codemate-ai 0.1.4

CodeMate

CodeMate: AI-Powered Jupyter Notebook Assistant

CodeMate is a Jupyter notebook extension that leverages powerful AI models for code assistance, debugging, and generation. It supports multiple Large Language Model (LLM) providers, including OpenAI, Anthropic, Google Gemini, and local models via Hugging Face, bringing a versatile range of tools right to your coding environment.

Features

• 🤖 Multiple LLM Provider Support: Seamlessly integrate with OpenAI, Anthropic Claude, Google Gemini, and Hugging Face.
• 🔍 Contextual Code Analysis: Analyze your code's context and structure for better understanding.
• 🐛 AI-Powered Debugging: Get suggestions to fix errors directly within the notebook.
• ✨ Code Generation: Generate new code based on your current project context.
• 📱 Local Model Support: Run models locally using Hugging Face’s Transformers library.
• 📱 Customization: Customize the look and feel of the interface and the LLM interacting with.
• 🚀 Easy Integration: Effortlessly integrate with your Jupyter notebooks for enhanced productivity.

Installation

Install CodeMate via pip:

pip install codemate_ai

Install CodeAssist via git clone:

git clone https://github.com/somethingreallycool123/CodeMate.git

Quick Start

Import the extension in your Jupyter notebook:

%load_ext codemate_ai

Set up your preferred LLM provider:

For API-based LLMs (OpenAI, Anthropic, Google Gemini, etc):

%set_llm_provider <provider_name>
%set_api_key <provider_name> <YOUR_API_KEY>

Local Model Support

CodeAssist also supports running models locally using Hugging Face’s Transformers library:

Download and run a model locally:

%set_llm_provider transformers_download <model_name> --8bit  # The --8bit option is optional for reduced memory usage

Use an already downloaded model:

%set_llm_provider transformers_local ./models/my_model

Magic Commands

%set_api_key: Set API keys for different providers.
%set_llm_provider: Configure the LLM provider (e.g., OpenAI, Anthropic, or local models).
%generate_code: Generate new code based on the current context.
%%debug_cell: Debug a cell with AI assistance.
%set_code_theme: Change the background theme of the code solutions.
%set_persona: Change the LLM behaviour to various types including but not limited to detailed, consise, beginnerfriendly etc.
%%refactor_code: Refactors code for modern best practices.
%%explain_code: Cell magic to generate detailed explanation of code.
%%optimize_code: Cell magic to suggest performance optimizations for code.
%%generate_test: Cell magic to generate unit tests for code.

To understand each individual magic command functionality and operations, see Magic function Documentation.

Example Usage

Set up OpenAI as the provider:

%set_api_key openai YOUR_API_KEY
%set_llm_provider openai

Set up a locally running model as provider:

%set_llm_provider transformers_download facebook/opt-350m --8bit

OR

%set_llm_provider transformers_local ./models/my_model

Generate a new function:

%generate_code create_data_pipeline

Debug a problematic cell:

%%debug_cell
def process_data(df):
    result = df.groupby('category').mean()
    return result['value'] / 0  # Intentional error

Change the background theme:

%set_code_theme rrt

Change the persona of the LLM:

%set_persona expert

Project Structure

CodeMate/
├── codemate_ai/
│   ├── __init__.py
│   ├── core.py
│   ├── magics.py
│   ├── providers.py
├── tests/
│   ├── test_ML.py
│   ├── test_setup.py
├── LICENSE
├── README.md
├── requirements.txt
├── setup.py
└── .gitignore

Contributing

We welcome contributions! If you'd like to improve or add features to CodeAssist, please submit a pull request.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contact

For inquiries, please contact us via:

Email: [[e-mail]]

GitHub: somethingreallycool123