How to use wjmcat/opt-350m-paddle with paddlenlp:
from paddlenlp.transformers import AutoTokenizer, OPTForCausalLM tokenizer = AutoTokenizer.from_pretrained("wjmcat/opt-350m-paddle", from_hf_hub=True) model = OPTForCausalLM.from_pretrained("wjmcat/opt-350m-paddle", from_hf_hub=True)
