Fan-Yun Sun

For mini-batch training, refer to the following snippet ``` train_edges = [(u,v) for u,v in G.edges()] n_nodes = G.number_of_nodes() for epoch in range(epochs): np.random.shuffle(train_edges) batch_edges = train_edges[:batch_size] batch = torch.LongTensor(batch_edges)...

For negative sampling, you can refer to this repository: https://github.com/DMPierre/LINE

Note that it will take a longer time to observe increase in nmi and modularity if you use negative sampling.

The code for node classification task is not in this repo. We use LibLinear package.

Results reported on the paper are obtained through extensive parameter search.

I believe you can solve this by [starting the controller under a headless setup](https://ai2thor.allenai.org/ithor/documentation/#headless-setup). Try adding the argument `platform=CloudRendering` here: https://github.com/allenai/Holodeck/blob/db15e52238135b9595bbd344eadde81b8a1b25ff/modules/small_objects.py#L220

Same issue here.

Hey Jonathan! Previously we had to export the scenes from Unity using the FBX Exporter Package. I believe you can technically write C# code to programmatically export a scene with...

Indeed, it can take a while to obtain results on the NBA dataset. If you would like to expedite it, consider - increasing the initial learning rate (and increasing the...

That is to convert the unit from Foot to Meter.