This repository houses an extended version of the ViT Base/16 model from 🤗 Transformers. In particular, it provides the following:
- A
SavedModel
that has the preprocessing and postprocessing operations embedded inside the computation graph of the model. - A
tar
archive of the SavedModel.
Please refer to the following blog post to know how the SavedModel was exported: Deploying TensorFlow Vision Models in Hugging Face with TF Serving.