How to pad the graph if the size of each graph is different

[Barcavin]

Hi,

I am trying to implement the SQL-to-Text paper with the Graph2Seq model on WikiSQL dataset. When preparing the g_ids for the graph, it seems that we need to pad the nodes from all the graphs to the same size. I am wondering if I just need to create PAD nodes isolated or the PAD nodes have to be connected to the graph in some way.

Thanks

[694040837]

Hi,
I have the same question now. Do you solve this problem now?

[Barcavin]

not really. tried several ideas but none can reproduce the paper

…

On Tue, Jan 12, 2021 at 3:15 AM Listener ***@***.***> wrote: Hi, I have the same question now. Do you solve this problem now? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#7 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGNQQITC2XZPTYA3TPPDNPDSZQALLANCNFSM4OFEL2RQ> .

[694040837]

all right..Does the way 'create PAD nodes isolated' work? I tried this way, but my acc is 0....maybe some bugs happen

[teddylfwu]

Dear All,

Please see our newly released graph4nlp library: https://github.com/graph4ai/graph4nlp, which have implemented many GNN methods like Graph2Seq and Graph2Tree models.

[LN-Curiosity]

all right..Does the way 'create PAD nodes isolated' work? I tried this way, but my acc is 0....maybe some bugs happen

hi, do you have other solutions to deal with this "PAD nodes" problem?

[LN-Curiosity]

not really. tried several ideas but none can reproduce the paper
…
On Tue, Jan 12, 2021 at 3:15 AM Listener @.***> wrote: Hi, I have the same question now. Do you solve this problem now? — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#7 (comment)>, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGNQQITC2XZPTYA3TPPDNPDSZQALLANCNFSM4OFEL2RQ .

hi, do you have a new idea to work out this problem?