eval en 2.0
8.58GPT-4o ScoreQwen2.5-7B-Instruction
Evaluation results
| Method | Links | |
|---|---|---|
| Qwen2.5-7B-InstructionBackbone=Qwen 2.5-7B, Reasoning Method=Instruction-tuned2025.01 | 8.58 | |
| Qwen2.5-7B-AdaMCOTBackbone=Qwen 2.5-7B, Reasoning Method=AdaMCOT2025.01 | 8.58 | |
| LLaMA3.1-8B-AdaMCOTBackbone=LLaMA 3.1-8B, Reasoning Method=AdaMCOT2025.01 | 8.35 | |
| LLaMA3.1-8B-InstructionBackbone=LLaMA 3.1-8B, Reasoning Method=Instruction-tuned2025.01 | 8.33 | |
| Qwen2.5-7B-AdaMCOTBackbone=Qwen 2.5-7B, Reasoning Method=AdaMCOT2025.01 | 8.16 | |
| LLaMA3.1-8B-AdaMCOTBackbone=LLaMA 3.1-8B, Reasoning Method=AdaMCOT2025.01 | 8.13 | |
| Qwen2.5-7B-InstructionBackbone=Qwen 2.5-7B, Reasoning Method=Instruction-tuned2025.01 | 7.85 | |
| LLaMA3.1-8B-InstructionBackbone=LLaMA 3.1-8B, Reasoning Method=Instruction-tuned2025.01 | 7.51 |