touches

app.py

@@ -1,24 +1,33 @@
 import gradio as gr
 import os
-
-def process_input(text_input, speaker_audio, speaker_name, option_selected):
-    if speaker_audio is None or speaker_name.strip() == "":
-        return "Please provide a valid audio file and speaker name."
-
-    # Save speaker audio under the name of the speaker
-    speaker_audio_path = f"{speaker_name}.wav"
-    speaker_audio.save(speaker_audio_path)
-
-    # Placeholder for generating the output audio
-    output_audio_path = f"generated_{speaker_name}.wav"
-
-    # Assuming some TTS or cloning process here
-    # Generate and save the output audio
-    # Replace this with your actual processing logic
-    with open(output_audio_path, "wb") as f:
-        f.write(b"Placeholder audio data")  # Replace with actual audio data
-
-    return output_audio_path
+from coqui_tts import run_audio_generation_v1
+
+def process_audio(input_text, speaker_audio, speaker_name, option_selected):
+    try:
+        # Ensure necessary directories exist
+        os.makedirs("./tmp/audio/input_src/", exist_ok=True)
+        os.makedirs("audio", exist_ok=True)
+        
+        # Save speaker audio to the required location
+        speaker_audio_path = "./tmp/audio/input_src/0.wav"
+        with open(speaker_audio_path, "wb") as f:
+            f.write(speaker_audio)
+        
+        # Check selected option and execute corresponding function
+        if option_selected == "Xtts_v2":
+            # Generate TTS audio using run_audio_generation_v1
+            run_audio_generation_v1(input_text)
+        else:
+            return f"The option '{option_selected}' is not implemented yet."
+        
+        # Save the output audio under the speaker's name
+        speaker_output_path = f"audio/{speaker_name}.wav"
+        os.rename("audio/output.wav", speaker_output_path)
+        
+        return speaker_output_path
+    
+    except Exception as e:
+        return str(e)
 
 # Gradio interface
 with gr.Blocks() as demo:
@@ -31,14 +40,14 @@ with gr.Blocks() as demo:
         speaker_audio = gr.Audio(label="Speaker Audio (to be cloned)",type='filepath', format='wav')
         speaker_name = gr.Textbox(label="Speaker Name", placeholder="Enter the speaker's name.")
 
-    option_selected = gr.Dropdown(choices=["Option 1", "Option 2", "Option 3"], label="Select an Option")
+    option_selected = gr.Dropdown(choices=["Xtts_v2", "metaVoice(not working at the moment)", "more"], label="Select an Option")
 
     submit_btn = gr.Button("Submit")
 
     output_audio = gr.Audio(label="Generated Audio Output")
 
     submit_btn.click(
-        process_input,
+        fn=process_audio,
         inputs=[text_input, speaker_audio, speaker_name, option_selected],
         outputs=output_audio,
     )