MKLab-ITI/pygrank

Torch GNN support

maniospas opened this issue · 1 comments

maniospas commented

Current helper methods for GNNs are centered on tensorflow and keras. Create backend operations to abstract them so that they can also be implemented through torch. This needs to add a new test to make sure everything is working.

Related tests: tests.test_gnn.test_appnp

maniospas commented

  • Split tests in two to differentiate between backends., provided native-looking implementations and implemented an abstraction on the core library.

  • GNN training is not integrated as part of the backend but as part of the benchmarks module, since it does not provide base operations.

  • In the future, creating a different module (or core submodule) to host more GNN operations can be considered.

Related tests: tests.test_gnn.test_appnp_tf, tests.test_gnn.test_appnp_torch