Noisy-RAG Question Answering on CoQA
92.4Exact Match (EM)Qwen2.5-OpAmp-72B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-OpAmp-72BParameters=72B, Adaptation=OpAmp2025.02 | 92.4 | |
| GPT-4o-0806Version=08062025.02 | 88.6 | |
| DeepSeek-V3Version=V32025.02 | 88.4 | |
| Llama3.3-70B-instParameters=70B, Type=Instruction-tuned2025.02 | 88.2 | |
| Qwen2.5-72B-instParameters=72B, Type=Instruction-tuned2025.02 | 85.8 | |
| Llama3.1-OpAmp-8BParameters=8B2025.02 | 85.4 | |
| Qwen2.5-7B-instParameters=7B2025.02 | 84.2 | |
| Llama3.1-8B-instParameters=8B2025.02 | 82.2 | |
| Llama3-ChatQA2-70BParameters=70B, Version=ChatQA22025.02 | 80.2 | |
| Llama3-ChatQA2-8BParameters=8B2025.02 | 78.2 | |
| Mistral-7B-inst-v0.3Parameters=7B2025.02 | 70.6 |