No recognized faces and error while searching
Hello everybody,
I've recently started using Immich on my home server and it is working really well as a backup tool. It is running quite smoothly on an Orange Pi 3 LTS.
I'm just having problems with the search function and the face recognition. Every time I try to search for something ("cake", "lake", "trees", etc) on the search bar I receive an "Internal server error 500" as you can see in the image. At the same time, I can't really find any kind of faces in the "Explore" tab, while the geo tagging works really well instead.
I understand that the hardware might be underpowered for the machine learning service but the machine-learning container seems to be running without issues. Furthermore from over 5000 photos in my library the "Recognize Face" jobs shows just a couple of failed jobs, just like the "Encode CLIP" and "Tag Objects" jobs. If those weren't working they should show 5000+ failed jobs, or something like that, right?
I can provide the docker-compose.yml and .env files if needed (too many characters for the post)
I've recently started using Immich on my home server and it is working really well as a backup tool. It is running quite smoothly on an Orange Pi 3 LTS.
I'm just having problems with the search function and the face recognition. Every time I try to search for something ("cake", "lake", "trees", etc) on the search bar I receive an "Internal server error 500" as you can see in the image. At the same time, I can't really find any kind of faces in the "Explore" tab, while the geo tagging works really well instead.
I understand that the hardware might be underpowered for the machine learning service but the machine-learning container seems to be running without issues. Furthermore from over 5000 photos in my library the "Recognize Face" jobs shows just a couple of failed jobs, just like the "Encode CLIP" and "Tag Objects" jobs. If those weren't working they should show 5000+ failed jobs, or something like that, right?
I can provide the docker-compose.yml and .env files if needed (too many characters for the post)
