jepa
jepa copied to clipboard
facebookresearch
PyTorch code and models for V-JEPA self-supervised learning from video.
The following 2 links don't work: ```bash curl --head https://dl.fbaipublicfiles.com/jepa/vitl16/in1k-probe.pth.tar curl --head https://dl.fbaipublicfiles.com/jepa/vith16/in1k-probe.pth.tar ```
Hey, I was wondering what the number of optimization steps (or epochs) and batch sizes are used for the Low-Shot Frozen Evaluation experiment (Table 7 in the V-JEPA paper). Is...
Note that [`pip install -e .` is preferred over `python setup.py install`](https://stackoverflow.com/questions/15724093/difference-between-python-setup-py-install-and-pip-install).
Shouldn't the frozen evaluation not use augmentations? After looking at the code and reading the paper, I see that you apply random augmentations when computing the embeddings. The weights of...
Hi, I can see there are various encoder class configs including a ViT-mini. Will you be releasing one in the future?
Metadata
Owner
Metadata
PyTorch code and models for V-JEPA self-supervised learning from video.