from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
# Load the base model first
base_model = AutoModelForCausalLM.from_pretrained("data4elm/Llama-400M-12L")
# Load the DoRA adapter
model = PeftModel.from_pretrained(base_model, "leonzc/llama400m-wikitext2-dora-adapter")
# Load the tokenizer from the base model
tokenizer = AutoTokenizer.from_pretrained("data4elm/Llama-400M-12L")
# Example usage
input_text = "What is the capital of France?"
inputs = tokenizer(input_text, return_tensors="pt")
outputs = model.generate(inputs.input_ids, max_length=50)
print(tokenizer.decode(outputs[0], skip_special_tokens=True))