Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Automatically provision models when accessing URL first time #27

Open
jcklie opened this issue Feb 25, 2021 · 5 comments
Open

Automatically provision models when accessing URL first time #27

jcklie opened this issue Feb 25, 2021 · 5 comments

Comments

@jcklie
Copy link
Contributor

jcklie commented Feb 25, 2021

It would be cool to add a function to ariadne so that you can just do something like this:

http://localhost:5000/ariadne/flair/sequencetagger/<modelname>

and then it just fetches the right model and that's it

@david-waterworth
Copy link

That would work quite nicely with the huggingface transformers library.

modelname = 'bert-base-cased'
model = AutoModel.from_pretrained(modelname)

@jcklie
Copy link
Contributor Author

jcklie commented Apr 21, 2021

We already use adapter for your transformer needs. I am currently working on v2 of this project, so I do not want to implement this and then archive the project.

@david-waterworth
Copy link

No problem, what's adapter? Is that something I can use now to implement external recommenders or is it a work in progress?

@jcklie
Copy link
Contributor Author

jcklie commented Apr 21, 2021

Adapters are a way to efficiently finetune transformers. You do not fine tune the whole thing, but insert a small layer and finetune that.

https://adapterhub.ml/
https://github.com/inception-project/inception-external-recommender/blob/master/ariadne/contrib/adapters.py

@david-waterworth
Copy link

Thanks!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants