Windows environment

[Teravus]

Hey there, Just a FYI, this works for me in Windows with Cuda 11.1 and a 3090.

conda create -n e2fgvi python=3.7 conda activate e2fgvi python -m pip install --upgrade pip pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts matplotlib==3.4.1 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts opencv-python==4.5.5.62 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts vispy==0.9.3 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts transforms3d==0.3.1 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts networkx==2.3 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts scikit-image==0.19.2 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts pyaml==21.10.1 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts torch==1.9.0+cu111 torchvision==0.10.0+cu111 torchaudio==0.9.0 -f https://download.pytorch.org/whl/torch_stable.html pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts moviepy==1.0.3 pip install --no-cache-dir --ignore-installed --force-reinstall --no-warn-conflicts pyqt6==6.3.0 pip install mmcv-full -f https://download.openmmlab.com/mmcv/dist/cu111/torch1.9/index.html pip install tensorboard matplotlib

The original environment solve bug thread resulted in getting to the inference portion after an abnormally long time with no feedback.. and then it crashing saying that there was no kernel on the device.

After setting up the environment this way, python test.py --model e2fgvi_hq --video examples/tennis --mask examples/tennis_mask --ckpt release_model/E2FGVI-HQ-CVPR22.pth processed the test video in about 10 seconds.

[LownyCGI]

After this i run

pip install -U openmim
mim install mmcv-full

Then it's ready