Found bug in pytorch reinforcement tutorials

reinforcement-learning
1

hello, i found bug in pytorch reinforcement tutorial, which is on pytorch.org.i run colab version and it gave me error on this code:
‘’’
env.reset()
plt.figure()
plt.imshow(get_screen().cpu().squeeze(0).permute(1, 2, 0).numpy(),

       interpolation='none')

plt.title(‘Example extracted screen’)
plt.show()
‘’’
and error says:
‘’’
Cannot connect to “None”
‘’’
i haven’t read code yet.so what is solution, is that a bug? and how to fix it.To see whole code, you can go to https://pytorch.org/tutorials/intermediate/reinforcement_q_learning.html, this is official website.

2

Which matplotlib version are you using?
Apparently the interpolation argument is not recognized.

3

i know, like i wrote in question i run the code on google colab.for more info go to the pytorch.org website, where the tutorials are.thanks

4

please help me, i just started RL, and that bug discouraged me :D.

5

The error message is still pointing to matplotlib, so which version are you using?

6

i am using 3.2.0 version.official tutorial uses that version.

7

Thanks for the information.
In that case I doubt the error is thrown from the posted lines of code but rather from the gym interaction.

8

thanks, but it is pytorch’s tutorial, i haven’t written any line of code.

9

I can reproduce the issue on Colab, while the tutorial runs on my local machine.
Could you please open an issue here?

10

I was facing the same problem so I opened the issue by following your link:

1 Like
11

abyaadrafid answered on it there on May 30:

!apt install xvfb -y
!pip install pyvirtualdisplay
!pip install piglet

from pyvirtualdisplay import Display
display = Display(visible=0, size=(1400, 900))
display.start()

using pyvirtualdisplay fixed it for me.

1 Like