-
Clone the repository.
-
cdto the cloned directory. -
Create a virtual environment with
uv sync --frozen -
Activate the virtual environment (
source .venv/bin/activate) -
Pull the required models for local use:
ollama pull hf.co/abtsousa/llama3.2-drugsES-csv-ft:Q4_K_M # finetuned model
ollama pull llama3.2:3b-instruct-q4_K_M # base model
ollama pull jina/jina-embeddings-v2-base-es # embedding model-
Run the app with
streamlit run app.py -
The app is now working fully locally. Enjoy :)