- Python >= 3.8
- PyTorch >= 1.7, please follow PyTorch official instructions at pytorch.org
- fvcore:
pip install 'git+https://github.com/facebookresearch/fvcore'
- iopath:
pip install -U iopath
orconda install -c iopath iopath
- torchvision that matches the PyTorch installation. You can install them together at pytorch.org to make sure of this.
- simplejson:
pip install simplejson
- FairScale:
pip install 'git+https://github.com/facebookresearch/fairscale'
- psutil:
pip install psutil
Clone the MViT repository.
git clone https://github.com/facebookresearch/mvit
cd mvit
python setup.py build develop
Download the ImageNet-1K classification dataset and structure the data as follows:
/path/to/imagenet-1k/
train/
class1/
img1.jpeg
class2/
img2.jpeg
val/
class1/
img3.jpeg
class2/
img4.jpeg