Upload 4 files

.gitattributes

@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
 *.zip filter=lfs diff=lfs merge=lfs -text
 *.zst filter=lfs diff=lfs merge=lfs -text
 *tfevents* filter=lfs diff=lfs merge=lfs -text
+stock_model.keras filter=lfs diff=lfs merge=lfs -text

Requirements.txt

@@ -0,0 +1,7 @@
+tensorflow
+yfinance
+pandas
+numpy
+scikit-learn
+joblib
+gradio

app.py

@@ -0,0 +1,78 @@
+import gradio as gr
+import yfinance as yf
+import numpy as np
+import pandas as pd
+from keras.models import load_model
+import joblib
+import os
+
+# Load the saved components
+# Using a check to ensure the app doesn't crash during build if files are missing
+MODEL_PATH = 'stock_model.keras'
+SCALER_PATH = 'scaler.gz'
+
+if os.path.exists(MODEL_PATH) and os.path.exists(SCALER_PATH):
+    model = load_model(MODEL_PATH)
+    scaler = joblib.load(SCALER_PATH)
+else:
+    model = None
+    scaler = None
+
+def predict_next_day(ticker):
+    if model is None or scaler is None:
+        return "Error: Model or Scaler files not found in the repository."
+    
+    try:
+        # 1. Fetch data
+        # We use multi_level_index=False to fix the common yfinance MultiIndex error
+        df = yf.download(ticker, period='90d', interval='1d', multi_level_index=False)
+        
+        if df.empty: 
+            return f"Error: No data found for ticker '{ticker}'."
+
+        # 2. Re-create technical indicators
+        df['MA7'] = df['Close'].rolling(window=7).mean()
+        df['MA21'] = df['Close'].rolling(window=21).mean()
+        df.dropna(inplace=True)
+
+        # 3. Prepare the sliding window (Last 60 days)
+        # Ensure we have enough rows after dropping NaNs
+        if len(df) < 60:
+            return "Error: Not enough historical data to generate a 60-day window."
+
+        features = df[['Close', 'MA7', 'MA21']].tail(60).values
+        
+        # 4. Scale and Reshape
+        scaled_features = scaler.transform(features)
+        # Reshape to (1, 60, 3) -> 1 sample, 60 timesteps, 3 features
+        input_data = np.reshape(scaled_features, (1, 60, 3))
+
+        # 5. Predict
+        prediction_scaled = model.predict(input_data, verbose=0)
+        
+        # 6. Inverse Transform
+        # Since the scaler was fit on 3 columns, we create a dummy with 3 cols
+        dummy = np.zeros((1, 3))
+        dummy[0, 0] = prediction_scaled[0, 0]
+        prediction_final = scaler.inverse_transform(dummy)[0, 0]
+
+        return f"Predicted Next Closing Price for {ticker}: ₹{prediction_final:.2f}"
+    
+    except Exception as e:
+        return f"Technical Error: {str(e)}"
+
+# Professional Gradio Interface
+interface = gr.Interface(
+    fn=predict_next_day,
+    inputs=gr.Textbox(
+        label="Stock Ticker", 
+        placeholder="Enter Ticker (e.g., TCS.NS for NSE, AAPL for NASDAQ)..."
+    ),
+    outputs=gr.Textbox(label="Forecasted Price"),
+    title="📈 StockPulse: LSTM Market Predictor",
+    description="This AI model uses Long Short-Term Memory (LSTM) networks to predict the next day's closing price based on a 60-day window of historical prices and Moving Averages.",
+    theme="soft"
+)
+
+if __name__ == "__main__":
+    interface.launch()

scaler.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:f703de8e33ce32de0fd864c748977d4096da844352397c81f28064705daccb72
+size 974

stock_model.keras

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5cdd032ee98f5f7ad32221888007b62d6cb4b3a71c64f72dd59e7a77f1bf9bd6
+size 1648465