macLlama: Native macOS GUI for Ollama

Welcome to macLlama! This macOS application, built with SwiftUI, provides a user-friendly interface for interacting with Ollama. Recent updates include the ability to start the Ollama server directly from the app and various UI enhancements.

System Requirements

• Operating System: macOS 14.0 Sonoma or later.
• Processor: Apple Silicon (e.g., M1, M2, M3, M4 series).
• Memory (RAM): Varies based on the Ollama models you plan to use. Larger models require more RAM. Refer to the official Ollama documentation for specific recommendations.
• Ollama Installation: A working installation of Ollama is required.

Getting Started

Prerequisites

Before using macLlama, you're going to need to install Ollama:

1. Install Homebrew (if needed):

   Open Terminal (/Applications/Utilities/Terminal) and run:

   /bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

2. Install Ollama:

   In Terminal, run:

   brew install ollama

3. Verify Installation:

   ollama --version

Installation & Running the App

1. Download the Application:

   • Go to the Releases Page of this repository.
   • Download the latest macLlama.app.zip file from the Assets section.

2. Install:

   • Unzip the downloaded file.
   • Drag the macLlama.app file to your /Applications folder.

3. Run the App:

   Open macLlama from your Applications folder.

Running the Ollama Server

macLlama can help you start the Ollama server. Alternatively, you can start it manually:

• From the App: Look for a button or menu option to start the server.

• Manually:

  ollama serve

Installing Ollama Models

Before chatting, you need to download models for Ollama.

• Via macLlama: Go to the Settings > Model Management tab, enter a model name(or location), and click "Pull".

• Via Terminal:

  ollama pull <model_name>

  Example:

  ollama pull llama3:8b-instruct

  Find available models on the Ollama Library.

Usage

1. Launch macLlama.
2. The app will attempt to load available models.
3. Select a model from the dropdown menu.
4. Type your message and press Enter or click the "Send" button.(or command + return to send)

Chat History (v1.0.5 - Build 2)

macLlama securely stores your chat conversations locally using SwiftData.

Viewing / Managing Chat History:

• Menu: Window > Chat History.

Resetting Chat History (Advanced - Use with Caution):

If you encounter issues with chat history, you can manually manage the SwiftData storage. This is an advanced operation and should only be attempted if you understand the potential consequences.

To reset your chat history:

1. Navigate to: ~/Library/Application Support/
2. Delete the following files:
   • default.store
   • default.store-shm
   • default.store-wal

WARNING: Deleting these files will remove all SwiftData storage for macLlama and may impact other applications using SwiftData. Back up your local storage before attempting this.

Contributing

Contributions are welcome! Fork the project, create a feature branch, commit your changes, push to the branch, and open a pull request.

License

Distributed under the Apache 2.0 License. See LICENSE.txt for more information.

Frameworks & Libraries

• SwiftUI: macLlama is built using SwiftUI for a modern user interface.
• MarkdownUI: We utilize the MarkdownUI library to render chat messages in Markdown.

Contact

• Project Discussions: https://github.com/hellotunamayo/macLlama/discussions
• Project Link: https://github.com/hellotunamayo/macLlama

Acknowledgements

• Ollama Team
• Contributors to libraries used

Enjoy using macLlama! 😊