Stable-Pix2Seq

Stable-Pix2Seq copied to clipboard

您好，请问还在维护这个项目吗

ross-Hr

I want to extract the token embedding as shown in figure 11 of the paper.  However, when looking at the code, I see that the tokens are predicted by...

elituan

CUDA Out-of-memory using V100

1

I'm using V100 for experiments, but still out of memory in the middle of the training process. Not sure what would be the reason at this momnet ```bash Namespace(aux_loss=True, backbone='resnet50',...

allanj

trained model

1

Is there any trained model for testing? Thank you！

opentld

I was training 'Stable Pix2Seq', everything goes fine until the 3rd training epoch. I wonder if there's any accumulate operation or some tensors or variable should have been deleted.

yshMars

In the original paper, learning rate was set to 3e-3 and weight decay was set to 5e-2, why do u use the learning rate 1e-5 and weight decay 1e-4 in...

youngsheen

Thank you for your work, I have a question about sequence embedding. The screenshot is from transformer.py When you get sequence embedding, the position embedding has already been added to...

lumiaomiao

Does it support instance segmentation？

2

I try to use this code for instance segmentation。I using the command ：python -m torch.distributed.launch --nproc_per_node=1 --use_env main.py --coco_path ... --masks But I have encountered some errors. Does this code...

Eamon-UCAS

Why use MLP/FFN during training only ?

1

https://github.com/gaopengcuhk/Stable-Pix2Seq/blob/12587302a2b697e2be8c131452e466a3f45c8c3e/models/detr.py#L64 Why use MLP/FFN during training only ?

karansapra