fix(docs): Add screenshot of cross-validation plot

This is a quick fix to address #594.

In the end, I took a screenshot manually (by using the "Download plot as png" button of the Plotly widget) and displayed it with matplotlib.

Here are all the alternatives I considered, and what didn't work with them:
- Use SVG instead of PNG
	- For that, I'd need to use Plotly's `fig.write_image`, and the library that Plotly uses for this, Kaleido, is currently broken: plotly/Kaleido#210
- Export plotly figure to a file during the example and read it 
	- For that I also need Plotly's `fig.write_image`
- Use Pillow for displaying, instead of matplotlib (which is not really made for that)
	- Only matplotlib is supported by sphinx-gallery, the library that renders the examples

examples/plot_03_cross_validate.py

@@ -36,3 +36,18 @@
 cv_results = cross_validate(lasso, X, y, cv=3, project=my_project_gs)
 
 my_project_gs.get_item("cross_validation").plot
+
+# %%
+# Because Plotly graphs currently do not yet properly render in our docs engine due to `a bug in Plotly <https://github.com/plotly/plotly.py/issues/4828>`_,
+# we also show a screenshot below.
+# Alternatively, if you zoom in or out the Plotly graph should display properly again.
+
+import matplotlib.pyplot as plt
+import matplotlib.image as mpimg
+
+img = mpimg.imread("plot_03_cross_validate_plot_screenshot.png")
+fig, ax = plt.subplots()
+# fig.tight_layout(pad=0.01)
+fig.subplots_adjust(left=0.01, right=0.99, bottom=0.01, top=0.99)
+ax.axis("off")
+ax.imshow(img)