Stuck at "Checking libtensorflow" in WASM mode

[JartanFTW]

Which version of recognize are you using?

6.1.1

Enabled Modes

Object recognition, Face recognition, Music recognition

TensorFlow mode

WASM mode

Downstream App

Memories App

Which Nextcloud version do you have installed?

28.0.3

Which Operating system do you have installed?

Ubuntu 20.04.6 LTS

Which database are you running Nextcloud on?

PostgreSQL 15.6

Which Docker container are you using to run Nextcloud? (if applicable)

Nextcloud AIO 28.0.3

How much RAM does your server have?

8GiB

What processor Architecture does your CPU have?

x86_64

Describe the Bug

1. "Checking libtensorflow" never completes.
2. I get the message "Could not check whether your machine supports native TensorFlow operation. Make sure your OS has GNU lib C, your CPU supports AVX instructions and you are running on x86. If one of these things is not the case, you will need to run in WASM mode." I have manually verified that my server meets each of these conditions.
3. Human action recognition is greyed out and cannot be enabled, I presume because of the libtensorflow check hanging.

The other checks appear to be functional. "Node.js v14.18.2 binary was installed successfully." and the Nice binary check doesn't say anything.

Expected Behavior

1. "Checking libtensorflow" should finish.
2. The system should recognize my server's ability to support native TensorFlow operation.
3. The Human Action Recognition feature should be toggleable.

To Reproduce

N/A, but here's some extra info: I'm using NextCloud AIO. I installed recognize after the initial AIO setup. Face, Object, Audio, and Landmark recognition are all working properly.

Debug log

I cannot find "debug" anywhere in config.php to enable. Please let me know what I need to do to provide any necessary logs for this.

[github-actions[bot]]

Hello :wave:

Thank you for taking the time to open this issue with recognize. I know it's frustrating when software causes problems. You have made the right choice to come here and open an issue to make sure your problem gets looked at and if possible solved. I try to answer all issues and if possible fix all bugs here, but it sometimes takes a while until I get to it. Until then, please be patient. Note also that GitHub is a place where people meet to make software better together. Nobody here is under any obligation to help you, solve your problems or deliver on any expectations or demands you may have, but if enough people come together we can collaborate to make this software better. For everyone. Thus, if you can, you could also look at other issues to see whether you can help other people with your knowledge and experience. If you have coding experience it would also be awesome if you could step up to dive into the code and try to fix the odd bug yourself. Everyone will be thankful for extra helping hands! One last word: If you feel, at any point, like you need to vent, this is not the place for it; you can go to the forum, to twitter or somewhere else. But this is a technical issue tracker, so please make sure to focus on the tech and keep your opinions to yourself. (Also see our Code of Conduct. Really.)

I look forward to working with you on this issue Cheers :blue_heart:

[marcelklehr]

Hi @JartanFTW

1. "Checking libtensorflow" should finish.

This is odd indeed.

2. The system should recognize my server's ability to support native TensorFlow operation.

This just doesn't work sometimes, but shouldn't prevent you from doing anything. You can disable WASM mode manually anyway.

3. The Human Action Recognition feature should be toggleable.

If you disable WASM mode this should be toggleable again. The checks are just informative in nature, they do not enforce anythig.

[JartanFTW]

@marcelklehr

This just doesn't work sometimes, but shouldn't prevent you from doing anything. You can disable WASM mode manually anyway.

Ok, fair. I can disable WASM mode manually, and everything already working continues to work with WASM disabled.

If you disable WASM mode this should be toggleable again. The checks are just informative in nature, they do not enforce anythig.

Unfortunately, this is not what I am experiencing. Upon disabling WASM mode, the Human Action Recognition feature does not become toggleable.

Please let me know how to proceed.

[ShapeShifter499]

I tried disabling "WASM Mode" as my CPU supports AVX but I'm not sure if that means I can run the "Human Action Recognition feature", it stays grayed out regardless of if GPU mode is enabled or not. Does that feature require a GPU with CUDA cores?

EDIT: Should add that I am also using Nextcloud AIO but my specs are:

OS: Arch Linux x86_64 
Host: Stumpy 1.0 
Kernel: 6.8.7-arch1-2 
Uptime: 2 days, 23 hours, 26 mins 
Packages: 1406 (pacman) 
Shell: bash 5.2.26 
Terminal: /dev/pts/1 
CPU: Intel i5-2450M (4) @ 3.100GHz 
GPU: Intel 2nd Generation Core Processor Family 
Memory: 5172MiB / 15876MiB 

[marcelklehr]

Ah, it seems that x86_64 is required for Human action recognition as well. You said you have x86_64 though. Can you check in the network tab of the web inspector for the recognize settings page for a network request to "platform" and paste the response of that network request here?

[marcelklehr]

This will probably be fixed by the next release of which will be v7

[JartanFTW]

Can confirm that this particular issue seems to be fixed in v7. However, because I am using AIO, I now get the issue presented in #1129.