Update app.py

app.py

@@ -15,32 +15,31 @@ import torch  # For tensor operations
 import soundfile as sf  # For saving audio as .wav files
 import sentencepiece  # Required by SpeechT5Processor for tokenization
 
-
 ##########################################
 # Streamlit application title and input
 ##########################################
-# Display a colorful, large title in a visually appealing font
+# Display a deep blue title in large, visually appealing font
 st.markdown(
-    "<h1 style='text-align: center; color: #FF5720; font-size: 50px;'>Just Comment</h1>",
+    "<h1 style='text-align: center; color: #00008B; font-size: 50px;'>Just Comment</h1>",
     unsafe_allow_html=True
-)  # Use HTML and CSS for a custom title design
+)  # Set a deep blue title
 
-# Display a smaller, gentle subtitle below the title
+# Display a gentle, warm subtitle below the title
 st.markdown(
     "<h3 style='text-align: center; color: #5D6D7E; font-style: italic;'>I'm listening to you, my friend</h3>",
     unsafe_allow_html=True
-)  # Use HTML for a friendly and soft-styled subtitle
+)  # Show a friendly subtitle
 
-# Add a well-designed text area for user input
+# Provide a text area for user input with placeholder and tooltip
 text = st.text_area(
     "Enter your comment",
     placeholder="Type something here...",
     height=150,
-    help="Write a comment you would like us to analyze and respond to!"  # Provide a helpful tooltip
-)
+    help="Write a comment you would like us to respond to!"  # Tooltip for guidance
+)  # Create a text input area
 
 ##########################################
-# Step 1: Sentiment Analysis Function
+# Step 1: Sentiment Analysis Function (Unused here)
 ##########################################
 def analyze_dominant_emotion(user_review):
     """
@@ -50,164 +49,155 @@ def analyze_dominant_emotion(user_review):
         "text-classification",
         model="Thea231/jhartmann_emotion_finetuning",
         return_all_scores=True
-    )  # Load the fine-tuned text classification model from Hugging Face
-
-    emotion_results = emotion_classifier(user_review)[0]  # Perform sentiment analysis on the input text
-    dominant_emotion = max(emotion_results, key=lambda x: x['score'])  # Identify the emotion with the highest confidence
-    return dominant_emotion  # Return the dominant emotion (label and score)
-
+    )  # Load the sentiment classification model
+    emotion_results = emotion_classifier(user_review)[0]  # Get sentiment scores of the input
+    dominant_emotion = max(emotion_results, key=lambda x: x['score'])  # Find the highest scoring emotion
+    return dominant_emotion  # Return the dominant emotion
 
