app.py

import gradio as gr
from transformers import pipeline, AutoModelForCausalLM

model = AutoModelForCausalLM.from_pretrained(
      'vitaliy-sharandin/wiseai',
      load_in_8bit=True,
      device_map = {"": 0}
  )

pipe = pipeline('text-generation', model=model, do_sample=True, temperature=0.7)

def generate_text(prompt):
    result = pipe(prompt, max_length=200)
    return result[0]['generated_text']

iface = gr.Interface(fn=generate_text, inputs="text", outputs="text")
iface.launch()