Shoufa Chen

I'm wrong. the `nn.Parameter` are registered to `model.parameters()`. should print ``` model.parameters() ``` rather than `odel.parameter`

hello, @TheShadow29, @tdurieux . I have finished a simple project to download the repository from anonymous4open. https://github.com/ShoufaChen/clone-anonymous4open This implementation is not very elegant now but works properly. Hope it will...

The higher one is tested with multi-scale setting. https://github.com/hszhao/semseg/blob/master/config/cityscapes/cityscapes_pspnet50.yaml#L54

Hi, @rosscg Thanks for your contribution. I found that you import ``` from .const import get_username, get_password, get_email, get_geckdriver_path ``` but `.const` does not contains `get_geckdriver_path`. Could you help me?

Thank you @rosscg.

I solved this problem by replacing the `Correlation` package here with the one provided by FlowNet2. First, download [correlation_package](https://github.com/NVIDIA/flownet2-pytorch/tree/master/networks/correlation_package) to `PWC-Net/PyTorch/models/`. That looks like: ``` PWC-Net/PyTorch/models/ correlation_package PWCNet.py __init__.py ```...

I solved this issue by adding a `version.py` at `pose/mmpose/version.py` which contains: ``` __version__ = '0.13.0' short_version = __version__ ```

Thanks for your reply. So the utilization in the above table is calculated by the percentage of FLOPS instead of `nvidia-smi`. Is that right?

Hi, @unixpickle Thanks for your help. My command: ``` MODEL_FLAGS="--attention_resolutions 32,16,8 --class_cond True --diffusion_steps 1000 --image_size 128 --learn_sigma True --num_channels 256 --num_heads 4 --num_res_blocks 2 --resblock_updown True --use_fp16 True --use_scale_shift_norm...

Thanks for your help. I will patch this bug and try again. I will post my results in about 2 days.