How to use Erland/gpt2-tokenizer with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("Erland/gpt2-tokenizer", device_map="auto")
