Elia: An Open Source Terminal UI for Interacting with LLMs
Elia: A Terminal User Interface for Language Models
Elia is a snappy, keyboard-centric terminal user interface designed to facilitate interactions with various large language models, including ChatGPT, Claude, Llama 3, Phi 3, and Mistral. Developed by Darren Burns, Elia offers an efficient and streamlined way to chat with these models directly from your terminal.
Features of Elia
1. Keyboard-Centric Interface: Designed for speed and efficiency, Elia allows users to navigate and interact with language models using keyboard shortcuts.
2. Local Conversation Storage: Conversations are stored locally in an SQLite database, ensuring privacy and easy access.
3. Multi-Model Support: Elia supports multiple language models, providing flexibility and choice for users.
4. Customization: Users can configure Elia through a configuration file to suit their preferences.
5. Local Model Support: Run local models using tools like ollama.
Installation and Configuration
Prerequisites
Ensure you have pipx installed. If not, you can install it using:
python3 -m pip install --user pipxpython3 -m pipx ensurepath
Installation
1. Install Elia using pipx:
pipx install elia
2. Verify the installation by running:
elia --help
Configuration
1. Create a configuration file in your home directory:
mkdir -p ~/.config/eliatouch ~/.config/elia/config.toml
2. Edit `config.toml` to specify your preferred language model and settings. For example:
toml[default]model = "chatgpt"api_key = "your_api_key_here"
3. Run Elia:
elia
Running Local Models
To run local models, you can use the `ollama` tool:
1. Install ollama:
pipx install ollama
2. Configure Elia to use a local model:
toml[default]model = "local"command = "ollama run model_name"
Additional Features
- Import Conversations: Import existing conversations into Elia.
- Database Reset: Easily reset the local SQLite database when needed.
Elia is a powerful tool for anyone needing to interact with large language models in a streamlined, terminal-based environment. For more details and updates, visit the Elia GitHub repository: https://github.com/darrenburns/elia .