docling-project / docling-project/docling
As multiple PDFs are parsed, the memory of Docling continues to increase and will not decrease.
Open
question
- Dominant language
- Python
- Stars
- 66.4k
- Forks
- 4.8k
- Avg merge
- 3d 4h
- Merged PRs (30d)
- 98
Description
### Question
Docling runs on the CPU and RAM，without GPU, it can't operate properly. I think the documents used for parsing each PDF should be able to be reused, instead of the memory continuously increasing as more PDFs are parsed.
Is the continuous growth of memory a bug, or is it just the way the implementation mechanism works?
Contributor guide
Assessment
This issue has not been assessed yet.