Update app.py

app.py

@@ -9,6 +9,7 @@ from pathlib import Path
 from openai import OpenAI
 import tempfile
 import shutil
+from scipy.io import wavfile
 
 # Initialize OpenAI clients
 llm = LangChainOpenAI(temperature=0.7)
@@ -103,6 +104,14 @@ def generate_speech(text, filename):
         print(f"Error generating speech: {e}")
         return None
 
+def save_wav(audio, sample_rate, filename):
+    try:
+        wavfile.write(filename, sample_rate, audio)
+        return filename
+    except Exception as e:
+        print(f"Error saving WAV file: {e}")
+        return None
+
 def generate_story_and_music(topic):
     with tempfile.TemporaryDirectory() as temp_dir:
         try:
@@ -136,9 +145,13 @@ def generate_story_and_music(topic):
         try:
             music_response = query(music_payload, config)
             music_audio, music_sample_rate = process_audio(music_response)
+            
+            # Save music as WAV file
+            music_file = "generated_music.wav"
+            save_wav(music_audio, music_sample_rate, music_file)
         except Exception as e:
-            print(f"Error generating music: {e}")
-            music_audio, music_sample_rate = None, None
+            print(f"Error generating or saving music: {e}")
+            music_audio, music_sample_rate, music_file = None, None, None
 
         # Generate speech for the entire story
         speech_file = os.path.join(temp_dir, "speech.mp3")
@@ -151,21 +164,13 @@ def generate_story_and_music(topic):
         else:
             speech_file = None
 
-        return story, final_music_prompt, music_audio, music_sample_rate, speech_file
+        return story, final_music_prompt, music_file, speech_file
 
 # Create the Gradio interface
 def gradio_interface(topic):
-    story, music_prompt, music_audio, music_sample_rate, speech_file = generate_story_and_music(topic)
-    
-    outputs = [story, music_prompt]
-    
-    if music_audio is not None and music_sample_rate is not None:
-        outputs.append((music_sample_rate, music_audio))
-    else:
-        outputs.append(None)
+    story, music_prompt, music_file, speech_file = generate_story_and_music(topic)
     
-    # Add speech output
-    outputs.append(speech_file)
+    outputs = [story, music_prompt, music_file, speech_file]
     
     return outputs
 
@@ -175,11 +180,11 @@ iface = gr.Interface(
     outputs=[
         gr.Textbox(label="Generated African-based Story"),
         gr.Textbox(label="Generated Music Prompt"),
-        gr.Audio(label="Generated Music", type="numpy", autoplay=True),
+        gr.Audio(label="Generated Music", type="filepath"),
         gr.Audio(label="Story Speech")
     ],
     title="African Story and Music Generator with Speech",
-    description="Enter a topic, and the AI will generate a short African-based story, matching background music, and speech for the entire story. The music will autoplay. If any part fails, you'll still see the successful parts."
+    description="Enter a topic, and the AI will generate a short African-based story, matching background music, and speech for the entire story. If any part fails, you'll still see the successful parts."
 )
 
 # Launch the Gradio app