Transformers How to use charmander-33/opt-66b-0.6x-eigenattn with Transformers:
# Load model directly
from transformers import AutoTokenizer, OPTForCausalLM_EigenAttn
tokenizer = AutoTokenizer.from_pretrained("charmander-33/opt-66b-0.6x-eigenattn")
model = OPTForCausalLM_EigenAttn.from_pretrained("charmander-33/opt-66b-0.6x-eigenattn") 
