app.py

import os
import subprocess

# Download Kronos model code at startup
if not os.path.exists("model"):
    subprocess.run([
        "git", "clone", "--depth=1",
        "https://github.com/shiyu-coder/Kronos",
        "kronos_tmp"
    ])
    os.rename("kronos_tmp/model", "model")
    subprocess.run(["rm", "-rf", "kronos_tmp"])

import spaces
import gradio as gr
import pandas as pd
import plotly.graph_objects as go
from model import Kronos, KronosTokenizer, KronosPredictor

tokenizer = KronosTokenizer.from_pretrained("NeoQuasar/Kronos-Tokenizer-base")
model = Kronos.from_pretrained("NeoQuasar/Kronos-base")
predictor = KronosPredictor(model, tokenizer, device="cuda", max_context=512)

@spaces.GPU
def forecast(file, pred_len, temperature, top_p):
    df = pd.read_csv(file.name)
    df['timestamps'] = pd.to_datetime(df['timestamps'])

    lookback = min(400, len(df) - int(pred_len))
    x_df = df.loc[:lookback-1, ['open', 'high', 'low', 'close']]
    x_ts = df.loc[:lookback-1, 'timestamps']
    y_ts = df.loc[lookback:lookback+int(pred_len)-1, 'timestamps']

    pred = predictor.predict(
        df=x_df,
        x_timestamp=x_ts,
        y_timestamp=y_ts,
        pred_len=int(pred_len),
        T=float(temperature),
        top_p=float(top_p),
        sample_count=1
    )

    fig = go.Figure()
    fig.add_trace(go.Scatter(
        x=x_ts, y=x_df['close'],
        name="Historical", line=dict(color='steelblue')
    ))
    fig.add_trace(go.Scatter(
        x=y_ts, y=pred['close'],
        name="Forecast", line=dict(color='orange', dash='dot')
    ))
    fig.update_layout(
        title="Kronos Forecast",
        xaxis_title="Time",
        yaxis_title="Price",
        template="plotly_dark"
    )
    return fig

with gr.Blocks(theme=gr.themes.Soft()) as demo:
    gr.Markdown("# 📈 Kronos Financial Forecaster")
    gr.Markdown("Upload a CSV with columns: `timestamps, open, high, low, close`")
    
    with gr.Row():
        file = gr.File(label="📂 Upload CSV")
    with gr.Row():
        pred_len = gr.Slider(10, 120, value=30, step=1, label="Forecast Steps")
        temp = gr.Slider(0.1, 2.0, value=1.0, step=0.1, label="Temperature")
        top_p = gr.Slider(0.1, 1.0, value=0.9, step=0.05, label="Top-p")

    btn = gr.Button("🔮 Run Forecast", variant="primary")
    plot = gr.Plot(label="Forecast Output")

    btn.click(
        fn=forecast,
        inputs=[file, pred_len, temp, top_p],
        outputs=plot
    )

demo.launch()