Update app.py

app.py

@@ -23,40 +23,50 @@ torch.load = original_load
 
 def generate_speech(reference_audio, text):
     """
-    Generate speech audio mimicking the voice from the reference audio using Bark.
+    Generate speech audio using a pre-defined speaker.
     
     Parameters:
-    reference_audio (str): Filepath to the uploaded voice sample.
+    reference_audio (str): Path to uploaded audio (ignored in this version).
     text (str): Text to convert to speech.
     
     Returns:
-    str: Path to the generated audio file
+    str: Path to the generated audio file.
     """
-    # Generate speech using the reference audio and text
-    audio_array = generate_audio(text, history_prompt=reference_audio)
+    # Use a pre-defined speaker since custom voice cloning isn't supported
+    history_prompt = "v2/en_speaker_6"  # Pre-defined speaker ID
+    audio_array = generate_audio(text, history_prompt=history_prompt)
     
-    # Create a temporary file to save the audio
-    temp_file = tempfile.NamedTemporaryFile(suffix=".wav", delete=False)
-    temp_file_path = temp_file.name
-    # Save the audio to the temporary file
-    write_wav(temp_file_path, SAMPLE_RATE, audio_array)
-    temp_file.close()
+    # Save the audio to a temporary file
+    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_file:
+        write_wav(temp_file.name, SAMPLE_RATE, audio_array)
+        temp_file_path = temp_file.name
     
     return temp_file_path
 
 # Build the Gradio interface
-with gr.Blocks(title="Voice Cloning TTS with Bark") as app:
-    gr.Markdown("## Voice Cloning Text-to-Speech with Bark")
-    gr.Markdown("Upload a short voice sample in English, then enter text to hear it in your voice!")
+with gr.Blocks(title="Text-to-Speech with Bark") as app:
+    gr.Markdown("## Text-to-Speech with Bark")
+    gr.Markdown(
+        "Enter text to hear it in a pre-defined voice. "
+        "Custom voice cloning from uploaded audio is not supported in this version."
+    )
     
-    with gr.Row():
-        audio_input = gr.Audio(type="filepath", label="Upload Your Voice Sample (English)")
-        text_input = gr.Textbox(label="Enter Text to Convert to Speech", placeholder="e.g., I love chocolate")
+    # Input components
+    audio_input = gr.Audio(
+        type="filepath",
+        label="Upload Your Voice Sample (English, Ignored)",
+        visible=True  # Kept for future functionality, but ignored
+    )
+    text_input = gr.Textbox(
+        label="Enter Text to Convert to Speech",
+        placeholder="e.g., I love chocolate"
+    )
     
-    generate_btn = gr.Button("Generate Speech")
+    # Output component
     audio_output = gr.Audio(label="Generated Speech", interactive=False)
     
-    # Connect the button to the generation function
+    # Button to trigger generation
+    generate_btn = gr.Button("Generate Speech")
     generate_btn.click(
         fn=generate_speech,
         inputs=[audio_input, text_input],