#Setting up Immich remote ML

1 messages · Page 1 of 1 (latest)

rancid sage
#

Hi, I setuped a immich-machine-learning container on a remote machine so that my raspi is not overwhelmed.

Is this the correct output when ML should be running ?

[01/13/26 15:35:13] INFO Loading detection model 'PP-OCRv5_mobile' to memory

[01/13/26 15:35:13] INFO Setting execution providers to

                         ['CPUExecutionProvider'], in descending order of   

                         preference

[01/13/26 15:35:13] INFO Loading visual model 'ViT-B-32__openai' to memory

[01/13/26 15:35:13] INFO Setting execution providers to

                         ['CPUExecutionProvider'], in descending order of   

                         preference

[01/13/26 15:35:14] INFO Loading detection model 'buffalo_l' to memory

[01/13/26 15:35:14] INFO Setting execution providers to

                         ['CPUExecutionProvider'], in descending order of   

                         preference

[01/13/26 15:35:14] INFO Loading recognition model 'buffalo_l' to memory

[01/13/26 15:35:14] INFO Setting execution providers to

                         ['CPUExecutionProvider'], in descending order of   

                         preference

[01/13/26 15:35:15] INFO Loading recognition model 'PP-OCRv5_mobile' to

                         memory

[01/13/26 15:35:15] INFO Setting execution providers to

                         ['CPUExecutionProvider'], in descending order of   

                         preference

[INFO] 2026-01-13 15:35:15,640 [RapidOCR] base.py:22: Using engine_name: onnxruntime
`

velvet wadiBOT
#

:wave: Hey @rancid sage,

Thanks for reaching out to us. Please carefully read this message and follow the recommended actions. This will help us be more effective in our support effort and leave more time for building Immich immich.

References

#

Checklist

I have...

  1. :ballot_box_with_check: verified I'm on the latest release(note that mobile app releases may take some time).
  2. :ballot_box_with_check: read applicable release notes.
  3. :ballot_box_with_check: reviewed the FAQs for known issues.
  4. :ballot_box_with_check: reviewed Github for known issues.
  5. :ballot_box_with_check: tried accessing Immich via local ip (without a custom reverse proxy).
  6. :ballot_box_with_check: uploaded the relevant information (see below).
  7. :ballot_box_with_check: tried an incognito window, disabled extensions, cleared mobile app cache, logged out and back in, different browsers, etc. as applicable

(an item can be marked as "complete" by reacting with the appropriate number)

Information

In order to be able to effectively help you, we need you to provide clear information to show what the problem is. The exact details needed vary per case, but here is a list of things to consider:

  • Your docker-compose.yml and .env files.
  • Logs from all the containers and their status (see above).
  • All the troubleshooting steps you've tried so far.
  • Any recent changes you've made to Immich or your system.
  • Details about your system (both software/OS and hardware).
  • Details about your storage (filesystems, type of disks, output of commands like fdisk -l and df -h).
  • The version of the Immich server, mobile app, and other relevant pieces.
  • Any other information that you think might be relevant.

Please paste files and logs with proper code formatting, and especially avoid blurry screenshots.
Without the right information we can't work out what the problem is. Help us help you ;)

If this ticket can be closed you can use the /close command, and re-open it later if needed.

velvet wadiBOT
rancid sage
#

What is really strange is that I got this once:
Fetching 11 files: 0%| | 0/11 [00:00<?, ?it/s]

Fetching 11 files: 9%|▉ | 1/11 [00:00<00:02, 4.62it/s]

Fetching 11 files: 45%|████▌ | 5/11 [00:11<00:14, 2.42s/it]

Fetching 11 files: 100%|██████████| 11/11 [00:11<00:00, 1.04s/it]

Fetching 4 files: 0%| | 0/4 [00:00<?, ?it/s]

Fetching 4 files: 25%|██▌ | 1/4 [00:00<00:00, 7.61it/s]

Fetching 4 files: 75%|███████▌ | 3/4 [00:02<00:01, 1.10s/it]

Fetching 4 files: 100%|██████████| 4/4 [00:08<00:00, 2.57s/it]

Fetching 4 files: 100%|██████████| 4/4 [00:08<00:00, 2.09s/it]

#

And never again.

