Update app.py

app.py

@@ -14,43 +14,38 @@ import soundfile as sf
 import os
 import numpy as np
 
+import gradio as gr
+
 
 # Initialize the Kokoro pipeline
 pipeline = KPipeline(lang_code='a')  # 'a' stands for American English
 
+# Define audio output path
+AUDIO_OUTPUT_PATH = "tools/output.wav"
+
 @tool
 def text_to_speech_kokoro(text: str, voice: str = 'af_heart', speed: float = 1.0) -> str:
-    """Convert text to speech using the Kokoro-82M model.
-    
-    Args:
-        text: The text to be converted to speech.
-        voice: The voice to use for speech synthesis (default is 'af_heart').
-        speed: The speed of the speech (default is 1.0).
-    
-    Returns:
-        An AgentAudio object with the relative URL to the generated audio file.
-    """
+    """Convert text to speech using Kokoro-82M model and return audio file path."""
     try:
-        # Generate speech audio
         generator = pipeline(text, voice=voice, speed=speed, split_pattern=r'\n+')
         audio_segments = []
         for _, _, audio in generator:
             audio_segments.append(audio)
+        
         if not audio_segments:
             raise ValueError("No audio generated.")
-        # Concatenate segments into one audio array
+        
         full_audio = np.concatenate(audio_segments)
         sample_rate = 24000  # Kokoro outputs at 24 kHz
-        # Ensure the static folder exists and save the file there
+        
         os.makedirs("tools", exist_ok=True)
-        filename = os.path.join("tools", "output.wav")
-        sf.write(filename, full_audio, sample_rate)
-        # Return an AgentAudio object pointing to the relative URL of the audio file
-        from smolagents.agent_types import AgentAudio
-        return AgentAudio(f"tools/output.wav")
+        sf.write(AUDIO_OUTPUT_PATH, full_audio, sample_rate)
+        
+        return AUDIO_OUTPUT_PATH
     except Exception as e:
         return f"Error generating speech: {str(e)}"
 
+
 @tool
 def search_dad_jokes(term: str) -> str:
     """A tool that searches for dad jokes containing a specific term.
@@ -111,4 +106,21 @@ agent = CodeAgent(
 )
 
 
-GradioUI(agent).launch()
+#GradioUI(agent).launch()
+
+# Gradio wrapper function
+def gradio_text_to_speech(text):
+    audio_path = text_to_speech_kokoro(text)
+    return audio_path if os.path.exists(audio_path) else "Error: No audio file generated."
+
+# Define the Gradio UI
+with gr.Blocks() as demo:
+    gr.Markdown("### Text-to-Speech with Kokoro AI 🎙️")
+    with gr.Row():
+        input_box = gr.Textbox(label="Enter text")
+        output_audio = gr.Audio(label="Generated Speech", type="filepath")
+    
+    btn = gr.Button("Generate Speech")
+    btn.click(gradio_text_to_speech, inputs=input_box, outputs=output_audio)
+
+demo.launch()