Document Visual Information Extraction (Document VQA format) on FUNSD (test)
81.62ANLSLayTokenLLM-8B
Evaluation results
| Method | Links | |
|---|---|---|
| LayTokenLLM-8BBackbone=Llama3-8B2025.03 | 81.62 | |
| LayTokenLLM-7BBackbone=Qwen1.5-7B2025.03 | 78.72 | |
| InternVL2-8BBackbone=8B2025.03 | 75.84 | |
| TextMonkey-7BBackbone=7B2025.03 | 65.49 | |
| QwenVL-7BBackbone=7B2025.03 | 47.09 |