Victor Sonck

Hi! This issue should be fixed as of: https://github.com/allegroai/clearml-serving/pull/50/files 1. You can just redeploy the endpoint, pointing to a different preprocessing script 2. Yes, the inference container should be able...

Hi @gaspard-met I'll try to recreate your environment and help out. When you say you have a clearml server running on Kubernetes, how is that running locally? Microk8s? Also, do...

Thanks for the detailed writeup @Muscle-Oliver ! So I've taken a look and it seems like a specific parameter is missing from the helm chart. The url `http://127.0.0.1:8080/serve` does not...

I have the same in an SSH environment, so it seems likely to be as @asford suggests, connected to any remote setup

I'm not OP, but had the same issue and it is indeed fixed. Thank you very much. That fact that it's possible to revert to old behaviour might make it...

Hi @glenn-jocher, You're right, this is not the best experience, We're working right now to fix this for clearml in general, the change will then propagate to YOLOv5 as well....