ImmichI
Immich•5mo ago
Oscar

Machine learning request failed with status 500: Internal Server Error

Hi all,

I've just been trying to set up Immich on my Synology DS1522+ using docker compose (just the default downloads from the site) and had set it up seemingly successfully though I kept getting an error in console [Nest] 6 - 07/30/2025, 4:15:13 AM WARN [Microservices:MachineLearningRepository] Machine learning request to "http://immich-machine-learning:3003" failed with status 500: Internal Server Error. I am new to all of this so thought it might be a problem with insufficient hardware resources, so I got docker (WSL2) on my computer and it successfully downloads the model, but again it gets the same 500 error when trying to do anything. If it helps or changes anything, I am using an external library and not the Immich library for these photos.

I have tried adding firewall rules in Windows/DSM and using a different model but I still get those errors for both requests.

PC Immich ML logs:
[07/29/25 17:50:09] WARNING  Returning existing local_dir
                             `/cache/clip/ViT-SO400M-16-SigLIP2-384__webli` as
                             remote repo cannot be accessed in
                             `snapshot_download` (429 Client Error: Too Many
                             Requests for url:
                             https://huggingface.co/api/models/immich-app/ViT-SO
                             400M-16-SigLIP2-384__webli/revision/main).
[07/29/25 17:50:09] INFO     Attempt #2 to load visual model
                             'ViT-SO400M-16-SigLIP2-384__webli' to memory
[07/29/25 17:50:25] WARNING  Failed to load visual model
                             'ViT-SO400M-16-SigLIP2-384__webli'. Clearing cache.
[07/29/25 17:50:25] INFO     Cleared cache directory for model
                             'ViT-SO400M-16-SigLIP2-384__webli'.
[07/29/25 18:33:38] INFO     Shutting down due to inactivity.
[07/29/25 18:33:38] INFO     Shutting down
[07/29/25 18:33:38] INFO     Waiting for application shutdown.
[07/29/25 18:33:38] INFO     Application shutdown complete.
[07/29/25 18:33:38] INFO     Finished server process [9]
[07/29/25 18:33:38] ERROR    Worker (pid:9) was sent SIGINT!
[07/29/25 18:33:38] INFO     Booting worker with pid: 1138
[07/29/25 18:33:40] INFO     Started server process [1138]
[07/29/25 18:33:40] INFO     Waiting for application startup.
[07/29/25 18:33:40] INFO     Created in-memory cache with unloading after 300s
                             of inactivity.
[07/29/25 18:33:40] INFO     Initialized request thread pool with 20 threads.
[07/29/25 18:33:40] INFO     Application startup complete.

CONTAINER ID   IMAGE                                                     COMMAND                  CREATED             STATUS                       PORTS                                         NAMES
bd22bd817ae2   ghcr.io/immich-app/immich-machine-learning:release-cuda   "tini -- python -m i…"   About an hour ago   Up About an hour (healthy)   0.0.0.0:3003->3003/tcp, [::]:3003->3003/tcp   immich_machine_learning


immich_server logs to follow.

Any help would be greatly appreciated, sorry if this is a very obvious problem—I am surprisingly oblivious sometimes.
Was this page helpful?