How to use jinmang2/dall-e-tokenizer with Transformers:
# Load model directly from transformers import DallEEncoder model = DallEEncoder.from_pretrained("jinmang2/dall-e-tokenizer", dtype="auto")
