[Tutorial] Batched Graph Classification

[mufeili]

Description

Motivated by the forum discussion, we decide to add a tutorial on batched graph classification.

Checklist

Please feel free to remove inapplicable items for your PR.

• The PR title starts with [$CATEGORY] (such as [Model], [Doc], [Feature]])
• Changes are complete (i.e. I finished coding on this PR)
• All changes have test coverage
• Code is well-documented
• To the my best knowledge, examples are either not affected by this change,
  or have been fixed to be compatible with this change
• Related issue is referred in this PR

Changes

• Added the tutorial as the last one in the basics section.
• Modified docs/index.rst accordingly

[BarclayII]

LGTM.

Should we also mention that currently batched graphs does not work with torch.nn.DataParallel?

[jermainewang]

@BarclayII better not. The support will be included in #356 .

[jermainewang]

@mufeili updated the tutorial using the new dataset. The CI error should be resolved after #364 is merged.

[mufeili]

@jermainewang

• It now takes about within 30s to run the tutorial with some performance drop (argmax accuracy can still reach 80%). I added the note "Note that for deployment of the tutorial, we restrict our running time and you are likely to get a higher accuracy (80 % ~ 90%) than the ones printed below."
• I also made some small fix to the doc for the dataset.

If this sounds good to you, just go ahead and merge it.

[jermainewang]

Really great job @mufeili !