How to use Jeonghwa/mdeberta-mcqa-lora with PEFT:
from peft import PeftModel from transformers import AutoModelForSequenceClassification base_model = AutoModelForSequenceClassification.from_pretrained("microsoft/mdeberta-v3-base") model = PeftModel.from_pretrained(base_model, "Jeonghwa/mdeberta-mcqa-lora")
Jeonghwa commited on 