I would like to ask about the first action if I want to add a new model at the source code level.
says ‘open “calls-for-model-addition” here and to contact us’, but the link is broken.
I know a way to make a Pull Request with the “PR for Model Addition” tag.
However, I would like to confirm with the community in advance whether the model I want to add is suitable.
What I want to do specifically:
I implemented and learned the aMLP model proposed in the paper “Pay Attention to MLPs”.
There are minor changes to adapt to NLP, not ViT.
Currently, it is available in my repository (GitHub - tanreinama/aMLP-japanese: aMLP Transformer Model for Japanese).
However, currently only Japanese is supported.
The method of text encoding is special. This is because it was created from scratch according to the characteristics of Japanese language. There is also handling to encode emojis into special tokens.
So there are two things I would like to know.
First, if I merge this model into Hagging Face, can I send pull request?
Second, if possible, I’d like some support to integrate it into Hagging Face. Is this possible?