Inference Endpoint not starting on HTTP request

Hello! I have an Inference Endpoint that autoscales to 0 after inactivity. Recently I noticed that the model doesn’t start to initialize after hitting the inactive endpoint with an HTTP request nor does it return a 503 as stated by the docs.

Is there an updated configuration to allow this behavior?

Thank you!

2 Likes

Hi @nick-livefront Thanks for reporting this! We’ll take a look into this and I’ll update you soon.

2 Likes

Having the same issue.