| # optional for chat with PDF | |
| langchain==0.0.202 | |
| pypdf==3.9.1 | |
| # avoid textract, requires old six | |
| #textract==1.6.5 | |
| # for HF embeddings | |
| sentence_transformers==2.2.2 | |
| # local vector db | |
| chromadb==0.3.25 | |
| # server vector db | |
| #pymilvus==2.2.8 | |
| # weak url support, if can't install opencv etc. If comment-in this one, then comment-out unstructured[local-inference]==0.6.6 | |
| # unstructured==0.6.6 | |
| # strong support for images | |
| # Requires on Ubuntu: sudo apt-get install libmagic-dev poppler-utils tesseract-ocr libreoffice | |
| unstructured[local-inference]==0.7.4 | |
| #pdf2image==1.16.3 | |
| #pytesseract==0.3.10 | |
| pillow | |
| pdfminer.six==20221105 | |
| urllib3 | |
| requests_file | |
| #pdf2image==1.16.3 | |
| #pytesseract==0.3.10 | |
| tabulate==0.9.0 | |
| # FYI pandoc already part of requirements.txt | |
| # JSONLoader, but makes some trouble for some users | |
| # jq==1.4.1 | |
| # to check licenses | |
| # Run: pip-licenses|grep -v 'BSD\|Apache\|MIT' | |
| pip-licenses==4.3.0 | |
| # weaviate vector db | |
| weaviate-client==3.20.0 | |