When I run the following code (which runs fine on Google Colab) my kernel crashes
from torchvision import datasets
from torchvision.transforms import ToTensor
import torch.nn as nn
import torch.nn as cnn
train_dataset = datasets.MNIST(
root = 'datasets',
train = True,
transform = ToTensor(),
download = True,
)
test_dataset = datasets.MNIST(
root = 'datasets',
train = False,
transform = ToTensor()
)
print(train_dataset)
print(test_dataset)
print(train_dataset.data.size())
print(test_dataset.data.size())
import matplotlib.pyplot as plot
plot.figure(1)
plot.imshow(train_dataset.data[0], cmap='gray')
plot.title('%i' % train_dataset.targets[0])
plot.show()
The very undescriptive error message i get is the following:
(This is the click here link shown in the screenshot-image above)
However if I run the following code a one or more cells above the previous code works without crashing the kernel. What is really difficult to understand is even with importing matplotlib
with different names doesn't effect this issue...
import matplotlib.pyplot as plt
plt.imshow([[1, 0], [0, 1]]) #this fixes the kernel from crashing if placed before previous code for some strange reason
I know some people (as seen here https://github.com/jupyter/notebook/issues/6219) suggest going back a version or two for different packages. This did not work for me. Since this code runs on colab it means there is some issue with my PC.
I have reinstalled Anaconda and all my python stuff and still am unsuccessful with solving this issue.
Currently my package versions are: Torch Version: 2.0.1+cu117 TorchVision Version: 0.15.2+cu117 Matplotlib Version: 3.5.0
source https://stackoverflow.com/questions/77035613/python-matplotlib-plt-imshow-crashes-kernel-in-an-unknown-way-when-plotting-imag
Comments
Post a Comment