How should I make the fine-tuned model available as an API to call for others? Without downloading the whole model
#71
by
JoeyXia
- opened
Hi all, I'm here to find some suggestion as described in the title.
I fine-tuned an MPT model and created a private repo for it. I know others could clone the repo with the config and parameters files to use that fine-tuned model. Is there any way I can store the model somewhere like a cloud service in AWS and make it an API call that people don't need to download the whole model to use it?
Any discussion and suggestion are appreciated.
This seems like a generic question related to creating a model serving product. I'd suggest a different forum for this question as I think you are looking for generic help making a hosted api that others can hit.
daking
changed discussion status to
closed