app.py

import os
from threading import Thread
from typing import Iterator
import gradio as gr
import spaces
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer, TextIteratorStreamer

DESCRIPTION = """\
# Qwen2 0.5B Instruct Text Completion

This is a demo of [`Qwen/Qwen2-0.5B-Instruct`](https://huggingface.co/Qwen/Qwen2-0.5B-Instruct), fine-tuned for instruction following.

Enter your text in the box below and click "Complete" to have the AI generate a completion for your input. The generated text will be appended to your input. You can stop the generation at any time by clicking the "Stop" button.
"""

MAX_MAX_NEW_TOKENS = 2048
DEFAULT_MAX_NEW_TOKENS = 1024
MAX_INPUT_TOKEN_LENGTH = int(os.getenv("MAX_INPUT_TOKEN_LENGTH", "4096"))

device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")

model_id = "Qwen/Qwen2-0.5B-Instruct"
tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(
    model_id,
    device_map="auto",
    torch_dtype=torch.bfloat16,
)
model.eval()

@spaces.GPU(duration=90)
def generate(
    message: str,
    max_new_tokens: int = 1024,
    temperature: float = 0.6,
    top_p: float = 0.9,
    top_k: int = 50,
    repetition_penalty: float = 1.2,
) -> Iterator[str]:
    input_ids = tokenizer.encode(message, return_tensors="pt")
    if input_ids.shape[1] > MAX_INPUT_TOKEN_LENGTH:
        input_ids = input_ids[:, -MAX_INPUT_TOKEN_LENGTH:]
        gr.Warning(f"Trimmed input as it was longer than {MAX_INPUT_TOKEN_LENGTH} tokens.")
    input_ids = input_ids.to(model.device)

    streamer = TextIteratorStreamer(tokenizer, timeout=20.0, skip_prompt=True, skip_special_tokens=True)
    generate_kwargs = dict(
        input_ids=input_ids,
        streamer=streamer,
        max_new_tokens=max_new_tokens,
        do_sample=True,
        top_p=top_p,
        top_k=top_k,
        temperature=temperature,
        num_beams=1,
        repetition_penalty=repetition_penalty,
    )
    t = Thread(target=model.generate, kwargs=generate_kwargs)
    t.start()

    partial_message = message
    for text in streamer:
        partial_message += text
        yield partial_message

with gr.Blocks(css="style.css", fill_height=True) as demo:
    gr.Markdown(DESCRIPTION)
    gr.DuplicateButton(value="Duplicate Space for private use", elem_id="duplicate-button")

    with gr.Row():
        with gr.Column(scale=4):
            text_box = gr.Textbox(
                label="Enter your text",
                placeholder="Type your message here...",
                lines=5
            )
        with gr.Column(scale=1):
            complete_button = gr.Button("Complete")
            stop_button = gr.Button("Stop")

    max_new_tokens = gr.Slider(
        label="Max new tokens",
        minimum=1,
        maximum=MAX_MAX_NEW_TOKENS,
        step=1,
        value=DEFAULT_MAX_NEW_TOKENS,
    )
    temperature = gr.Slider(
        label="Temperature",
        minimum=0.1,
        maximum=4.0,
        step=0.1,
        value=0.6,
    )
    top_p = gr.Slider(
        label="Top-p (nucleus sampling)",
        minimum=0.05,
        maximum=1.0,
        step=0.05,
        value=0.9,
    )
    top_k = gr.Slider(
        label="Top-k",
        minimum=1,
        maximum=1000,
        step=1,
        value=50,
    )
    repetition_penalty = gr.Slider(
        label="Repetition penalty",
        minimum=1.0,
        maximum=2.0,
        step=0.05,
        value=1.2,
    )

    complete_button.click(
        generate,
        inputs=[text_box, max_new_tokens, temperature, top_p, top_k, repetition_penalty],
        outputs=[text_box],
    )
    stop_button.click(
        None,
        None,
        None,
        cancels=[complete_button.click]
    )

if __name__ == "__main__":
    demo.queue(max_size=20).launch()