Long Question Answering on MM-Telco Long QA
41ROUGE-1GPT-4o
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| GPT-4o2025.11 | 41 | 15 | 23 | 91 | 75.39 | 10 | |
| Llama3.2 3BParameters=3B2025.11 | 39 | 13 | 22 | 76 | 45.41 | 8.25 | |
| Phi 4 14BParameters=14B2025.11 | 39 | 13 | 20 | 89 | 61.14 | 8.53 | |
| Llama3.1 8BParameters=8B2025.11 | 39 | 12 | 20 | 89 | 60.2 | 8.29 | |
| Nemotron 70BParameters=70B2025.11 | 38 | 12 | 20 | 90 | 72.04 | 6.58 | |
| Qwen2.5VL 7BParameters=7B2025.11 | 28 | 11 | 18 | 88 | 43.23 | 3.4 |