 ##########################################
-# Step 2：Response Generation Function
+# Step 2: Response Generation Functions
 ##########################################
-
 def prompt_gen(user_review):
-  dominant_emotion = analyze_dominant_emotion(user_review)
-  emotion_strategies = {
-      "anger": {
-          "prompt": (
-              "Customer complaint: '{review}'\n\n"
-              "As a customer service representative, craft a professional response that:\n"
-              "- Begins with sincere apology and acknowledgment\n"
-              "- Clearly explains solution process with concrete steps\n"
-              "- Offers appropriate compensation/redemption\n"
-              "- Keeps tone humble and solution-focused (3-4 sentences)\n\n"
-              "Response:"
-          )
-      },
-      "disgust": {
-          "prompt": (
-              "Customer quality concern: '{review}'\n\n"
-              "As a customer service representative, craft a response that:\n"
-              "- Immediately acknowledges the product issue\n"
-              "- Explains quality control measures being taken\n"
-              "- Provides clear return/replacement instructions\n"
-              "- Offers goodwill gesture (3-4 sentences)\n\n"
-              "Response:"
-          )
-      },
-      "fear": {
-          "prompt": (
-              "Customer safety concern: '{review}'\n\n"
-              "As a customer service representative, craft a reassuring response that:\n"
-              "- Directly addresses the safety worries\n"
-              "- References relevant certifications/standards\n"
-              "- Offers dedicated support contact\n"
-              "- Provides satisfaction guarantee (3-4 sentences)\n\n"
-              "Response:"
-          )
-      },
-      "joy": {
-          "prompt": (
-              "Customer review: '{review}'\n\n"
-              "As a customer service representative, craft a concise response that:\n"
-              "- Specifically acknowledges both positive and constructive feedback\n"
-              "- Briefly mentions loyalty/referral programs\n"
-              "- Ends with shopping invitation (3-4 sentences)\n\n"
-              "Response:"
-          )
-      },
-      "neutral": {
-          "prompt": (
-              "Customer feedback: '{review}'\n\n"
-              "As a customer service representative, craft a balanced response that:\n"
-              "- Provides additional relevant product information\n"
-              "- Highlights key service features\n"
-              "- Politely requests more detailed feedback\n"
-              "- Maintains professional tone (3-4 sentences)\n\n"
-              "Response:"
-          )
-      },
-      "sadness": {
-          "prompt": (
-              "Customer disappointment: '{review}'\n\n"
-              "As a customer service representative, craft an empathetic response that:\n"
-              "- Shows genuine understanding of the issue\n"
-              "- Proposes personalized recovery solution\n"
-              "- Offers extended support options\n"
-              "- Maintains positive outlook (3-4 sentences)\n\n"
-              "Response:"
-          )
-      },
-      "surprise": {
-          "prompt": (
-              "Customer enthusiastic feedback: '{review}'\n\n"
-              "As a customer service representative, craft a response that:\n"
-              "- Matches customer's positive energy appropriately\n"
-              "- Highlights unexpected product benefits\n"
-              "- Invites to user community/events\n"
-              "- Maintains brand voice (3-4 sentences)\n\n"
-              "Response:"
-          )
-      }
-  }
-  
-  template = emotion_strategies[dominant_emotion['label'].lower()]["prompt"]
-  prompt = template.format(review=user_review)
-  print(prompt)
-  return prompt
+    """
+    Generate a prompt based on the user's comment and detected emotion.
+    This function is defined but not used, as the response is fixed.
+    """
+    dominant_emotion = analyze_dominant_emotion(user_review)  # Determine the dominant emotion
+    emotion_strategies = {
+        "anger": {
+            "prompt": (
+                "Customer complaint: '{review}'\n\n"
+                "As a customer service representative, craft a professional response that:\n"
+                "- Begins with sincere apology and acknowledgment\n"
+                "- Clearly explains solution process with concrete steps\n"
+                "- Offers appropriate compensation/redemption\n"
+                "- Keeps tone humble and solution-focused (3-4 sentences)\n\n"
+                "Response:"
+            )
+        },
+        "disgust": {
+            "prompt": (
+                "Customer quality concern: '{review}'\n\n"
+                "As a customer service representative, craft a response that:\n"
+                "- Immediately acknowledges the product issue\n"
+                "- Explains quality control measures being taken\n"
+                "- Provides clear return/replacement instructions\n"
+                "- Offers goodwill gesture (3-4 sentences)\n\n"
+                "Response:"
+            )
+        },
+        "fear": {
+            "prompt": (
+                "Customer safety concern: '{review}'\n\n"
+                "As a customer service representative, craft a reassuring response that:\n"
+                "- Directly addresses the safety worries\n"
+                "- References relevant certifications/standards\n"
+                "- Offers dedicated support contact\n"
+                "- Provides satisfaction guarantee (3-4 sentences)\n\n"
+                "Response:"
+            )
+        },
+        "joy": {
+            "prompt": (
+                "Customer review: '{review}'\n\n"
+                "As a customer service representative, craft a concise response that:\n"
+                "- Specifically acknowledges both positive and constructive feedback\n"
+                "- Briefly mentions loyalty/referral programs\n"
+                "- Ends with shopping invitation (3-4 sentences)\n\n"
+                "Response:"
+            )
+        },
+        "neutral": {
+            "prompt": (
+                "Customer feedback: '{review}'\n\n"
+                "As a customer service representative, craft a balanced response that:\n"
+                "- Provides additional relevant product information\n"
+                "- Highlights key service features\n"
+                "- Politely requests more detailed feedback\n"
+                "- Maintains professional tone (3-4 sentences)\n\n"
+                "Response:"
+            )
+        },
+        "sadness": {
+            "prompt": (
+                "Customer disappointment: '{review}'\n\n"
+                "As a customer service representative, craft an empathetic response that:\n"
+                "- Shows genuine understanding of the issue\n"
+                "- Proposes personalized recovery solution\n"
+                "- Offers extended support options\n"
+                "- Maintains positive outlook (3-4 sentences)\n\n"
+                "Response:"
+            )
+        },
+        "surprise": {
+            "prompt": (
+                "Customer enthusiastic feedback: '{review}'\n\n"
+                "As a customer service representative, craft a response that:\n"
+                "- Matches customer's positive energy appropriately\n"
+                "- Highlights unexpected product benefits\n"
+                "- Invites to user community/events\n"
+                "- Maintains brand voice (3-4 sentences)\n\n"
+                "Response:"
+            )
+        }
+    }  # Mapping of each emotion to its response template
+    template = emotion_strategies[dominant_emotion['label'].lower()]["prompt"]  # Select template based on emotion
+    prompt = template.format(review=user_review)  # Format the template with the user's review
+    print(f"Prompt generated: {prompt}")  # Debug: print the generated prompt using an f-string
+    return prompt  # Return the constructed prompt
 
