MultimodalHugs

MultimodalHugs is an extension of Hugging Face that offers a generalized framework for training, evaluating, and using multimodal AI models with minimal code differences, ensuring seamless compatibility with Hugging Face pipelines.

Intallation

1. Clone the repository.

2. Navigate to the repository and Install the package.

   • Standard installation:

     cd /multimodalhugs
     pip install .

   • Developer installation:

     cd /multimodalhugs
     pip install -e .[dev]

Usage

Here you can find an example for Multimodal Translation.

Citing this work

@misc{multimodalhugs2024, 
    title={MultimodalHugs: Extending HuggingFace for Generalized Multimodal AI Model Training and Evaluation},
    author={Sant, Gerard},
    howpublished={\url{https://github.com/GerrySant/multimodalhugs}},
    year={2024}
}