More Linux struggles to document. And more to come since I’ve been spending a lot of time with Tensorflow lately.
I’ve been training some CNNs lately that requires matplotlib() graphics to render. The problem is I can’t get X11 Quartz to show up when I run python scripty.py and want to visualize training at each epoch.
I kept getting this error about incorrect display end points.
File "/usr/lib/python3.5/tkinter/__init__.py", line 1871, in __init__
self.tk = _tkinter.create(screenName, baseName, className, interactive, wantobjects, useTk, sync, use)
_tkinter.TclError: couldn't connect to display ":0.0"
I tried just changing the backend. Turns out there are a lot of interesting matplotlib backend options that will alter where and when you view output of matplotlib graphics.
matplotlib.use('GTK'). Check out the Matplotlib FAQ.
Turns out the solution was as simple as setting secure X11 forwarding during login by adding a -Y parameter. Voila!
ssh -Y homer@simpsons.com
Special thanks to this useful writeup from the Indiana School of Informatics and Computing:
Now back to training!