from transformers import AutoTokenizer, pipeline
from peft import PeftModel, PeftConfig
from transformers import AutoModelForCausalLM

# Load adapter
adapter_id = "imrahulwarkade/tinyllama-toneopbot-lora"
config = PeftConfig.from_pretrained(adapter_id)
base_model = AutoModelForCausalLM.from_pretrained(config.base_model_name_or_path)
model = PeftModel.from_pretrained(base_model, adapter_id)

tokenizer = AutoTokenizer.from_pretrained(config.base_model_name_or_path)
pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)

# Prompt
messages = [
  {"role": "user", "content": "How can I lose weight in a healthy way?"}
]
prompt = tokenizer.apply_chat_template(messages, tokenize=False)
response = pipe(prompt, max_new_tokens=150)[0]["generated_text"]
print(response)