NameError: name '_update_fn' is not defined when using the SAM

[wingvortex]

Hi,

I installed the ml-backend according to the official steps: git clone https://github.com/HumanSignal/label-studio-ml-backend.git cd label-studio-ml-backend/ pip install -e .

Then I installed SAM and started label-studio-ml: pip install -r label_studio_ml/examples/segment_anything_model/requirements.txt label-studio-ml start label_studio_ml/examples/segment_anything_model

It resulted in a success that shows it was running on http://127.0.0.1:8080

But when I started the label studio, and config the SAM model:

I get this error: label-studio-ml-backend/label_studio_ml/model.py", line 112, in fit if _update_fn: NameError: name '_update_fn' is not defined

Could you help with this issue?

[wingvortex]

I understand the suggested approach is to install by Docker, but as I'm working with a Redhat linux, it needs sudo to run podman which I do not have. Anyway, the previous step seems very likely to work, the SAM is connected correctly: But whenever I click the items in 'ML-Assisted Labeling' or 'Save' during the configuration of ML-backend, I would get the previous error in the terminal:

And the SAM auto annotation could not give out any result as well.

[xiahaifeng1995]

I've encountered the same issue. Have you managed to resolve it?

[wingvortex]

Not yet.

[oltiMetdaan]

Has anyone resolved this?

[wonka80]

same error

[song-GAO-1992]

same error

[wonka80]

收到您的邮件，谢谢！

[bcwein]

Same error here.

Hope this will be resolved.

[wonka80]

收到您的邮件，谢谢！

[CLF3721]

Same error here, as well.

[wonka80]

收到您的邮件，谢谢！

[korneliaWatson]

Same error when attempting to follow instructions for using tesseract backend.

[goodfella47]

A workaround that worked for me is to add a dummy fit method to the SamMLBackend class in segment_anything_model/model.py

def fit(self, event, data, **additional_params):
    pass