Long PDF Understanding on PaperPDF Chinese 1.0
40.9ANLSPDF-WuKong
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| PDF-WuKong#Param=8.5B, Token=2107, Solution Category=Parsed Interleaved Text and Images2024.10 | 40.9 | 47.8 | 48.6 | 57.8 | |
| InternVL2#Param=8B, Token=6220, Solution Category=Parsed Interleaved Text and Images2024.10 | 28.5 | 40.6 | 42 | 54.7 | |
| InternVL2+BGE-M3*#Param=8.5B, Token=849, Solution Category=Plain Text Solution2024.10 | 28.4 | 34.2 | 34.5 | 46.3 | |
| InternVL2+BGE-M3#Param=8.5B, Token=583, Solution Category=Plain Text Solution2024.10 | 24.6 | 27.8 | 26.8 | 43 | |
| InternVL2#Param=8B, Token=4051, Solution Category=Plain Text Solution2024.10 | 24.4 | 28.2 | 27.7 | 30.8 | |
| Qwen2.5-VL#Param=7B, Token=5878, Solution Category=Pure Vision Solution2024.10 | 23.5 | 34.4 | 38.7 | 40.8 | |
| IXC2-VL+BGE-M3#Param=8.5B, Token=623, Solution Category=Plain Text Solution2024.10 | 23.3 | 29.8 | 32.3 | 38.1 | |
| IXC2-VL#Param=8B, Token=6217, Solution Category=Parsed Interleaved Text and Images2024.10 | 22.5 | 29.5 | 29.2 | 31.4 | |
| IXC2-VL#Param=8B, Token=4644, Solution Category=Plain Text Solution2024.10 | 21.1 | 28.5 | 28.7 | 30.8 | |
| MiniCPM-V 2.6 + VisRAG*(3p)#Param=10B, Token=2199, Solution Category=Pure Vision Solution2024.10 | 20.8 | 27.9 | 30.8 | 37.7 | |
| MiniCPM-V 2.6 + VisRAG*(2p)#Param=10B, Token=1444, Solution Category=Pure Vision Solution2024.10 | 20.5 | 27 | 30.1 | 35.9 | |
| InternVL2#Param=8B, Token=5008, Solution Category=Pure Vision Solution2024.10 | 20.3 | 33.9 | 37.8 | 37.7 | |
| IXC2-VL#Param=8B, Token=4712, Solution Category=Pure Vision Solution2024.10 | 15.9 | 19.5 | 22.2 | 27.4 |