How to use yuchiz/cache with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("yuchiz/cache", dtype="auto")
