Update app.py

app.py

@@ -65,35 +65,23 @@ st.markdown("""
 
 @st.cache_resource
 def load_movie_data():
-    # Option 1: Try loading with trust_remote_code
     try:
-        dataset = load_dataset(
-            "eraser-benchmark/movie_rationales", 
-            split="train[:2000]", 
-            trust_remote_code=True  # Explicitly allow trusted code
-        )
-        print(dataset.column_names)
+        st.write("🔄 Attempting to load dataset...")
+
+        dataset = load_dataset("eraser-benchmark/movie_rationales", split="train[:5]", trust_remote_code=True)
         df = pd.DataFrame(dataset)
-        df['context'] = "Title: " + df['movie'] + "\n" + \
-                       "Plot: " + df['plot'] + "\n" + \
-                       "Year: " + df['year'].astype(str)
+
+        st.write("✅ Columns in dataset:")
+        st.write(df.columns.tolist())
+        st.write("📄 Sample row:")
+        st.write(df.iloc[0])
+
         return df
-        
+
     except Exception as e:
-        # Option 2: Fallback to synthetic data
-        st.warning("Using high-quality synthetic movie data")
-        st.error(f"Dataset loading failed: {str(e)}")
-        return pd.DataFrame([
-            {
-                "context": "Title: The Dark Knight\nPlot: Batman faces the Joker...\nYear: 2008\nCast: Christian Bale, Heath Ledger\nDirector: Christopher Nolan"
-            },
-            {
-                "context": "Title: Inception\nPlot: A thief who enters dreams...\nYear: 2010\nCast: Leonardo DiCaprio\nDirector: Christopher Nolan"
-            },
-            {
-                "context": "Title: Pulp Fiction\nPlot: Interconnected stories of criminals...\nYear: 1994\nCast: John Travolta\nDirector: Quentin Tarantino"
-            }
-        ])
+        st.error(f"❌ Dataset loading failed: {str(e)}")
+        return pd.DataFrame()
+
 
 
 @st.cache_resource