app.py

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

MODEL = "THUDM/LongWriter-glm4-9b"

TITLE = "<h1><center>LongWriter-glm4-9b</center></h1>"

PLACEHOLDER = """
<center>
<p>Hi! I'm LongWriter-glm4-9b, capable of generating 10,000+ words. How can I assist you today?</p>
</center>
"""

CSS = """
.duplicate-button {
    margin: auto !important;
    color: white !important;
    background: black !important;
    border-radius: 100vh !important;
}
h3 {
    text-align: center;
}
"""

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

tokenizer = AutoTokenizer.from_pretrained(MODEL, trust_remote_code=True)
model = AutoModelForCausalLM.from_pretrained(MODEL, torch_dtype=torch.bfloat16, trust_remote_code=True, device_map="auto")
model = model.eval()

@spaces.GPU
def stream_chat(
    message: str, 
    history: list,
    system_prompt: str,
    temperature: float = 0.5, 
    max_new_tokens: int = 32768, 
    top_p: float = 1.0, 
    top_k: int = 50,
):
    print(f'message: {message}')
    print(f'history: {history}')

    chat_history = []
    for prompt, answer in history:
        chat_history.append((prompt, answer))

    response, _ = model.chat(
        tokenizer,
        message,
        history=chat_history,
        max_new_tokens=max_new_tokens,
        top_p=top_p,
        top_k=top_k,
        temperature=temperature,
    )
    
    yield response


chatbot = gr.Chatbot(height=600, placeholder=PLACEHOLDER)

with gr.Blocks(css=CSS, theme="soft") as demo:
    gr.HTML(TITLE)
    gr.DuplicateButton(value="Duplicate Space for private use", elem_classes="duplicate-button")
    gr.ChatInterface(
        fn=stream_chat,
        chatbot=chatbot,
        fill_height=True,
        additional_inputs_accordion=gr.Accordion(label="⚙️ Parameters", open=False, render=False),
        additional_inputs=[
            gr.Textbox(
                value="You are a helpful assistant capable of generating long-form content.",
                label="System Prompt",
                render=False,
            ),
            gr.Slider(
                minimum=0,
                maximum=1,
                step=0.1,
                value=0.5,
                label="Temperature",
                render=False,
            ),
            gr.Slider(
                minimum=1024,
                maximum=32768,
                step=1024,
                value=32768,
                label="Max new tokens",
                render=False,
            ),
            gr.Slider(
                minimum=0.0,
                maximum=1.0,
                step=0.1,
                value=1.0,
                label="Top p",
                render=False,
            ),
            gr.Slider(
                minimum=1,
                maximum=100,
                step=1,
                value=50,
                label="Top k",
                render=False,
            ),
        ],
        examples=[
            ["Write a 10000-word comprehensive guide on artificial intelligence and its applications."],
            ["Create a detailed 5000-word business plan for a space tourism company."],
            ["Compose a 3000-word short story about time travel and its consequences."],
            ["Develop a 7000-word research proposal on the potential of quantum computing in cryptography."],
        ],
        cache_examples=False,
    )

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