AttributeError: 'FigureCanvasAgg' object has no attribute 'tostring_rgb'

[Mehmet2407]

Whenever I press "train model" it goes up to 5.3 seconds and then says error on the progress bar. The command window shows this error in the last line. I have no idea what's wrong

[kunibald413]

same here

[Sol-III-Observer]

+1

[kunibald413]

i think this might be only related to tensorboard logs
if you don't need them you can just comment out the code or skip it

in infer/lib/train/utils.py find the summarize function and try catch it

def summarize(
    writer,
    global_step,
    scalars={},
    histograms={},
    images={},
    audios={},
    audio_sampling_rate=22050,
):
    try:
        for k, v in scalars.items():
            writer.add_scalar(k, v, global_step)
        for k, v in histograms.items():
            writer.add_histogram(k, v, global_step)
        for k, v in images.items():
            writer.add_image(k, v, global_step, dataformats="HWC")
        for k, v in audios.items():
            writer.add_audio(k, v, global_step, audio_sampling_rate)
    except Exception as e:
        print(f"Exception when summarizing  {e}. Skip, we don't care!")

[5uperninj4]

+1. Looking for solution

[xpdd123]

pip install matplotlib==3.7.0,
It work for me

[TheTrustedComputer]

I also faced this error in the terminal window once training started. The workaround is downgrading matplotlib to 3.9 with: pip install matplotlib<3.10, likely due to breaking API changes with 3.10.

[anurag12-webster]

same, downgrading to pip install matplotlib==3.7.0 worked for me,

[Charles-yueyue831]

pip install matplotlib==3.7.0, It work for me

I agree with you

[cahayamahardika16]

But you got this error thing was matplotlib this expired.

[meisei4]

I think this might be related? (matplotlib/matplotlib#25484), I am not using this repos software, but i am using matplotlib version 3.10 and getting this error from trying to show plots in the jetbrains IDE PyCharm 2024.1.3 (Professional Edition, Runtime version: 17.0.11+1-b1207.24 x86_64, macOS 12.7.6) viewer:

trace:

...
lib/python3.12/site-packages/matplotlib/pyplot.py", line 614, in show
    return _get_backend_mod().show(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/Applications/PyCharm.app/Contents/plugins/python/helpers/pycharm_matplotlib_backend/backend_interagg.py", line 41, in __call__
    manager.show(**kwargs)
  File "/Applications/PyCharm.app/Contents/plugins/python/helpers/pycharm_matplotlib_backend/backend_interagg.py", line 144, in show
    self.canvas.show()
  File "/Applications/PyCharm.app/Contents/plugins/python/helpers/pycharm_matplotlib_backend/backend_interagg.py", line 85, in show
    buffer = self.tostring_rgb()
             ^^^^^^^^^^^^^^^^^
AttributeError: 'FigureCanvasInterAgg' object has no attribute 'tostring_rgb'. Did you mean: 'tostring_argb'?

Process finished with exit code 1
...

Workaround is using matplotlib version 3.9.4 and below i think... (at least i tested with 3.9.4 and that solved my issue

[cahayamahardika16]

I got failure Preprocessing and Feature Extraction "AttributeError: 'NoneType' object has no attribute 'loader'"

[BoyceCecil]

I am pretty sure the update of the matplotlib break its support with seaborn

[ikamensh]

someone should report this to matplotlib. I got this error while plotting generic data.

[FHeizenberg]

I had the same problem and the following helps me:
`import matplotlib

matplotlib.use('TkAgg')`
You can try to select another backend manually, here is the documentation https://matplotlib.org/stable/users/explain/figure/backends.html