Hope you are doing well.
I was following this tutorial: Google Colab
and I was wondering how we could load the saved model in the end. It seems like using the regular approach with Hugging Face doesn’t work because we are using a defined Class and have no config. I also found this on how to build custom models Sharing custom models, however I don’t understand the BLOCK_MAPPING part.
Can anyone help me understand the best way to load the model in my case?