Error in GPU selection

[Pacomito]

Hello,

Thank you very much for this software.

I just installed the plugin from napari and tried to run the Semantic Segmentation (APOC) with various 2D example data from napari.

When I click on 'run' after selecting the image and label I get this error :

File ~/anaconda3/envs/napari-env/lib/python3.9/site-packages/pyclesperanto_prototype/_tier0/_device.py:43, in get_device()
     41 def get_device() -> Device:
     42     """Get the current device GPU class."""
---> 43     return _current_device._instance or select_device()
        _current_device._instance = None

File ~/anaconda3/envs/napari-env/lib/python3.9/site-packages/pyclesperanto_prototype/_tier0/_device.py:72, in select_device(name=None, dev_type=None, score_key=None)
     68 except:
     69     pass
---> 72 device = filter_devices(name, dev_type, score_key)[-1]
        name = None
        dev_type = None
        score_key = None
     73 if _current_device._instance and device == _current_device._instance.device:
     74     return _current_device._instance

IndexError: list index out of range

Do you have any idea why and how I could debug it ?

Thank you in advance,
Pacome

[haesleinhuepf]

Hi @Pacomito ,

interesting issue! Could you please run this code in Python or a Jupyter notebook and paste its output?

import pyclesperanto_prototype as cle

print(cle.available_device_names())
print(cle.select_device())

Thanks!

Best,
Robert

[Pacomito]

Thanks for your answer,
Here is the output :

>>> print(cle.available_device_names())
cl_get_gt_device(): error, unknown device: 9bc8
[]
>>> print(cle.select_device())
cl_get_gt_device(): error, unknown device: 9bc8
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/localadmin/anaconda3/envs/napari-env/lib/python3.9/site-packages/pyclesperanto_prototype/_tier0/_device.py", line 72, in select_device
    device = filter_devices(name, dev_type, score_key)[-1]
IndexError: list index out of range
>>> 

BTW, I am on Ubuntu 22.04.1 LTS with Processor Intel® Core™ i5-10500 CPU @ 3.10GHz × 12 and graphic card Mesa Intel® UHD Graphics 630 (CML GT2)

[haesleinhuepf]

Ok, can you try to install the graphics cards driver and / or the ICD loader? See hints here:

https://github.com/clEsperanto/pyclesperanto_prototype#troubleshooting-graphics-cards-drivers

[Pacomito]

Now it works with:

>>> import pyclesperanto_prototype as cle

>>> 
>>> print(cle.available_device_names())
['pthread-Intel(R) Core(TM) i5-10500 CPU @ 3.10GHz']
>>> print(cle.select_device())
<pthread-Intel(R) Core(TM) i5-10500 CPU @ 3.10GHz on Platform: Portable Computing Language (2 refs)>

However, if i now open napari, load the sample image 'Lund' , make a label, select Semantic segmentation (APOC) and then run on 'Train' like below

napari closes with 'Segmentation error', here are the console logs:

WARNING: Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.
09:28:16 WARNING Warning: Ignoring XDG_SESSION_TYPE=wayland on Gnome. Use QT_QPA_PLATFORM=wayland to run on Wayland anyway.
/home/localadmin/anaconda3/envs/napari-env/lib/python3.9/site-packages/jupyter_client/__init__.py:23: UserWarning: Could not import submodules
  warnings.warn("Could not import submodules")
Requirement already satisfied: torch in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (1.8.1+cpu)
Requirement already satisfied: torchvision in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (0.9.1+cpu)
Requirement already satisfied: torchaudio in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (0.8.1)
Requirement already satisfied: grad-cam==1.4.6 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (1.4.6)
Requirement already satisfied: tqdm in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (4.64.1)
Requirement already satisfied: scikit-learn in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (1.2.1)
Requirement already satisfied: ttach in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (0.0.3)
Requirement already satisfied: numpy in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (1.23.5)
Requirement already satisfied: matplotlib in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (3.6.3)
Requirement already satisfied: opencv-python in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (4.5.5.62)
Requirement already satisfied: Pillow in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from grad-cam==1.4.6) (9.4.0)
Requirement already satisfied: typing-extensions in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from torch) (4.4.0)
Requirement already satisfied: fonttools>=4.22.0 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (4.38.0)
Requirement already satisfied: kiwisolver>=1.0.1 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (1.4.4)
Requirement already satisfied: python-dateutil>=2.7 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (2.8.2)
Requirement already satisfied: cycler>=0.10 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (0.11.0)
Requirement already satisfied: contourpy>=1.0.1 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (1.0.7)
Requirement already satisfied: pyparsing>=2.2.1 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (3.0.9)
Requirement already satisfied: packaging>=20.0 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from matplotlib->grad-cam==1.4.6) (23.0)
Requirement already satisfied: scipy>=1.3.2 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from scikit-learn->grad-cam==1.4.6) (1.10.0)
Requirement already satisfied: threadpoolctl>=2.0.0 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from scikit-learn->grad-cam==1.4.6) (3.1.0)
Requirement already satisfied: joblib>=1.1.1 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from scikit-learn->grad-cam==1.4.6) (1.2.0)
Requirement already satisfied: six>=1.5 in ./anaconda3/envs/napari-env/lib/python3.9/site-packages (from python-dateutil>=2.7->matplotlib->grad-cam==1.4.6) (1.16.0)
09:28:18 WARNING Java backend is not available. This could be due to a missing dependency (jpype).
train PixelClassifier
num images 1
object annotation value 2
features gaussian_blur=1 difference_of_gaussian=1 laplace_box_of_gaussian_blur=1 sobel_of_gaussian_blur=1
depth 2
num trees 100
file /home/localadmin/PixelClassifier.cl
annotation shape (100, 256, 256)
Erreur de segmentation (core dumped)

[haesleinhuepf]

Can you try if the classifier works from python, e.g. using this notebook?
https://github.com/haesleinhuepf/apoc/blob/main/demo/demo_object_segmenter.ipynb

[Pacomito]

Thank you for sticking with this issue,

Indeed running the script on jupyter notebook did not throw any error and the model managed to train.

I don't know then why napari crashes. I can run other modules such as CellPose or Stardist for example for nuclei detection without problems.

[haesleinhuepf]

Yes, for me it's also hard to debug remotely. If you have time, try different versions of napari (0.4.15, 0.4.17), qt and/or magicgui. And let us know if you figure out what's the problem