Update requirements.txt

requirements.txt

@@ -1,14 +1,48 @@
-gradio
-langchain>=0.2.0
-langchain-community
-huggingface_hub
-unstructured
-pdfminer.six
-pi-heif
-unstructured-inference
-PyMuPDF
-pdf2image
-pytesseract
-transformers
-sentence-transformers
-faiss-cpu
+# Core dependencies for multimodal RAG system
+gradio==4.44.0
+huggingface-hub==0.25.0
+transformers==4.45.0
+
+# LangChain dependencies
+langchain==0.3.0
+langchain-community==0.3.0
+langchain-core==0.3.0
+langchain-text-splitters==0.3.0
+langchain-huggingface==0.1.0
+
+# Vector store and multimodal embeddings
+faiss-cpu==1.8.0
+sentence-transformers==3.0.1
+clip-by-openai==1.0
+
+# PDF processing and multimodal support
+unstructured[pdf]==0.15.12
+pillow==10.4.0
+pdf2image==1.17.0
+pymupdf==1.24.0
+pdfplumber==0.11.0
+
+# Image processing and computer vision
+opencv-python==4.10.0.84
+numpy==1.26.4
+ftfy==6.1.1
+regex==2023.12.25
+
+# Document processing
+python-docx==1.1.2
+python-pptx==0.6.23
+
+# Additional utilities
+requests==2.32.3
+torch==2.4.0
+torchvision==0.19.0
+
+# For better PDF extraction
+#detectron2 @ git+https://github.com/facebookresearch/detectron2.git
+#layoutparser[paddlepaddle]==0.3.4
+
+# System dependencies (these might be needed for unstructured)
+poppler-utils
+#tesseract-ocr
+#libgl1-mesa-glx
+#libglib2.0-0