bartholemewii/stans-ml-stack repository overview

This Docker image contains Stan's ML Stack, a comprehensive machine learning stack for AMD GPUs.

⁠Features

• Complete ML Stack: Includes all components of Stan's ML Stack
• AMD GPU Support: Optimized for AMD GPUs with ROCm
• Easy Installation: Simple setup with clear instructions
• Comprehensive Documentation: Detailed documentation included

⁠Usage

To run this container with GPU access:

docker run --device=/dev/kfd --device=/dev/dri --group-add video -it bartholemewii/stans-ml-stack:latest

⁠Verification

To verify the installation, run the following command inside the container:

/workspace/verify_ml_stack.sh

⁠Components

The Docker image includes:

• ONNX Runtime: For machine learning model interoperability
• Transformers: For working with transformer-based models
• DeepSpeed: For optimizing large model training
• MPI: For distributed computing

⁠Post-Installation

After starting the container, you need to install PyTorch with ROCm support and MIGraphX. Use the scripts in /workspace/Stans_MLStack/scripts/ to install these components.

⁠Documentation

Documentation is available in the /workspace/Stans_MLStack/docs directory.

⁠Examples

Example code is available in the /workspace/Stans_MLStack/examples directory.

⁠Tags

• latest: The most recent version of Stan's ML Stack

⁠Architecture Support

This image is optimized for AMD GPUs and requires:

• AMD GPU with ROCm support
• Docker with GPU passthrough capabilities

⁠Author

Stanley Chisango (Scooter Lacroix)

⁠License

See the license information in the repository.

⁠Issues and Contributions

For issues or contributions, please visit the GitHub repository⁠.