app.py

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

# Load model
MODEL_NAME = "mistralai/Mistral-7B-Instruct-v0.3"
tokenizer = AutoTokenizer.from_pretrained(MODEL_NAME)
model = AutoModelForCausalLM.from_pretrained(MODEL_NAME, device_map="auto", torch_dtype=torch.float16)

def generate(prompt):
    inputs = tokenizer(prompt, return_tensors="pt").to(model.device)
    with torch.no_grad():
        outputs = model.generate(**inputs, max_new_tokens=300)
    text = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return text

# Gradio interface
iface = gr.Interface(
    fn=generate,
    inputs=gr.Textbox(lines=5, placeholder="Enter your prompt here"),
    outputs=gr.Textbox(label="AI Response"),
)

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