search

DeepGraphLearning / GearNet

GearNet and Geometric Pretraining Methods for Protein Structure Representation Learning, ICLR'2023 (https://arxiv.org/abs/2203.06125)
MIT License
253 stars 28 forks source link

What information that the hidden dimensions respectively represent #28

Closed BenardChase closed 1 year ago

BenardChase commented 1 year ago

Hello! GearNet is really a good work! But I have a problem. I see that the hidden dimension set in the config file is [512,512,512,512,512,512]. Since I don't know much about the specific principle of graph neural network, I want to know what information these dimensions respectively represent.Thank you!

Oxer11 commented 1 year ago

Hi, thanks for your question! These denote the hidden dimensions of the graph neural network, which means the network will have six hidden layer with 512 dimensional representations. You can find the detailed description in the doc.

BenardChase commented 1 year ago

Thank you for your kind reply.But in the GearNet instance mentioned in another document, the dimensions of the hidden dimensions were only set as [512, 512, 512]. I would like to know what the difference between them.

Oxer11 commented 1 year ago

The tutorial you refer to is only for illustration, so we only use a shallow graph neural network to run on colab. In practice, we need at least 6 layers to get good performance.