app.py

import os
import gradio as gr
import tensorflow as tf
import joblib
import numpy as np

# -------------------------------------------------------
# CONFIG
# -------------------------------------------------------
MODEL_REPO = "munem420/stock-forecaster-lstm"
MODEL_FILENAME = "model_lstm.h5"
SCALER_FILENAME = "scalers.joblib"

os.environ["CUDA_VISIBLE_DEVICES"] = "-1"

# -------------------------------------------------------
# LOAD MODEL AND SCALERS
# -------------------------------------------------------
print("📦 Loading model and scalers...")

try:
    model_path = tf.keras.utils.get_file(
        MODEL_FILENAME,
        f"https://huggingface.co/{MODEL_REPO}/resolve/main/{MODEL_FILENAME}"
    )
    scalers_path = tf.keras.utils.get_file(
        SCALER_FILENAME,
        f"https://huggingface.co/{MODEL_REPO}/resolve/main/{SCALER_FILENAME}"
    )

    model = tf.keras.models.load_model(
        model_path,
        custom_objects={"mse": tf.keras.losses.MeanSquaredError()}
    )
    scalers = joblib.load(scalers_path)

    print("✅ Model and scalers loaded successfully.")
except Exception as e:
    print(f"❌ Error loading model or scalers: {e}")
    model, scalers = None, None

# -------------------------------------------------------
# FORECAST FUNCTION
# -------------------------------------------------------
def forecast_stock(ticker):
    if not model or not scalers:
        return "❌ Model or scalers not loaded properly."

    ticker = ticker.strip().upper()
    if ticker not in scalers:
        return f"⚠️ No scaler found for ticker '{ticker}'. Please check spelling."

    # Dummy inference example (replace with actual data fetching or preprocessing)
    # Here we just simulate 60 normalized close prices for inference
    scaler = scalers[ticker]
    dummy_data = np.linspace(0.9, 1.1, 60).reshape(-1, 1)
    X_pred = dummy_data.reshape(1, 60, 1)

    # Predict scaled value
    pred_scaled = model.predict(X_pred, verbose=0)[0][0]

    # Inverse transform prediction
    pred_actual = scaler.inverse_transform(np.array([[pred_scaled]]))[0][0]

    return f"🔮 Predicted next day close for **{ticker}**: ${pred_actual:.2f}"

# -------------------------------------------------------
# GRADIO INTERFACE
# -------------------------------------------------------
iface = gr.Interface(
    fn=forecast_stock,
    inputs=gr.Textbox(label="Enter Ticker or Company Name"),
    outputs=gr.Markdown(label="Prediction Result"),
    title="📊 Stock Price Forecaster (LSTM)",
    description="Enter a stock ticker or company name to predict the next day's closing price using a trained LSTM model."
)

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