docling-project / docling-project/docling
Puzzling Docling/Granite
- Dominant language
- Python
- Stars
- 66.4k
- Forks
- 4.8k
- Avg merge
- 2d 21h
- Merged PRs (30d)
- 84
Description
### Question
I ran the following command (one of your examples):
docling --to md --show-layout --pipeline vlm --vlm-model granite_docling "https://arxiv.org/pdf/2501.17887"
I did it just as a test of how well the software works.
When I looked at the output, I saw the following in the references section. It's #5 from the end and it has the following text:
"Wang, B.; Xu, C.; Zhao, X.; Ouyang, L.; Wu, F.; Zhao, Z.; Xu, R.; Liu, K.; Qu, Y.; Shang, F.; Zhang, B.; Wei, L.; Sui, Z.; Li, W.; Shi, B.; Qiao, Y.; Lin, D.; and He, C. 2024. MinerU: An Open-Source Solution for Precise Document Content Extraction. arXiv:2409.1839."
The issue is "arXiv:2409.1839". It's the wrong identifier. The correct one is "arXiv:2409.18839" as shown in the following screenshot:
At first I thought it was an error in the original paper, but it's not.
The reason it's puzzling is that I'm assuming you read the PDF and I would think that it should get the extra 8 just fine.
Am I missing something obvious? The difference is small, but it happens to reference a real document; it's about Quantum Physics. My concern is obviously that this could really cause problems if I were trying to build a reference tree. And it's not in a location where I would expect to see errors.
Thanks!
Contributor guide
Assessment
This issue has not been assessed yet.