app.py

import transformers
transformers.__version__

import torch
torch.cuda.is_available()

from transformers import pipeline

generator = pipeline("text-generation")

generator = pipeline("text-generation", model = "Qwen/Qwen3-4B-Instruct-2507",
                     torch_dtype= "auto", device_map="auto")

messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "Tell me a short joke"}
]

outputs = generator(
    messages,
    max_new_tokens = 100,
    do_sample=True,
    temperature = 0.7,
    return_full_text = False)

print(outputs[0]["generated_text"])

messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": "Türkiyenin başkenti neresidir?"}
]

outputs = generator(
    messages,
    max_new_tokens = 100,
    do_sample=True,
    temperature = 0.7,
    return_full_text = False)

print(outputs[0]["generated_text"])

import gradio as gr

def generate_text(prompt):
  messages = [
    {"role": "system", "content": "You are a helpful AI assistant."},
    {"role": "user", "content": prompt}]
  outputs = generator(
    messages,
    max_new_tokens = 100,
    do_sample=True,
    temperature = 0.7,
    return_full_text = False)
  return outputs[0]["generated_text"]

demo = gr.Interface(
    fn= generate_text,
    inputs = gr.Textbox(label="Give an input"),
    outputs = gr.Textbox(label="Output"),
    title = "Text Generation"
)

demo.launch()