app.py

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

base_model_name = "meta-llama/Llama-2-7b-chat-hf"
adapter_name = "mostafa33/llama2-ecommerce-lora2"

bnb_config = BitsAndBytesConfig(
    load_in_4bit=True,
    bnb_4bit_compute_dtype=torch.float16,
    bnb_4bit_use_double_quant=True,
    bnb_4bit_quant_type="nf4",
)

# Load model
base_model = AutoModelForCausalLM.from_pretrained(
    base_model_name,
    quantization_config=bnb_config,
    device_map="auto"
)
model = PeftModel.from_pretrained(base_model, adapter_name)

tokenizer = AutoTokenizer.from_pretrained(base_model_name)
tokenizer.pad_token = tokenizer.eos_token

def generate_response(prompt):
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
    outputs = model.generate(
        **inputs,
        max_new_tokens=256,
        temperature=0.7,
        top_p=0.9,
        do_sample=True,
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

# Gradio Interface
demo = gr.Interface(
    fn=generate_response,
    inputs=gr.Textbox(lines=4, label="Enter your product or query"),
    outputs=gr.Textbox(label="Model Response"),
    title="🦙 LLaMA2 LoRA Ecommerce Chatbot",
    description="Fine-tuned LoRA model for ecommerce text generation and chat.",
)

demo.launch()