app.py

import gradio as gr
from transformers import AutoTokenizer, AutoModelForCausalLM
import torch

# Load Falcon-7B-Instruct model with 4-bit quantization
model_id = "tiiuae/falcon-7b-instruct"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    torch_dtype=torch.float16,
    device_map="auto",
    load_in_4bit=True,
    trust_remote_code=True
)

# Function to generate response
def generate_response(prompt):
    inputs = tokenizer(prompt, return_tensors="pt").to("cuda" if torch.cuda.is_available() else "cpu")
    outputs = model.generate(
        **inputs,
        max_length=100,
        do_sample=True,
        top_k=50,
        top_p=0.95,
        num_return_sequences=1
    )
    return tokenizer.decode(outputs[0], skip_special_tokens=True)

# Gradio interface
iface = gr.Interface(
    fn=generate_response,
    inputs=gr.Textbox(label="Enter your prompt", placeholder="Type here..."),
    outputs=gr.Textbox(label="Zephyrix Response"),
    title="Zephyrix - Pakistan's First AI Model",
    description="Built with love in Pakistan!"
)

# Launch app
iface.launch(server_name="0.0.0.0", server_port=7860)