Questions

About the Questions category (1)

Can Reduce-function return Tensors with different dimension? (7)

I am confused about the calculation order of GAT (2)

Question about update_all (2)

Why DGL is fast than the Pytorch or MXNet? (7)

Question about the tree-LSTM implementation (6)

How to read the source code? (2)

Implemention:copy-out vs. zero-copy (2)

Multiple-graph classification for graphs with different graph sizes (3)

Way to recursively 'split' graphs into subgraphs? (6)

Broadcast readout output to each node in BatchedDGLGraph? (2)

Question about the reduce function (7)

Problem to reproduce the result in DGL gcn_concat (15)

Dose DGL support distributed training & inference? (4)

Why is graph.send much slower when not using all edges? (7)

How to Implement the Renormalization Trick (4)

Is DGL support user define ops? (6)

Questions about BatchNode mailbox (7)

Source Node Features are Emptied (7)

How would you like your repo to be cited in papers? (2)

Users should not make assumption in the batch dimension (2)

How to fix random seeds in dgl in training? (2)

Is there any ERROR in the model SPARSITY? (10)

About setting initializer (3)

How to classify entire graphs in batches? (14)

How to join the development and is there any roadmap? (如何加入开发，项目计划有没有呢?) (4)

Error occurred when running the code of GCN from the tutorial (9)

AttributeError: 'NoneType' object has no attribute 'varcount' (3)

Would dgl.mean_nodes(batch_graph, ''hidden") break computational graph? (5)