fixed the csv results issue

app.py

@@ -9,7 +9,6 @@ import numpy as np
 from functools import lru_cache
 
 # Cache the model loading
-# @gr.cache
 @lru_cache(maxsize=1)
 def load_model():
     model_path = "MMADS/MoralFoundationsClassifier"
@@ -144,34 +143,48 @@ def process_text(text):
     
     return scores_text, bar_chart
 
-def process_csv(file):
+def process_csv(file, progress=gr.Progress()):
     """Process CSV file with multiple texts"""
     if file is None:
-        return "Please upload a CSV file", None, None
+        return "Please upload a CSV file", None, None, None
     
     try:
         # Read CSV
         df = pd.read_csv(file.name)
         
         if 'text' not in df.columns:
-            return "Error: CSV must contain a 'text' column", None, None
+            return "Error: CSV must contain a 'text' column", None, None, None
         
         texts = df['text'].tolist()
         
         # Load model and process in batches
+        progress(0, desc="Loading model...")
         model, tokenizer, label_names = load_model()
         
         # Process in batches of 32
         batch_size = 32
         all_results = []
+        total_batches = (len(texts) + batch_size - 1) // batch_size
         
         for i in range(0, len(texts), batch_size):
+            batch_num = i // batch_size + 1
+            progress(batch_num / total_batches, desc=f"Processing batch {batch_num}/{total_batches}")
+            
             batch_texts = texts[i:i+batch_size]
             batch_results = predict_batch(batch_texts, model, tokenizer, label_names)
             all_results.extend(batch_results)
         
+        progress(0.9, desc="Creating visualizations...")
+        
         # Create summary
         summary = f"**Processed {len(texts)} texts**\n\n"
+        summary += "**Average Scores Across All Texts:**\n\n"
+        
+        # Calculate average scores
+        avg_scores = {}
+        for label in label_names:
+            avg_scores[label] = np.mean([r['scores'][label] for r in all_results])
+            summary += f"{label.replace('_', ' ').title()}: {avg_scores[label]:.4f}\n"
         
         # Create visualizations
         bar_chart = create_visualization(all_results)
@@ -185,12 +198,14 @@ def process_csv(file):
             } for r in all_results
         ])
         
-        results_df.to_csv('results.csv', index=False)
+        # Save to a temporary file and return the path
+        output_path = "results.csv"
+        results_df.to_csv(output_path, index=False)
         
-        return summary + "Results saved to results.csv", bar_chart, heatmap
+        return summary, bar_chart, heatmap, output_path
         
     except Exception as e:
-        return f"Error processing CSV: {str(e)}", None, None
+        return f"Error processing CSV: {str(e)}", None, None, None
 
 # Create example texts
 example_texts = [
@@ -244,6 +259,8 @@ with gr.Blocks(title="Moral Foundations Classifier") as demo:
         gr.Markdown("""
         Upload a CSV file with a 'text' column containing the texts to analyze.
         The app will process all texts and provide aggregate visualizations.
+                    
+        A sample CSV file is available for download [here](https://huggingface.co/spaces/MMADS/MoralFoundationsClassifier-app/tree/main/examples            
         """)
         
         csv_input = gr.File(
@@ -257,12 +274,15 @@ with gr.Blocks(title="Moral Foundations Classifier") as demo:
         
         with gr.Row():
             bar_output = gr.Plot(label="Average Scores")
-            heatmap_output = gr.Plot(label="Scores Heatmap")
+            heatmap_output = gr.Plot(label="Scores Heatmap (First 20 texts)")
+        
+        # Add download component
+        download_output = gr.File(label="Download Results", visible=True)
         
         process_btn.click(
             fn=process_csv,
             inputs=csv_input,
-            outputs=[summary_output, bar_output, heatmap_output]
+            outputs=[summary_output, bar_output, heatmap_output, download_output]
         )
     
     gr.Markdown("""