app.py

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

model_name = "entfane/math-virtuoso-7B"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForCausalLM.from_pretrained(
    model_name,
    device_map="auto",
)
model.eval()

eos_token = tokenizer.eos_token_id

@spaces.GPU()
def chat_math_virtuoso(message: str, temperature: float = 0.7, max_new_tokens: int = 2048):
    """
    Generate a streaming response with Math Virtuoso 7B.
    """
    conversation = [{"role": "user", "content": message}]
    
    input_ids = tokenizer.apply_chat_template(
        conversation,
        add_generation_prompt=True,
        return_tensors="pt"
    ).to(model.device)

    streamer = TextIteratorStreamer(
        tokenizer,
        timeout=10.0,
        skip_prompt=True,
        skip_special_tokens=True
    )

    generate_kwargs = dict(
        input_ids=input_ids,
        streamer=streamer,
        max_new_tokens=max_new_tokens,
        temperature=temperature,
        eos_token_id=eos_token,
        pad_token_id=eos_token
    )

    thread = Thread(target=model.generate, kwargs=generate_kwargs)
    thread.start()

    outputs = []
    for token in streamer:
        outputs.append(token)
        yield "".join(outputs)

with gr.Blocks(css="""
    body {
        background: linear-gradient(135deg, #f0f2ff, #d9e4ff);
        font-family: 'Helvetica Neue', Arial, sans-serif;
    }
    .gr-block {
        border-radius: 20px;
        box-shadow: 0 10px 30px rgba(0,0,0,0.12);
        padding: 20px;
    }
    .gr-button {
        background: linear-gradient(90deg, #6C63FF, #8B78FF);
        color: white;
        font-weight: bold;
        border-radius: 12px;
        padding: 10px 25px;
        transition: all 0.3s ease;
    }
    .gr-button:hover {
        transform: translateY(-2px);
        box-shadow: 0 6px 15px rgba(0,0,0,0.2);
    }
    .gr-textbox {
        border-radius: 12px;
        border: 1px solid #ccc;
        padding: 10px;
        transition: border-color 0.3s;
    }
    .gr-textbox:focus {
        border-color: #6C63FF;
        box-shadow: 0 0 8px rgba(108, 99, 255, 0.3);
    }
    .gr-slider .noUi-connect {
        background: #6C63FF;
    }
    h2 {
        text-align: center;
        color: #6C63FF;
        font-weight: 700;
        margin-bottom: 25px;
    }
""") as demo:

    gr.Markdown("<h2>🤖 Math Virtuoso 7B Chat</h2>")
    
    with gr.Row():
        with gr.Column(scale=2):
            msg = gr.Textbox(label="Your question", placeholder="Type a math question...", lines=2)
            temperature = gr.Slider(0, 1, value=0.7, step=0.1, label="Temperature")
            max_tokens = gr.Slider(50, 2048, value=512, step=50, label="Max new tokens")
            submit_btn = gr.Button("Ask")
        with gr.Column(scale=3):
            output = gr.Textbox(label="Model reply", lines=15, interactive=False)

    submit_btn.click(
        fn=chat_math_virtuoso,
        inputs=[msg, temperature, max_tokens],
        outputs=output
    )

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