app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM
from peft import PeftModel

# Load base GPT-2 + your LoRA adapter from Hugging Face Hub
base_model = AutoModelForCausalLM.from_pretrained("gpt2")
tokenizer = AutoTokenizer.from_pretrained("n-sudheer/ns-lora-gpt2-demo")
model = PeftModel.from_pretrained(base_model, "n-sudheer/ns-lora-gpt2-demo")

def generate_text(prompt, max_length=50):
    inputs = tokenizer(prompt, return_tensors="pt")
    outputs = model.generate(
        **inputs,
        max_length=max_length,
        do_sample=True,
        top_k=50
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

demo = gr.Interface(
    fn=generate_text,
    inputs=[
        gr.Textbox(label="Prompt"),
        gr.Slider(10, 200, value=50, step=5, label="Max length")
    ],
    outputs=gr.Textbox(label="Generated Text"),
    title="LoRA GPT-2 Demo",
    description="A GPT-2 model fine-tuned with LoRA, deployed on Hugging Face Spaces."
)

if __name__ == "__main__":
    demo.launch()