app.py

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

# Загрузка модели и токенизатора
model_name = "mistralai/Mistral-7B-Instruct-v0.3"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype=torch.float16).to("cuda")

# Функция для поиска
def search(query):
    inputs = tokenizer(query, return_tensors="pt").to("cuda")
    outputs = model.generate(inputs.input_ids, max_length=256)
    result = tokenizer.decode(outputs[0], skip_special_tokens=True)
    return result

# Настройка интерфейса Gradio
iface = gr.Interface(fn=search, inputs="text", outputs="text", title="AI Search Engine")

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