deepseek-ai / deepseek-ai/DeepSeek-OCR
出现了很离谱的解析结果
- Dominant language
- Python
- Stars
- 23.9k
- Forks
- 2.2k
- PR merge metrics
- No merged PRs in 30d
Description
目前遇到了两个case,都出现了很奇怪的解析结果:
(1)针对下面这个文件
[yanbaopptmerge_yanbaoPPT_4570.pdf](https://github.com/user-attachments/files/23952242/yanbaopptmerge_yanbaoPPT_4570.pdf)
解析结果很离谱,输出了很多不存在的内容,我做了可视化方便查看:
[2.md](https://github.com/user-attachments/files/23952446/2.md)

(2)针对下面这个文件:
[jiaocaineedrop_jiaocai_needrop_en_2604.pdf](https://github.com/user-attachments/files/23952352/jiaocaineedrop_jiaocai_needrop_en_2604.pdf)
解析结果也是很奇怪,多了很多类似(<|ref|>text<|/ref|><|det|>[[440, 812, 562, 825]]<|/det|>
\(x = 8\) )不存在的内容,我做了可视化方便查看:
[res.md](https://github.com/user-attachments/files/23952404/res.md)

这是什么原因呢?
Contributor guide
No contributing guide indexed for this repository
Assessment
This issue has not been assessed yet.