app.py

import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer, AutoModelForSeq2SeqLM


def greet(input):

    model_name = "Qwen/Qwen3-8B"

    # load the tokenizer and the model
    tokenizer = AutoTokenizer.from_pretrained(model_name)
    model = AutoModelForCausalLM.from_pretrained(model_name)

    tokenizer.save_pretrained("./qwen3")
    model.save_pretrained("./qwen3")

    # prepare the model input
    prompt = "Give me a short introduction to large language model."
    prompt = input
    messages = [{"role": "user", "content": prompt}]
    text = tokenizer.apply_chat_template(
        messages,
        tokenize=False,
        add_generation_prompt=True,
        enable_thinking=True,  # Switches between thinking and non-thinking modes. Default is True.
    )
    model_inputs = tokenizer([text], return_tensors="pt").to(model.device)

    # conduct text completion
    generated_ids = model.generate(**model_inputs, max_new_tokens=32768)
    output_ids = generated_ids[0][len(model_inputs.input_ids[0]) :].tolist()

    # parsing thinking content
    try:
        # rindex finding 151668 (</think>)
        index = len(output_ids) - output_ids[::-1].index(151668)
    except ValueError:
        index = 0

    thinking_content = tokenizer.decode(
        output_ids[:index], skip_special_tokens=True
    ).strip("\n")
    content = tokenizer.decode(output_ids[index:], skip_special_tokens=True).strip("\n")

    # print("thinking content:", thinking_content)
    # print("content:", content)

    return "thinking content:" + thinking_content + "\n" + "content:" + content

demo = gr.Interface(fn=greet, inputs="text", outputs="text")
demo.launch()