Update src/streamlit_app.py

src/streamlit_app.py

@@ -3,7 +3,6 @@ import pandas as pd
 import joblib
 from tensorflow import keras
 from datetime import date
-from huggingface_hub import hf_hub_download
 import os
 
 from utils import forecast_to_annual_return, plot_forecast_vs_actual
@@ -20,17 +19,11 @@ from forecast import forecast_lstm
 tickers = ["TSLA", "BND", "SPY"]
 sequence_length = 60
 start_date_data = "2015-01-01" 
-hf_repo_id = "abnsol/tsla_price_lstm_model"
-model_filename = "tsla_price_lstm_model.keras"
-scaler_filename = "scaler.joblib"
 
 APP_DIR = os.path.dirname(os.path.abspath(__file__))
 PROJECT_ROOT = os.path.dirname(APP_DIR)
-
-proc_dir = os.path.join(PROJECT_ROOT,"src","data", "processed")
-os.environ["HF_HOME"] = "./.hf_cache"
-os.environ["HF_HUB_CACHE"] = "./.hf_cache"
-os.environ["TRANSFORMERS_CACHE"] = "./.hf_cache"
+proc_dir = os.path.join(PROJECT_ROOT, "src", "data", "processed")
+models_dir = os.path.join(PROJECT_ROOT, "models")
 
 # --- App Layout ---
 st.set_page_config(layout="wide")
@@ -47,17 +40,12 @@ if st.sidebar.button("Run Optimization"):
         end_date_data = date.today().strftime("%Y-%m-%d")
         data = load_processed_prices(proc_dir, tickers)
 
-    # --- 2. Load trained LSTM model & scaler from Hugging Face Hub ---
-    with st.spinner("Loading forecasting model from Hugging Face Hub..."):
-        # Load the Keras model directly using the hf:// protocol
-        # model = keras.models.load_model(f"hf://{hf_repo_id}/{model_filename}")
-        # Available backend options are: "jax", "torch", "tensorflow".
-        os.environ["KERAS_BACKEND"] = "tensorflow"        
-        model = keras.models.load_model("hf://abnsol/tsla_price_lstm_model")
-
-        
-        # Download the scaler file from the Hub and then load it with joblib
-        scaler_path = hf_hub_download(repo_id=hf_repo_id, filename=scaler_filename)
+    # --- 2. Load trained LSTM model & scaler from local models folder ---
+    with st.spinner("Loading forecasting model..."):
+        model_path = os.path.join(models_dir, "lstm_model.keras")
+        scaler_path = os.path.join(models_dir, "scaler.joblib")
+
+        model = keras.models.load_model(model_path)
         scaler = joblib.load(scaler_path)
 
     # --- 3. Prepare TSLA data & forecast ---
@@ -111,7 +99,6 @@ if st.sidebar.button("Run Optimization"):
             allocation_df = (weights_df * investment_capital).map('${:,.2f}'.format)
             st.dataframe(allocation_df)
 
-
     # --- 8. Efficient Frontier Plot ---
     st.subheader("Efficient Frontier")
     fig_frontier = plot_efficient_frontier(mu, cov, general_results)