Wiku is an artificial intelligence designed to be your unconditional support while you study. Inspired by the lo-fi and cozy aesthetic, Wiku is not just a technical tool; it's a learning environment designed to reduce stress and foster deep concentration.
Where once there were cold processes, now there is an experience guided by our mascot: a Yorkshire terrier ready to help you break down the toughest topics.
- 📚 Friendly Explanations: Wiku translates complex concepts into simple, easy-to-digest language.
- 🗣️ Natural Text-to-Speech: Listen to Wiku's responses with high-quality neural voices (supports English and Spanish).
- 💬 Multi-Language Support: Wiku automatically detects your language and responds accordingly (English/Spanish).
- 📄 PDF Analysis: Upload study documents, and Wiku will answer questions based strictly on the content provided.
- 🎧 Focus Mode: Interface designed to minimize distractions and maximize workflow.
- 📝 Smart Summaries: Ability to synthesize large volumes of information into clear outlines.
- 🌱 Priority Organization: Helps you decide where to start when you feel overwhelmed.
- Install the required dependencies:
pip install fastapi uvicorn llama-cpp-python PyPDF2 python-multipart edge-tts
-
Navigate to the
/appfolder. -
Start the FastAPI server:
python -m uvicorn main:app --reload
-
Install the packages:
npm install
-
Installing Rust (Required for Tauri)
To run the Tauri development server, you need to have the Rust toolchain installed.
-
Download Rust
Visit the official installer page:
https://rustup.rs
Run the installer and accept the default options.
-
Verify installation Restart your terminal and check that Rust is available:
cargo --version
If this command prints a version number (e.g. cargo 1.75.0), Rust is correctly installed.
-
Run the development server:
npx @tauri-apps/cli dev
Do you want to help make Wiku smarter or cozier? We'd love your help!
-
Fork the project.
-
Create a new Branch.
-
Submit a Pull Request explaining your changes.
- Model used: Qwen2.5-3B-Instruct-GGUF
- Qwen 2.5: Qwen2.5-72B-Instruct
- Llama cpp python: Github repository
Made with 🤎 for the studens community.