Yao Li

Demo code

8

I think it is better to have some demo code in this package to visualize the segmentation output from hrnet.

No Depth-wise convolution for BiFPN?

1

The EfficientDet paper uses depth-wise convolution of BiFPN, however, in this implementation, clearly depth-wise convolution is not used (in module.py). `conv_cfg = { 'Conv': nn.Conv2d, 'ConvWS': ConvWS2d, # TODO: octave...

data loading error

2

I did not change the code. Got data loading error: trainer.train() File "train.py", line 115, in train for step, batch in batch_iterator: File "/usr/local/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 345, in __next__ data =...

size mismatch error

3

When I run the "python3 app.py" for demo, it cannot load the pretrained model naver-clova-ix/donut-base-finetuned-docvqa, there is a size miss match error pretrained_model = DonutModel.from_pretrained(args.pretrained_path) File "/home/local/Project/chart/donut/donut/model.py", line 597, in...

problem of running the code on Berkeley Attributes of People Dataset

2

I am trying to run the code on the Berkeley Attributes of People Dataset (BAPD) . After downloading the BAPD dataset and pre-trained models from the provided links, I modified...