Conversational Question Answering on TopiOCQA (test)
20.9EMQWEN - Base SFT
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| QWEN - Base SFTType=ANCE2026.01 | 20.9 | 46.7 | 54 | 55.1 | 60.2 | |
| GroGUType=ANCE, LLM=QWEN2026.01 | 20.8 | 46.9 | 54.6 | 54.7 | 60.7 | |
| GroGUType=BM25, LLM=QWEN2026.01 | 20.3 | 44.7 | 52.6 | 51.3 | 56.2 | |
| QWEN - OriginalType=ANCE2026.01 | 19.2 | 43.1 | 50.5 | 50.5 | 54.3 | |
| QWEN - Base SFTType=BM252026.01 | 19 | 42.5 | 50.3 | 48.7 | 52.4 | |
| QWEN - OriginalType=BM252026.01 | 17.7 | 39.8 | 47.5 | 46.3 | 48.8 | |
| GroGUType=ANCE, LLM=PHI2026.01 | 7 | 37.7 | 31.2 | 73.1 | 70.1 | |
| GroGUType=BM25, LLM=PHI2026.01 | 6.6 | 35.9 | 30 | 69.8 | 65.8 | |
| PHI - OriginalType=ANCE2026.01 | 6.2 | 36 | 29.8 | 70.2 | 66.7 | |
| PHI - Base SFTType=ANCE2026.01 | 6.2 | 36.1 | 29.9 | 70.5 | 66.2 | |
| PHI - OriginalType=BM252026.01 | 5.7 | 32.3 | 27 | 63.6 | 56.4 | |
| PHI - Base SFTType=BM252026.01 | 5.5 | 32.2 | 26.8 | 63.7 | 57.2 |