GAT example not working

[alperyilmaz]

I changed the preprocessing part

train_X = Float32.(features) |> gpu  # dim: num_features * num_nodes
train_y = Float32.(labels) |> gpu  # dim: target_catg * num_nodes

into

train_X = Matrix{Float32}(features) |> gpu  # dim: num_features * num_nodes
train_y = Matrix{Float32}(labels) |> gpu  # dim: target_catg * num_nodes

in order to avoid some GPU errors. However, the training function is stuck at epoch 1 even after many minutes.

model(train_X) gives 56×1 array, but train_y is 7×2708 array. Is there a problem with data dimension or reshaping?

thanks.

[yuehhua]

Computing gradient for the whole graph could take longer time. It is suggested to compute less than 64 nodes for gradient for each updates. A proper node number to update would be 32. I will release a new feature for node sampling to act like batch size in classical deep learning model. It will update gradient for some nodes in a graph randomly.

[yuehhua]

Check new version here. If you have any questions, it's welcome to reopen the issue.