Transformers How to use fusing/gpt2_optimus with Transformers:
# Load model directly
from transformers import AutoTokenizer, GPT2OptimusForLatentConnector
tokenizer = AutoTokenizer.from_pretrained("fusing/gpt2_optimus")
model = GPT2OptimusForLatentConnector.from_pretrained("fusing/gpt2_optimus") 
