Easily chat with your markdown files using a local AI model such as llama3, mistral
or others. System will generate embeddings and use surrealDB
to store
them.
FastAPI exposes and endpoint to query the model and chat with your own docs without needing to use paid models.
Project/Tool | Description |
---|---|
Ollama | A platform for creating, sharing, and running machine learning models. |
Langchain | A framework for developing applications powered by language models. |
SurrealDB | A scalable, distributed, document-graph database. |
FastAPI | A modern, fast web framework for building APIs with Python 3.6+. |
Just | A handy way to save and run project-specific commands. |
Once all the prerequisites are installed, use the just
commands to:
just setup
will create a python virtual environment, install required packages and check if all the prerequisites are correctly installedjust embed [./path/to/docs]
will create embeddings from the docsjust run [model]
will start SurrealDB, expose the API on the8282
port and run the server. Default model isllama3
but you can change it to any locally installedollama
modeljust chat [question]
will send the query to model and generate a response