Fix initialization and add README

README.md

@@ -1,12 +1,14 @@
 ---
-title: AI Virtual
-emoji: 🐨
-colorFrom: pink
+title: AI Virtual Therapist
+emoji: 🧠
+colorFrom: blue
 colorTo: purple
 sdk: gradio
-sdk_version: 5.12.0
+sdk_version: 4.12.0
 app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+# AI Virtual Therapist
+
+This application provides text emotion analysis, voice emotion analysis, and chat with text-to-speech capabilities.

app.py

@@ -1,5 +1,5 @@
 import gradio as gr
-from fastapi import FastAPI, UploadFile, File, HTTPException
+from fastapi import FastAPI
 import librosa
 import openai
 from transformers import pipeline
@@ -14,9 +14,9 @@ app = FastAPI()
 # Initialize emotion classifier
 text_emotion_classifier = pipeline("text-classification", 
                                  model="bhadresh-savani/distilbert-base-uncased-emotion",
-                                 device=-1)  # Use CPU
+                                 device=-1)
 
-# Environment variables will be set in Hugging Face Spaces
+# Environment variables
 OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")
 ELEVEN_LABS_API_KEY = os.getenv("ELEVEN_LABS_API_KEY")
 VOICE_ID = os.getenv("VOICE_ID", "9BWtsMINqrJLrRacOk9x")
@@ -31,16 +31,16 @@ def analyze_text_emotion(text):
 
 def analyze_voice_emotion(audio):
     try:
-        # Convert audio to numpy array
+        if audio is None:
+            return "Please upload an audio file"
+        
         y = audio[1]
         sr = audio[0]
         
-        # Extract features
         pitch = float(librosa.feature.spectral_centroid(y=y, sr=sr).mean())
         intensity = float(librosa.feature.rms(y=y).mean())
         tempo, _ = librosa.beat.beat_track(y=y, sr=sr)
 
-        # Simple emotion classification
         if pitch < 150 and intensity < 0.02:
             emotion = "sadness"
         elif pitch > 200 and intensity > 0.05:
@@ -52,11 +52,13 @@ def analyze_voice_emotion(audio):
 
         return f"Emotion: {emotion}\nPitch: {pitch:.2f}\nIntensity: {intensity:.2f}\nTempo: {tempo:.2f}"
     except Exception as e:
-        return f"Error: {str(e)}"
+        return f"Error analyzing audio: {str(e)}"
 
 def chat_and_tts(message):
     try:
-        # Get ChatGPT response
+        if not OPENAI_API_KEY or not ELEVEN_LABS_API_KEY:
+            return "API keys not configured", None
+            
         openai.api_key = OPENAI_API_KEY
         chat_response = openai.ChatCompletion.create(
             model="gpt-3.5-turbo",
@@ -67,7 +69,6 @@ def chat_and_tts(message):
         )
         response_text = chat_response['choices'][0]['message']['content'].strip()
 
-        # Convert to speech using Eleven Labs
         url = f"https://api.elevenlabs.io/v1/text-to-speech/{VOICE_ID}"
         headers = {
             "xi-api-key": ELEVEN_LABS_API_KEY,
@@ -85,7 +86,6 @@ def chat_and_tts(message):
         if response.status_code != 200:
             return response_text, None
             
-        # Save audio temporarily
         audio_path = "response.mp3"
         with open(audio_path, "wb") as f:
             f.write(response.content)
@@ -95,7 +95,9 @@ def chat_and_tts(message):
         return f"Error: {str(e)}", None
 
 # Create Gradio interface
-with gr.Blocks(title="AI Therapist") as demo:
+demo = gr.Blocks(title="AI Therapist")
+
+with demo:
     gr.Markdown("# AI Virtual Therapist")
     
     with gr.Tab("Text Emotion Analysis"):
@@ -117,5 +119,5 @@ with gr.Blocks(title="AI Therapist") as demo:
         audio_output = gr.Audio(label="Voice Response")
         chat_button.click(chat_and_tts, inputs=chat_input, outputs=[chat_output, audio_output])
 
-# Mount Gradio app to FastAPI
-app = gr.mount_gradio_app(app, demo, path="/")
+if __name__ == "__main__":
+    demo.launch(server_name="0.0.0.0", server_port=7860)

requirements.txt

@@ -1,5 +1,5 @@
+gradio==4.12.0
 fastapi
-gradio
 uvicorn
 python-multipart
 openai
@@ -8,4 +8,5 @@ transformers
 torch
 requests
 python-dotenv
-soundfile
+soundfile
+numpy