Training time and logs

[Jarvis73]

Hi, Seokju-Cho

Good work! This is a quite surprising job for few-shot segmentation.

I am trying to re-train VAT on PASCAL, but I find that it may take about 5 days on 4 Tesla V100 GPUs for 300 epochs.

• I wonder how long did it use in your experiments (with 3090 or other GPU)?
• And could you share a copy of the training log of the experiment for reference?

Thanks a lot!

[Anant4830]

Hello Jarvis,

I am also trying to train on pascal dataset. I am getting an error and not able to enter the for loop. Here is the code:- for idx, batch in enumerate(dataloader):

1. Hypercorrelation Squeeze Networks forward pass

    print("idx and batch are ========= ", idx,"\t", batch)
    batch = utils.to_cuda(batch)
    logit_mask = model(batch['query_img'], batch['support_imgs'].squeeze(1), batch['support_masks'].squeeze(1))
    pred_mask = logit_mask.argmax(dim=1)

Here, I am not able to enter the for loop.

How did you solve it? any solution is appreciated.