OllamaPDFInsight

OllamaPDFInsight is a small RAG (Retrieval-Augmented Generation) system that leverages Ollama to create embeddings from a PDF file, stores these embeddings in a Weaviate vector store, and uses Ollama to answer questions regarding the PDF content.

Steps

• Use LangChain document loader to turn the PDF into a set of documents.
• Create a collection from these documents in the Weaviate vector store.
• Use Ollama to generate embeddings from the documents, and store the embeddings in the collection.
• Query and retrieve information from the PDF using Ollama.

Run & Modify

1. Create a virtual environment:

pyenv install 3.8.10
pyenv virtualenv 3.8.10 ollama-pdf-insight-env
pyenv activate ollama-pdf-insight-env

2. Once the virtual environment is activated, install the requirements from the requirements.txt file.

pip install -r requirements.txt

3. Clone the repository:

git clone https://github.com/yourusername/OllamaPDFInsight.git
cd OllamaPDFInsight

4. Prepare the data and the weaviate vector store.

python load_data.py

5. Prepare the template, llm, and run the prompt.

python retrieve_context.py

References

I essentially followed the steps in this article and added my own touch to it.