Upload folder using huggingface_hub

hate_speech_demo.py

@@ -66,17 +66,18 @@ def process_retrieval_text(retrieval_text, user_input):
                 content_match = re.search(r'Content:\s*(.*)', doc_info, re.DOTALL)
                 content = content_match.group(1).strip() if content_match else "No content available"
                 
-                # Format with clear section headers
+                # Format with clear section headers and better spacing
                 formatted_html = f"""
                 <div class='doc-section'>
-                    <strong>Evidence Document {i+1}</strong>
-                    <br>
-                    <strong>Document Title:</strong> {doc_name}
-                    <br>
-                    <strong>Section:</strong> Page {page}
-                    <br>
-                    <strong>Content:</strong>
-                    <div class='doc-content'>{content}</div>
+                    <h3 class="doc-number">Evidence Document {i+1}</h3>
+                    <div class="doc-info">
+                        <p><strong>Document Title:</strong> {doc_name}</p>
+                        <p><strong>Section:</strong> Page {page}</p>
+                    </div>
+                    <div class="doc-content-container">
+                        <h4>Content:</h4>
+                        <div class='doc-content'>{content}</div>
+                    </div>
                 </div>
                 """
                 chunks.append(formatted_html)
@@ -107,8 +108,12 @@ def process_retrieval_text(retrieval_text, user_input):
             chunks.append(current_chunk)
         
         # Format each chunk with better section styling
-        chunks = [f"<div class='doc-section'><div class='doc-title'>Evidence Document {i+1}</div><div class='doc-content'>{chunk.strip()}</div></div>" 
-                  for i, chunk in enumerate(chunks)]
+        chunks = [f"""
+        <div class='doc-section'>
+            <h3 class="doc-title">Evidence Document {i+1}</h3>
+            <div class='doc-content'>{chunk.strip()}</div>
+        </div>
+        """ for i, chunk in enumerate(chunks)]
     
     # Extract keywords from user input (longer than 3 chars)
     keywords = re.findall(r'\b\w{4,}\b', user_input.lower())
@@ -127,7 +132,72 @@ def process_retrieval_text(retrieval_text, user_input):
         
         highlighted_chunks.append(highlighted_chunk)
     
-    return "<div class='knowledge-sections'>" + "".join(highlighted_chunks) + "</div>"
+    # Add some additional CSS for the knowledge sections
+    additional_css = """
+    <style>
+    .knowledge-sections {
+        border-radius: 8px;
+        background: #f9f9f9;
+        padding: 15px;
+        font-family: 'All Round Gothic Demi', 'Poppins', sans-serif !important;
+    }
+    
+    .doc-section {
+        margin-bottom: 25px;
+        padding: 15px;
+        background: white;
+        border-radius: 8px;
+        box-shadow: 0 2px 5px rgba(0,0,0,0.05);
+    }
+    
+    .doc-number, .doc-title {
+        margin-top: 0;
+        padding-bottom: 10px;
+        border-bottom: 1px solid #eee;
+        color: #222;
+        font-size: 18px;
+    }
+    
+    .doc-info {
+        margin: 10px 0;
+        padding: 8px;
+        background: #f5f5f5;
+        border-radius: 4px;
+    }
+    
+    .doc-info p {
+        margin: 5px 0;
+    }
+    
+    .doc-content-container {
+        margin-top: 15px;
+    }
+    
+    .doc-content-container h4 {
+        margin-bottom: 8px;
+        font-size: 16px;
+    }
+    
+    .doc-content {
+        padding: 12px;
+        background: #f9f9f9;
+        border-left: 3px solid #FCA539;
+        line-height: 1.6;
+        border-radius: 4px;
+        white-space: pre-line;
+    }
+    
+    .highlight-match {
+        background-color: #FCA539;
+        color: black;
+        font-weight: bold;
+        padding: 0 2px;
+        border-radius: 2px;
+    }
+    </style>
+    """
+    
+    return additional_css + "<div class='knowledge-sections'>" + "".join(highlighted_chunks) + "</div>"
 
 # API Keys - hardcoded for convenience
 # Replace these with your actual API keys