Contributing Guidelines#

Thank you for your interest in contributing to MAMKit! We welcome contributions from everyone and look forward to your input to help improve and expand our toolkit. To ensure a smooth contribution process, please follow the guidelines below.

How to Contribute#

1. Implementing New Datasets or Models#

If you want to contribute by adding new datasets or models, please follow these steps:

  1. Understand the MAMKit Structure:

    • Review the current MAMKit structure and how datasets and models are implemented. This will help you align your contribution with the existing framework.

  2. Extend Existing Interfaces:

    • If you are adding a new dataset or model that fits within the current modalities and tasks, extend the existing interfaces accordingly. Make sure to follow the design patterns and coding standards used in MAMKit.

  3. Propose New Interfaces:

    • If your dataset introduces a new modality or your model requires a unique interface, propose new interfaces. Ensure that these interfaces are well-documented and integrate seamlessly with the existing system.

  4. Preprocessing Strategies:

    • Similarly, if you have new preprocessing strategies, follow the existing structure for extending or proposing new strategies. Document your approach and ensure it is compatible with the current setup.

2. Contributing Benchmarks#

If you want to contribute benchmarks, follow these steps:

  1. Prepare Your Benchmarks:

    • Implement benchmarking scripts and metrics for evaluating models using your dataset. Ensure that your benchmarks are thorough and adhere to the standards used in MAMKit.

  2. Document Your Benchmarks:

    • Provide clear documentation for your benchmarks, including details on evaluation metrics, experimental setup, and expected outcomes.

3. Submitting Your Code#

Once you have implemented your dataset, model, preprocessing strategy, or benchmarks, and you are happy with your code, follow these steps to submit it:

  1. Create a Pull Request:

    • Fork the MAMKit repository and create a new branch for your changes. Push your code to this branch and open a pull request (PR) against the main repository.

  2. Describe Your Changes:

    • In your pull request, provide a clear and detailed description of what your contribution does. Explain any new interfaces or changes to existing ones, and include relevant documentation.

  3. Collaborate on Improvements:

    • After you submit your pull request, our team will review your code. We may suggest improvements or ask questions. Please engage in this process constructively, and collaborate with us to refine your contribution.

4. Reviewing and Merging#

Our team will review your pull request to ensure it aligns with the MAMKit standards and integrates well with the existing codebase. We appreciate your patience during this process. If there are any revisions needed, we will work with you to address them.

Thank you for contributing to MAMKit! Your efforts help make this toolkit more robust and valuable for everyone. If you have any questions or need assistance, feel free to reach out to us.

Happy coding! 🚀