New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Running machine translation using different GNNs #536
Comments
Please try a smaller batch_size or try another GPU with larger memory. |
@AlanSwift I already tried a smaller batch size. What I find suprising is:
Its the same dataset. But GGNN and GraphSage fails to run while GCN and GAT works. So GGNN/GraphSage needs more resource for some reason? Super interested to know why? |
We haven't investigated the memory efficiency for dgl :). |
@AlanSwift I get this OOM error at runtime for GGNN:
Any idea? |
@AlanSwift came across this discussion at DGL: Memory consumption of the GGNN module |
It seems the dgl sacrifices memory efficiency for time efficiency. We will pay attention to this problem. Thank you for letting us know it! |
@AlanSwift can you please provide me with a fix/suggestion 🙏 |
@AlanSwift, this is interesting. I also faced the same problem. Wondering do you have any solution to this? |
@AlanSwift do you have a plan to address the GGNN implementation limitation? |
Currently, this is not on my plan since it is related to the DGL. |
❓ Questions and Help
I am running the NMT example on the same dataset with GNN variants:
While the execution runs with GCN, I get Out-of-Memory (OOM) for GGNN and GraphSage. Can anyone help me with this query?
The text was updated successfully, but these errors were encountered: