I’m unable to restart a private Space which used to run before for the past several days. During starting, it errors out (after a long time) with a message “Scheduling failure: Unable to schedule” without any additional logs. I tried Factory Reboot, but didn’t help. How can I debug this ?
Hi @anilgs The version of Gradio has been updated. Can you please update it in the README.md file of your Space? Let us know if you keep running into issues after that.
1 Like
It helped once for my streamlit space, but then for my Gradio space changing the version in the readme never helped. It appears that something major has gone wrong with HF… Three of my friends have similar issue… wait for 15 min for their spaces to startup but in vain.
3 Likes
I’m getting this same error on two dedicated inference endpoints today when they fail to start after initializing for many minutes. Might start a new post since it’s a different service, but replying here for cross-visibility. I can’t find any documentation on these error codes.
3 Likes
Getting the same thing on a Streamlit HF space today. So not related to Gradio.
3 Likes
Also, I’m getting billed for all of this, for no fault of mine. Would you be able to refund the amount spent on debugging the issue… It has been over 24 hours the error has been persisting.
1 Like
was having the same issue. I’m not certain what fixed it, but a few things that may have done it: changing the port in question from the Gradio port. Also scheduling on different hardware (and going back to the original)
1 Like
Reporting the same error for a HF Endpoint. It is continuously happening when I try to use an AWS-powered NVIDIA T4 instance, but gets fixed if I switch to any other kind of instance(from the 3-4 I have tried out). Seems like a provisioning issue on HF side as reported here.
1 Like