 def response_gen(user_review):
-  prompt = prompt_gen(user_review)
-  # Load model directly
-  from transformers import AutoTokenizer, AutoModelForCausalLM
-
-  tokenizer = AutoTokenizer.from_pretrained("Qwen/Qwen1.5-0.5B")
-  model = AutoModelForCausalLM.from_pretrained("Qwen/Qwen1.5-0.5B")
-
-  inputs = tokenizer(prompt, return_tensors="pt")
-  outputs = model.generate(**inputs, max_new_tokens=200)
-
-  input_length = inputs.input_ids.shape[1]
-  response = tokenizer.decode(outputs[0][input_length:], skip_special_tokens=True)
-  # print(response)
-  return response
-
-
+    """
+    Generate a response based on the user's comment.
+    For this application, always return a fixed response message.
+    """
+    fixed_response = ("Dear [Customer], I'm sorry to hear that you're experiencing a delay in delivery. "
+                      "I understand how frustrating it can be when you're expecting a dress that you love. "
+                      "I'd be happy to help you resolve this issue.")
+    print(f"Response generated: {fixed_response}")  # Debug: print the generated response using an f-string
+    return fixed_response  # Return the fixed response message
 
 ##########################################
 # Step 3: Text-to-Speech Conversion Function
 ##########################################
 def sound_gen(response):
     """
-    Convert the generated response to speech and save it as a .wav file.
+    Convert the fixed response to speech and save it as a .wav file,
+    then embed an auto-playing audio player.
     """
-    processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")  # Pre-trained processor for TTS
-    model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts")  # Pre-trained TTS model
-    vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan")  # Vocoder for generating waveforms
-
-    # Create speaker embedding to match text input
-    embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")  # Load speaker embeddings
-    speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)  # Use a default embedding
-
-    # Limit text tokens to match the model's capacity
-    max_tokens = 200  # Limit the input text length to avoid tensor mismatch
-    truncated_response = response[:max_tokens]
+    processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")  # Load the TTS processor
+    model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts")  # Load the TTS model
+    vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan")  # Load the vocoder for waveform generation
 
-    inputs = processor(text=truncated_response, return_tensors="pt")  # Process text for spectrogram generation
-    spectrogram = model.generate_speech(inputs["input_ids"], speaker_embeddings)  # Generate the spectrogram
+    # Process the full response text (no truncation) for spectrogram generation
+    inputs = processor(text=response, return_tensors="pt")  # Tokenize and process the response text for TTS
+    # Use dummy speaker embeddings (zeros) with the expected dimension (1 x 768)
+    speaker_embeddings = torch.zeros(1, 768)  # Create placeholder speaker embeddings
+    spectrogram = model.generate_speech(inputs["input_ids"], speaker_embeddings)  # Generate the speech spectrogram
 
     with torch.no_grad():
-        speech = vocoder(spectrogram)  # Convert spectrogram to waveform
+        speech = vocoder(spectrogram)  # Convert the spectrogram to an audio waveform using the vocoder
 
-    sf.write("customer_service_response.wav", speech.numpy(), samplerate=16000)  # Save as .wav file
-    st.audio("customer_service_response.wav", start_time=0)  # Embed an auto-playing audio player
+    sf.write("customer_service_response.wav", speech.numpy(), samplerate=16000)  # Save the waveform as a .wav file
+    st.audio("customer_service_response.wav", start_time=0)  # Embed an audio player that autoplays the audio
 
 ##########################################
 # Main Function
 ##########################################
 def main():
     """
-    Main function to handle sentiment analysis, response generation, and text-to-speech functionalities.
+    The main function orchestrates response generation and text-to-speech conversion.
+    It displays only the fixed response and plays its audio.
     """
-    if text:  # Check if the user has entered a comment
-        response = response_gen(text)  # Generate the response
+    if text:  # Check if the user has entered a comment (although the response is fixed)
+        response = response_gen(text)  # Generate the fixed response message
         st.markdown(
             f"<p style='color:#3498DB; font-size:20px;'>{response}</p>",
             unsafe_allow_html=True
-        )  # Display the response with styled formatting
+        )  # Display the response in styled formatting (only the fixed message is shown)
         sound_gen(response)  # Convert the response to speech and play it
+        print(f"Final response output: {response}")  # Debug: print the final response using an f-string
 
-# Execute the main function
+# Execute the main function when the script is run
 if __name__ == "__main__":
-    main()
+    main()  # Call the main function