About GNN Re-ranking training program

[viet2411]

Hello @layumi , thank you for your work

I was trying to reproduce the result in paper "Understanding Image Retrieval Re-Ranking: A Graph Neural Network Perspective" using your pytorch code, but I'm having some trouble in running the program.

The program needs "market_88_test.pkl" as input data for re-ranking process, but I don't understand how to generate it properly.

Could you give some advices on how to use this code?

Thank you and best regards.

[viet2411]

Update:
I have found how the program works, but when I applied on University-1652 dataset, it returned CUDA out of memory error:

RuntimeError: CUDA out of memory. Tried to allocate 10.10 GiB (GPU 0; 11.91 GiB total capacity; 10.22 GiB already allocated; 667.00 MiB free; 10.31 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

It seems like the reranking process took too much memory.
Have you ever met this problem before? Could you give some advices to deal with this problem?
Thank you and best regards.

[layumi]

@viet2411
You may try to set the half float, i.e., float 16.