You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
This repository was archived by the owner on Jun 11, 2026. It is now read-only.
I am trying to port the gnn model in this work into our code generation model which runs on JAVA. I implemented most inputs used in the gnn model in the repo. However, I am a bit confused by the use of num_eg_nodes / eg_node_labels in the data samples and how it connects to one of the inputs to the gnn (placeholders['eg_initial_node_ids']).
I also notice that there is a eg_node_offset, which is added to many inputs of gnn (such as eg_sending_node_ids, receiving_node_ids). This variable is also related to eg_node_labels.
I wonder if I can get more information on what num_eg_nodes / eg_node_labels represents in the input AST.
Hi,
Thanks again for releasing this amazing library!
I am trying to port the gnn model in this work into our code generation model which runs on JAVA. I implemented most inputs used in the gnn model in the repo. However, I am a bit confused by the use of num_eg_nodes / eg_node_labels in the data samples and how it connects to one of the inputs to the gnn (placeholders['eg_initial_node_ids']).
I also notice that there is a eg_node_offset, which is added to many inputs of gnn (such as eg_sending_node_ids, receiving_node_ids). This variable is also related to eg_node_labels.
I wonder if I can get more information on what num_eg_nodes / eg_node_labels represents in the input AST.