Smart search doesn't work at all, returns "Failed to Search Smart - Immich Server Error" #10968

kikendo · 2024-07-09T04:15:35Z

kikendo
Jul 9, 2024

The bug

Whenever I try any search, the application gives this error. I have analyzed all the photos in the library for machine learning.

The OS that Immich Server is running on

Debian Bullseye

Version of Immich Server

v1.107.2

Version of Immich Mobile App

1.107.1 build.162

Platform with the issue

Server
Web
Mobile

Your docker-compose.yml content

Your .env content

Reproduction steps

Do a smart search like "car"

Relevant log output

[Nest] 17  - 07/09/2024, 4:00:29 AM   ERROR [Api:Error: Machine learning request '{"clip":{"textual":{"modelName":"ViT-B-32__openai"}}}' failed with status 500: Internal Server Error
    at MachineLearningRepository.predict (/usr/src/app/dist/repositories/machine-learning.repository.js:22:19)
    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
    at async MachineLearningRepository.encodeText (/usr/src/app/dist/repositories/machine-learning.repository.js:47:26)
    at async SearchService.searchSmart (/usr/src/app/dist/services/search.service.js:97:27)~pavzxtov] Failed to search smart
[Nest] 17  - 07/09/2024, 4:00:29 AM   ERROR [Api:Error: Machine learning request '{"clip":{"textual":{"modelName":"ViT-B-32__openai"}}}' failed with status 500: Internal Server Error
    at MachineLearningRepository.predict (/usr/src/app/dist/repositories/machine-learning.repository.js:22:19)
    at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
    at async MachineLearningRepository.encodeText (/usr/src/app/dist/repositories/machine-learning.repository.js:47:26)
    at async SearchService.searchSmart (/usr/src/app/dist/services/search.service.js:97:27)~pavzxtov] Error: Machine learning request '{"clip":{"textual":{"modelName":"ViT-B-32__openai"}}}' failed with status 500: Internal Server Error

Additional information

I just noticed that the container "immich_machine_learning" doesn't seem to have a port attached to it when I do "docker ps". I am not sure if these problems are related or if they are, how to fix them.

mertalev · 2024-07-09T05:03:14Z

mertalev
Jul 9, 2024
Maintainer

The error means something failed in the machine learning container. The machine learning logs would be helpful to see what’s going wrong. It could be running out of memory, failing to download the model, etc.

13 replies

kikendo Jul 17, 2024
Author

Oh, I see. Well, I just have a default setup. I didn't do anything to it, just followed the instructions on the Wiki for installation.

alextran1502 Jul 17, 2024
Maintainer

Do you mind spinning up a new instance to see if you are still facing the issue? What is your server spec?

kikendo Jul 17, 2024
Author

I'm not sure I can spin up a new instance at the moment, especially with the amount of photos this one has.
The server is running on system with a Xeon E5-2620 v2 @ 2.10GHz and 32GB of RAM. I don't see how that would affect the ability to download stuff though.

alextran1502 Jul 17, 2024
Maintainer

I mean just spinning up to test out the machine learning portion. And when you perform the search, what are the logs on machine-learning container?

kikendo Jul 17, 2024
Author

I posted the logs above for the machine learning container.

goodman-b · 2024-07-24T22:19:35Z

goodman-b
Jul 24, 2024

Similar issue here. Tested on privileged/unprivilaged Debian 12 and Ubuntu 22 containers on proxmox.
ML throws errors when using OpenVINO.

immich_machine_learning | [07/24/24 22:08:58] INFO Attempt #2 to load textual model 'ViT-B-32__openai' to memory
immich_machine_learning | [07/24/24 22:08:58] INFO Setting execution providers to ['OpenVINOExecutionProvider', 'CPUExecutionProvider'], in descending order of preference
immich_machine_learning | 2024-07-24 22:08:58.167388287 [E:onnxruntime:Default, env.cc:254 ThreadMain] pthread_setaffinity_np failed for thread: 158, index: 1, mask: {4, }, error code: 22 error msg: Invalid argument. Specify the number of threads explicitly so the affinity is not set.

immich_machine_learning | [07/24/24 22:14:54] ERROR Exception in ASGI application
immich_machine_learning |
immich_machine_learning | ╭─────── Traceback (most recent call last) ───────╮
immich_machine_learning | │ /usr/src/app/main.py:152 in predict │
immich_machine_learning | │ │
immich_server | [Nest] 18 - 07/24/2024, 10:14:54 PM DEBUG [Api:LoggingInterceptorwvneolal] POST /api/search/smart 200 365.96ms ::ffff:192.168.50.211
immich_server | [Nest] 18 - 07/24/2024, 10:14:54 PM VERBOSE [Api:LoggingInterceptorwvneolal] {"page":1,"withExif":true,"isVisible":true,"query":"dog"}
immich_server | [Nest] 18 - 07/24/2024, 10:14:54 PM DEBUG [Api:HttpExceptionFilter~wvneolal] HttpException(500) {"message":"Failed to search smart","error":"Internal Server Error","statusCode":500}
immich_server | [Nest] 18 - 07/24/2024, 10:14:54 PM ERROR [Api:Error: Machine learning request '{"clip":{"textual":{"modelName":"ViT-B-32__openai"}}}' failed with status 500: Internal Server Error

0 replies

kikendo · 2024-07-31T13:06:48Z

kikendo
Jul 31, 2024
Author

I'm still getting this error

2 replies

alextran1502 Jul 31, 2024
Maintainer

@kikendo did you remove the model_cache volume so that the new model get downloaded?

powellclan Aug 5, 2024

I have the same issue. I deleted the model_cache volume and re-created the container. Same result.

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Smart search doesn't work at all, returns "Failed to Search Smart - Immich Server Error" #10968

{{title}}

Replies: 3 comments 15 replies

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

{{title}}

Select a reply

Smart search doesn't work at all, returns "Failed to Search Smart - Immich Server Error" #10968

kikendo Jul 9, 2024

The bug

The OS that Immich Server is running on

Version of Immich Server

Version of Immich Mobile App

Platform with the issue

Your docker-compose.yml content

Your .env content

Reproduction steps

Relevant log output

Additional information

Replies: 3 comments · 15 replies

mertalev Jul 9, 2024 Maintainer

kikendo Jul 17, 2024 Author

alextran1502 Jul 17, 2024 Maintainer

kikendo Jul 17, 2024 Author

alextran1502 Jul 17, 2024 Maintainer

kikendo Jul 17, 2024 Author

goodman-b Jul 24, 2024

kikendo Jul 31, 2024 Author

alextran1502 Jul 31, 2024 Maintainer

powellclan Aug 5, 2024

kikendo
Jul 9, 2024

Replies: 3 comments 15 replies

mertalev
Jul 9, 2024
Maintainer

kikendo Jul 17, 2024
Author

alextran1502 Jul 17, 2024
Maintainer

kikendo Jul 17, 2024
Author

alextran1502 Jul 17, 2024
Maintainer

kikendo Jul 17, 2024
Author

goodman-b
Jul 24, 2024

kikendo
Jul 31, 2024
Author

alextran1502 Jul 31, 2024
Maintainer