Hi fellow self-hoster.
Almost one year ago i did experiment with Immich and found, at the time, that it was not up to pair to what i was expecting from it. Basically my use case was slightly different from the Immich user experience.
After all this time i decided to give it another go and i am amazed! It has grown a lot, it now has all the features i need and where lacking at the time.
So, in just a few hours i set it up and configured my external libraries, backup, storage template and OIDC authentication with authelia. All works.
Great kudos to the devs which are doing an amazing work.
I have documented all the steps of the process with the link on top of this post, hope it can be useful for someone.
I’m curious;
Which ML CLIP model did you go with, and how accurate are you finding the search results?
I found the default kinda sub-par, particularly when it came to text in images.
Switched to “immich-app/XLM-Roberta-Large-Vit-B-16Plus” and it’s improved a bit; but I still find the search somewhat lacking.
The best one I have found was one of the newer ones that was added a few months ago. ViT-B-16-SigLIP__webli
Really impressed with the accuracy even with multi word search like “espresso machine”
How well does it do with text in images?
I often find searching for things like ‘horse’ will do a decent job bringing up images of horses, but will often miss images containing the word ‘horse’.
It does ok with that. better than the default model, but worse than the built in search on my phone.