Time to event prediction using Graph Neural Networks

Hi! I am trying to extend the tutorials 3 and 4 to implement the Logistic Hazard loss in a Graph Neural Network for Graph-level prediction of survival. This is the[ example](https://docs.dgl.ai/tutorials/blitz/5_graph_classification.html#sphx-glr-tutorials-blitz-5-graph-classification-py) I am following.

Nevertheless, I am getting some errors when training the network. Specifically in "model.fit_dataloader" where I get:

```
AttributeError Traceback (most recent call last)
in ()
2 epochs = 50
3 verbose = True
----> 4 log = model.fit_dataloader(dl_train, epochs, callbacks, verbose, val_dataloader=dl_test)

13 frames
/usr/local/lib/python3.7/dist-packages/torchtuples/tupletree.py in shapes_of(data)
109 def shapes_of(data):
110 """Apply x.shape to elemnts in data."""
--> 111 return data.shape
112
113 @apply_leaf

AttributeError: 'DGLHeteroGraph' object has no attribute 'shape'
```

Do you have any suggestions on how to fix this error and/or approach this issue?

Thanks!

Start by reading tutorials 3 and 4 alongside the linked DGL graph-classification example. Reproduce the failure at model.fit_dataloader with the DGLHeteroGraph input and determine the scope of supporting Logistic Hazard loss for graph-level survival prediction. Done means the requested graph neural network training approach works without the reported shape error.

Written by the indexing model from the issue text.