Gengzigang / PCT

This is an official implementation of our CVPR 2023 paper "Human Pose as Compositional Tokens" (https://arxiv.org/pdf/2303.11638.pdf)
MIT License
332 stars 21 forks source link

environment requirement versions #48

Open kristinbranson opened 6 days ago

kristinbranson commented 6 days ago

I needed to change a few package versions to get the tokenizer training to work (haven't tested the rest). I have CUDA 11.7 on my computer. My updated requirements file:

torch==1.13.0 --index-url https://download.pytorch.org/whl/cu117 torchvision==0.14.0 --index-url https://download.pytorch.org/whl/cu117 opencv-python numpy yapf==0.32.0 mmcv==1.7.0 mmpose==0.29.0 tqdm yacs json_tricks tensorboardX pycocotools
scipy timm albumentations==1.3.1