Model deployment


Do the methods used for deployment to services like render, sagemaker in the course19 docs still apply today. Last tried them out a while back

For clarity am talking of the docs in the below link

I believe that course19 was based on fastai V1 libraries.
I believe the has examples of deploying on more current platforms using the fastai v2

1 Like

Hello Edward,

Thanks for the reply, I really appreciate it

So I have done the above class, I was actually just getting done with it.

The course video shows how to deploy via Hugging Face spaces with gradio.
Book shows you how to use ipython widgets in combination with voilĂ 

I was hoping to find some references for how to deploy using sagemaker, render etc for the current version of fastai something similar to what was there for course19

I did manage to find ways to deploy using bentoML with a fastai example and some easy to understand docs, will try implementing then sharing
I was hoping to basically test out the different deployment platforms to see what fits.

I also haven’t gone through the whole book or all the course videos hence there is a chance the above is referenced, but I just haven’t seen it yet.
Another option is maybe all the rest of the platforms I last tried out have become obsolete options since I last checked.