This project facilitates conversational interaction with any public GitHub repository utilizing advanced technologies such as IBM WatsonX, Langchain, FAISS vector database, and Streamlit. It operates as an RAG (Retrieval Augmented Generation) application, enhancing the user experience with efficient code exploration and understanding.

• IBM WatsonX Integration: Users input their IBM WatsonX API Key along with the GitHub Repository link.
• Repository Cloning and Chunking: The specified repository is cloned locally and segmented into manageable chunks for processing.
• Embedding Conversion: Utilizing Langchain, the code segments are transformed into embeddings.
• FAISS Vector Database: The embeddings are stored in a FAISS vector database for efficient retrieval and querying.
• QA Retriever: Langchain constructs a QA retriever, enabling users to engage in conversational queries related to their code.

1. Clone the repository:

git clone https://github.com/IBM/gitpal-watsonx-ai-github-chatbot.git
cd <your_project_directory>

2. Create a new virtual environment using a tool like virtualenv or conda, and activate the environment:

python3 -m venv <name_of_virtual_env>
source /venv/bin/activate

3. Install dependencies:

pip install -r requirements.txt

4. Create .env file in the root of your project with following keys:

PROJECT_ID=<your_watsonx.ai_project_id>
ENDPOINT_URL=<your_watsonx.ai_endpoint_url>
PROMPT_MODEL_ID=meta-llama/llama-3-70b-instruct
EMBEDDING_MODEL_ID=ibm/slate-125m-english-rtrvr

You can find the project_id as follows. Open the prompt lab in watsonx.ai. At the very top of the UI, there will be Projects / /. Click on the link. Then get the project_id from Project's Manage tab (Project -> Manage -> General -> Details).

5. Run the Streamlit app:

streamlit run chatbot.py

5. Access the chat interface:

Open your web browser and navigate to http://localhost:8501.

• Input your IBM WatsonX API key and the name of your GitHub repository.
• Engage with the chatbot by asking questions or providing instructions using natural language.

• Language Model Dependency: The chatbot's performance is contingent upon the accuracy and capabilities of the underlying language model.
• Processing Time: Large codebases or repositories with intricate structures may incur longer processing times for chunking and embedding.
• Response Quality: The accuracy and quality of responses are influenced by both the language model's accuracy and the effectiveness of code embeddings.

Launch App on Streamlit

Feel free to contribute, provide feedback, or report issues via GitHub. Your contributions are valued and help enhance the functionality and usability of this project.