This PR adds a model card and metadata for the RAVLT models, including the `image-classification` pipeline tag and an MIT license (assumed, please verify). The model card includes a description of the model, its key features, and performance metrics on ImageNet-1k. Placeholders for usage examples and the code repository link are included and should be updated once the code becomes publicly available. I've included the performance metrics and checkpoints for image classification, as that's the primary focus. The README includes performance for Object Detection, Instance Segmentation and Semantic Segmentation, but I've omitted those in favor of keeping the card concise and focused.
