Serverless ML part 4

Gradio

Backtrack

There was a bit of a pause in the course so I ended up deleting resources for earlier lectures, but now the models trained earlier are needed again, so I have to rerun the whole pipeline from couple of previous lectures:

Deploying to Huggingface Spaces

Bit of a convoluted process, with having to connect to Hopsworks from within the Huggingface deployment:

At least the Iris example is working:

Serverless ml pt.5 : notes