Spaces:
Running
on
Zero
Running
on
Zero
The model doesn't answer
#1
by
aguillenATC
- opened
Hi,
I was testing the API, the client is created without problems but the prediction call freezes...
Thanks.
Are you using this demo Space, or replicating it locally on your machine?
I'm using the demo space
The demo space is configured to use the free ZeroGPU infrastructure provided by Hugging Face, so performance can vary based on availability of GPUs.
When you tried it before it was (mis-)configured to run on CPU so performance was very bad. Most of the time it should be much better now, but there are no guarantees.
This comment has been hidden
Now it works prefectly, thanks!