GNNLux feature parity

[CarloLucibello]

• heterograph convolutional layers
• tutorials and examples ([lux] add tutorials #544)
• pooling layers

[aurorarossi]

To keep track of the progress made.
Pooling layers:

• GlobalAttentionPool
• GlobalPool Add Graph Classification tutorial #568
• Set2Set
• TopKPool

Tutorials:

• Hands on Add Hands On tutorial GNNLux #549
• Node classification Add Node Classification tutorial GNNLux  #567
• Graph classification Add Graph Classification tutorial #568
• Traffic prediction (node autoregression)
• Temporal graph classification