Future downloadable small vision sound models

[hpssjellis]

Not sure if it fits the HuggingFace business model but there could be some use for a few small vision sound models to be downloaded to the desktop computer. The strong argument is data security but what I am interested in is speed.

[coyotte508]

you can do this with the python lib, find a small model among the 100k+ models on https://huggingface.co/ - see https://huggingface.co/tasks

It's python however, but see #82 too

[hpssjellis]

you can do this with the python lib, find a small model among the 100k+ models on https://huggingface.co/ - see https://huggingface.co/tasks

It's python however, but see #82 too

I am looking for Javascript clientside. I got squeeznet running about 4 years ago here, but no worries, HuggingFace is amazing.