#

So what happens is that it crashes my raspi after like 150 + photos synchronised

#

so I'm trying to offput the load to my main computer

versed iron
#

It should only happen once

#

Your first post is the normal logs for a running ML container

solid breach
#

everything looks good, remember to add the url for the remote machine at the top

rancid sage
#

how long does that take usually ? Because i tried that for about 100 photos, and it never started

rancid sage
#

yes i set up the URL as well

#

any way to troubleshoot further?

solid breach
#

how long does what take?

  • you should attach your Immich ML settings
  • immich server logs when processing images
  • dockerfile used to configure the remote ML instance and information of platform it's running on
rancid sage
#

The remote machine is my main windows 11 computer. It runs docker desktop

#

Here is the dockerfile used :

#

name: immich_remote_ml

services:
immich-machine-learning:
container_name: immich_machine_learning
# For hardware acceleration, add one of -[armnn, cuda, rocm, openvino, rknn] to the image tag.
# Example tag: ${IMMICH_VERSION:-release}-cuda
image: ghcr.io/immich-app/immich-machine-learning:${IMMICH_VERSION:-release}
# extends:
# file: hwaccel.ml.yml
# service: # set to one of [armnn, cuda, rocm, openvino, openvino-wsl, rknn] for accelerated inference - use the -wsl version for WSL2 where applicable
volumes:
- model-cache:/cache
restart: always
ports:
- 3003:3003

volumes:
model-cache:

rancid sage
solid breach
rancid sage
# solid breach yes

i'll get them in a minute, i just have another weird error "Impossible de vérifier la version" litteraly "Impossible to verify version" in my IOS app, so i can't sync

#

lemme try to fix that

#

Setuping Immich remote ML

solid breach
#

setting up 😉

rancid sage
#

thx man

#

Setting up Immich remote ML

#

Here is the immich server logs, i attempted to sync it

#

and here is the ml remote

image.png
rancid sage
# solid breach yes

i let it run for about 15 photos upload because it crashes my Pi if i let it run for too long

solid breach
#

ML looks fine, probably your pi is getting overloaded. try reducing the number of jobs to 1

rancid sage
#

For which one?

#

it just crashed, here is the log

image.png
solid breach
#

all of them? pi are very limited, it will run, but barely.
try lowering everything and see if that helps, then you can start increasing one by one to test

rancid sage
#

i'll get starting on lowering them and testing

solid breach
rancid sage
#

yeah i get that

#

oooh actually you're right

#

it might be another jobs that clocks it

solid breach
rancid sage
#

locally

#

could it be thumbnail not generated because of a crash ?

solid breach
#

there are lot of interrupted uploads on your log. try lowering the jobs and see if that helps, maybe the pi can't handle more uploads and it's corrupting the files, causing the errors

rancid sage
#

oh okay

#

i'll try that

#

i just lowered them all

#

i'll test it now

#

it slow af but it doesn't seem to crash the pi

#

@solid breach could i put more faces jobs concurently ?

#

because i have remote ml ?

#

"faces job" sounds bad but you get what i mean 🤣

solid breach
#

yes, you could try increasing face detection

rancid sage
#

rn the bottleneck seems to be the video transcoding

#

but the pi is running at full capacity and that's the main task

solid breach
#

jobs are independent, so you shouldn't worry about that

rancid sage
solid breach
#

video transcoding requires a lot of rss, especially for a pi.
you should be able to increase thumbs generation safely

rancid sage
#

i made it 3

#

seems to be doing fine

solid breach
rancid sage
#

i need to upgrade

#

but i can't be bothered to set it up again

rancid sage
#

pov: my pi receiving 1500 photos