app.py

import gradio as gr
import pandas as pd
import numpy as np
from datasets import load_dataset
from sklearn.ensemble import RandomForestRegressor
from sklearn.metrics import r2_score
import matplotlib.pyplot as plt

# Загрузка датасета
ds = load_dataset("QSBench/QSBench-Core-v1.0.0-demo")

# Функция для отображения данных выбранного сплита
def show_data(split):
    df = pd.DataFrame(ds[split])
    return df.head(10)

# Функция для обучения модели
def train_and_plot():
    feature_cols = ["total_gates", "gate_entropy", "meyer_wallach"]
    target_col = "ideal_expval_Z_global"
    X_train = pd.DataFrame(ds["train"])[feature_cols]
    y_train = pd.DataFrame(ds["train"])[target_col]
    X_test = pd.DataFrame(ds["test"])[feature_cols]
    y_test = pd.DataFrame(ds["test"])[target_col]

    model = RandomForestRegressor(n_estimators=100, random_state=42)
    model.fit(X_train, y_train)
    y_pred = model.predict(X_test)
    r2 = r2_score(y_test, y_pred)

    fig, ax = plt.subplots()
    ax.scatter(y_test, y_pred, alpha=0.5)
    ax.plot([y_test.min(), y_test.max()], [y_test.min(), y_test.max()], 'r--')
    ax.set_xlabel("True value")
    ax.set_ylabel("Predicted")
    ax.set_title(f"Predictions vs. Truth (R² = {r2:.4f})")
    return fig

with gr.Blocks(title="QSBench Demo Explorer") as demo:
    gr.Markdown("""
    # QSBench Core Demo Explorer
    Interactive demo of the **QSBench Core Demo** dataset – 200 synthetic quantum circuits (6 qubits, depth 4).
    This space shows how to load the data, inspect it, and train a simple model on the ideal expectation values.

    👉 **Full datasets (up to 200k samples, noisy versions, 10‑qubit transpilation packs) are available for purchase.**  
    [Visit the QSBench website](https://qsbench.github.io/)
    """)
    with gr.Tabs():
        with gr.TabItem("Data Explorer"):
            split_selector = gr.Dropdown(choices=["train", "validation", "test"], label="Choose a split", value="train")
            data_table = gr.Dataframe()
            split_selector.change(fn=show_data, inputs=split_selector, outputs=data_table)
        with gr.TabItem("Model Demo"):
            train_button = gr.Button("Train Random Forest")
            plot_output = gr.Plot()
            train_button.click(fn=train_and_plot, outputs=plot_output)
    gr.Markdown("---")
    gr.Markdown("### Get the full datasets\n- **QSBench Core** – 75k clean circuits (8 qubits)\n- **Depolarizing Noise Pack** – 150k circuits with depolarizing noise\n- **Amplitude Damping Pack** – 150k circuits with T1‑like relaxation\n- **Transpilation Hardware Pack** – 200k circuits (10 qubits) with hardware‑aware transpilation\n\n🔗 [Browse all datasets and purchase licenses](https://qsbench.github.io/)")

demo.launch()