Yolov12 attribute Error

[DolapoSalim]

I tried to run inference on my custom YOLOV12 model but keep running into this error

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
Cell In[7], [line 1](vscode-notebook-cell:?execution_count=7&line=1)
----> [1](vscode-notebook-cell:?execution_count=7&line=1) results = model.predict(image_path)
      [2](vscode-notebook-cell:?execution_count=7&line=2) results[0].show()

File c:\Users\xxxxx\anaconda3\envs\opencv-env\Lib\site-packages\ultralytics\engine\model.py:550, in Model.predict(self, source, stream, predictor, **kwargs)
    [548](file:///C:/Users/xxxx/anaconda3/envs/opencv-env/Lib/site-packages/ultralytics/engine/model.py:548) if prompts and hasattr(self.predictor, "set_prompts"):  # for SAM-type models
    [549](file:///C:/Users/xxx/anaconda3/envs/opencv-env/Lib/site-packages/ultralytics/engine/model.py:549)     self.predictor.set_prompts(prompts)
--> [550](file:///C:/Users/x/axxxnaconda3/envs/opencv-env/Lib/site-packages/ultralytics/engine/model.py:550) return self.predictor.predict_cli(source=source) if is_cli else self.predictor(source=source, stream=stream)

File c:\Users\dolap\anaconda3\envs\opencv-env\Lib\site-packages\ultralytics\engine\predictor.py:214, in BasePredictor.__call__(self, source, model, stream, *args, **kwargs)
    [212](file:///C:/Users/xxx/anaconda3/envs/opencv-env/Lib/site-packages/ultralytics/engine/predictor.py:212)     return self.stream_inference(source, model, *args, **kwargs)
    [213](file:///C:/Users/x/xxanaconda3/envs/opencv-env/Lib/site-packages/ultralytics/engine/predictor.py:213) else:
--> [214](file:///C:/Users/xxx/anaconda3/envs/opencv-env/Lib/site-packages/ultralytics/engine/predictor.py:214)     return list(self.stream_inference(source, model, *args, **kwargs))

File c:\Users\xxxxx\anaconda3\envs\opencv-env\Lib\site-packages\torch\utils\_contextlib.py:36, in _wrap_generator.<locals>.generator_context(*args, **kwargs)
     [33](file:///C:/Users/x/axxxnaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:33) try:
     [34](file:///C:/Users/xxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:34)     # Issuing `None` to a generator fires it up
     [35](file:///C:/Users/xxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:35)     with ctx_factory():
---> [36](file:///C:/Users/xxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:36)         response = gen.send(None)
     [38](file:///C:/Users/xxxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:38)     while True:
     [39](file:///C:/Users/xxxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:39)         try:
     [40](file:///C:/Users/xxxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/utils/_contextlib.py:40)             # Forward the response to our caller and get its next request
...
-> [1928](file:///C:/Users/x/xxxxanaconda3/envs/opencv-env/Lib/site-packages/torch/nn/modules/module.py:1928) raise AttributeError(
   [1929](file:///C:/Users/xxxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/nn/modules/module.py:1929)     f"'{type(self).__name__}' object has no attribute '{name}'"
   [1930](file:///C:/Users/xxxxx/anaconda3/envs/opencv-env/Lib/site-packages/torch/nn/modules/module.py:1930) )

AttributeError: 'AAttn' object has no attribute 'qkv'

[miftahurrrizki]

I have the same problem, have you solved it? 😢

[DolapoSalim]

I have the same problem, have you solved it? 😢

No, I haven’t

[mehmeteaydin]

Could anyone solve it ? I have same problem

[panchaldhruv27223]

!! ALERT ---- URGENT

I ALSO HAVE THE SAME ISSUE.
I HAVE TRIED ALL THE POSSIBLE STEPS TO REOSLVE THIS.
MY MODEL IS LOADING FINE BUT AT THE FINETUNING TIME OR PREDICTION TIME MODEL IS GIVING THIS ERROR.
LAST TIME WHEN I USED IN MARCH MONTH IT IS WORKING FINE BUT NOW MY SCRIPT IS GIVING THIS ERRROR.
CAN ANYONE PLEASED GUIDE ME.

[SkalskiP]

@DolapoSalim, @miftahurrrizki, @mortvivant, @panchaldhruv27223 could any of you update installation command to pip install -q git+https://github.com/sunsmarterjie/yolov12.git@3a336a4adf3683d280e1a50d03fa24bbe7f24a5b and check if that solves the issue?

[miftahurrrizki]

@DolapoSalim, @miftahurrrizki, @mortvivant, @panchaldhruv27223 could any of you update installation command to pip install -q git+https://github.com/sunsmarterjie/yolov12.git@3a336a4adf3683d280e1a50d03fa24bbe7f24a5b and check if that solves the issue?

Thanks, my problem is solved with this!

[DolapoSalim]

@DolapoSalim, @miftahurrrizki, @mortvivant, @panchaldhruv27223 could any of you update installation command to pip install -q git+https://github.com/sunsmarterjie/yolov12.git@3a336a4adf3683d280e1a50d03fa24bbe7f24a5b and check if that solves the issue?

@SkalskiP Now, the inferencing works fine. Thank you!

[Someet-consuma]

@DolapoSalim, @miftahurrrizki, @mortvivant, @panchaldhruv27223 could any of you update installation command to pip install -q git+https://github.com/sunsmarterjie/yolov12.git@3a336a4adf3683d280e1a50d03fa24bbe7f24a5b and check if that solves the issue?

Hey i tried this thing but still not able to solve it, getting the same error

[mehmeteaydin]

@DolapoSalim, @miftahurrrizki, @mortvivant, @panchaldhruv27223 could any of you update installation command to pip install -q git+https://github.com/sunsmarterjie/yolov12.git@3a336a4adf3683d280e1a50d03fa24bbe7f24a5b and check if that solves the issue?

Hey i tried this thing but still not able to solve it, getting the same error

Interesting, when I tried `pip install -q git+https://github.com/sunsmarterjie/yolov12.git@3a336a4adf3683d280e1a50d03fa24bbe7f24a5b'
it works. Are u sure that u trained the model on right way ?

[Someet-consuma]

Yes I'm able to do the inference also but when i tried to copy the model file put it on different machine, I'm getting the error

[mehmeteaydin]

Hmm, I'll research for solution, If I can find it , I'll write comment again

[Someet-consuma]

Thank you

[Andrii0607]

@Someet-consuma
I had the same problem with you.
As has been mentioned in this issue, probably you used wrong weights for training.
You should not use the "Turbo" version weights. Use the "v1.0" weights instead.

[faruktkgl]

Go to ultralytics/nn/modules/block.py in the Yolov12 folder. Add the following line of code to the Attn class: self.qkv = self.qk
your problem should be solved.