/langchain-rag

Ask questions to files

Primary LanguagePython

LangChain Document QA

Interface for interacting with files using langchain. Inspired by ollama examples. It follows the usual structure:

  1. Load documents
  2. Chunk them
  3. Create embeddings and store them into a vector datastore
  4. Retrieve from the datastore using a LLM

Setup

Run Llama 2 locally:

ollama run llama2

Install the required dependencies:

pip install -r requirements.txt

Run

Run the project:

python main.py

Example query:

Query: Can you summarise the content of the file?