Update app.py

app.py

@@ -1,39 +1,92 @@
 import gradio as gr
 from transformers import pipeline
 
-# Load Whisper for speech-to-text
-whisper = pipeline("automatic-speech-recognition", model="openai/whisper-medium")
-
-# Load DistilBERT for sentiment analysis
-sentiment_analyzer = pipeline("sentiment-analysis", model="distilbert-base-uncased-finetuned-sst-2-english")
-
-# Function to process audio and analyze tone
-def analyze_call(audio_file):
-    # Step 1: Transcribe audio to text
-    transcription = whisper(audio_file)["text"]
+def create_speech_analyzer():
+    # Initialize models with error handling
+    try:
+        # Load Whisper model for speech recognition
+        transcriber = pipeline(
+            "automatic-speech-recognition", 
+            model="openai/whisper-medium",
+            max_new_tokens=128
+        )
+        
+        # Load sentiment analysis model
+        sentiment_model = pipeline(
+            "sentiment-analysis",
+            model="distilbert-base-uncased-finetuned-sst-2-english"
+        )
+        
+        return transcriber, sentiment_model
     
-    # Step 2: Analyze sentiment of the transcription
-    sentiment_result = sentiment_analyzer(transcription)[0]
+    except Exception as e:
+        raise RuntimeError(f"Error loading models: {str(e)}")
+
+def analyze_speech(audio_file):
+    """
+    Analyze speech audio for transcription and sentiment.
     
-    return {
-        "transcription": transcription,
-        "sentiment": sentiment_result["label"],
-        "confidence": sentiment_result["score"]
-    }
+    Args:
+        audio_file: Path to audio file or audio data
+        
+    Returns:
+        dict: Contains transcription, sentiment and confidence score
+    """
+    try:
+        # Get model instances
+        transcriber, sentiment_model = create_speech_analyzer()
+        
+        # Transcribe audio
+        transcription = transcriber(audio_file)["text"]
+        
+        # Analyze sentiment
+        sentiment_result = sentiment_model(transcription)[0]
+        
+        return {
+            "transcription": transcription,
+            "sentiment": sentiment_result["label"],
+            "confidence": f"{sentiment_result['score']:.2%}"
+        }
+        
+    except Exception as e:
+        return {
+            "transcription": f"Error processing audio: {str(e)}",
+            "sentiment": "ERROR",
+            "confidence": "0%"
+        }
+
+def create_interface():
+    """Create and configure the Gradio interface"""
+    return gr.Interface(
+        fn=analyze_speech,
+        inputs=gr.Audio(
+            source="microphone",
+            type="filepath",
+            label="Upload or Record Audio"
+        ),
+        outputs=[
+            gr.Textbox(label="Transcription"),
+            gr.Textbox(label="Sentiment Analysis"),
+            gr.Textbox(label="Confidence Score")
+        ],
+        title="Real-Time Speech Sentiment Analyzer",
+        description="""
+        This tool transcribes speech and analyzes its sentiment in real-time.
+        Upload an audio file or record directly through your microphone.
+        """,
+        theme=gr.themes.Soft(),
+        examples=[],  # Add example audio files here if desired
+        cache_examples=True
+    )
 
-# Gradio Interface
-interface = gr.Interface(
-    fn=analyze_call,
-    inputs=gr.Audio(source="microphone", type="filepath"),
-    outputs=[
-        gr.Textbox(label="Transcription"),
-        gr.Textbox(label="Sentiment"),
-        gr.Textbox(label="Confidence")
-    ],
-    live=True,  # Enable real-time processing
-    title="Real-Time Call Analysis",
-    description="Upload or record audio to analyze tone and sentiment."
-)
+def main():
+    # Create and launch the interface
+    interface = create_interface()
+    interface.launch(
+        share=True,  # Enable sharing via public URL
+        debug=True,  # Enable debug mode for better error messages
+        server_name="0.0.0.0"  # Allow external connections
+    )
 
-# Launch the app
-interface.launch()
+if __name__ == "__main__":
+    main()