Update app.py

app.py

@@ -1,62 +1,35 @@
 import gradio as gr
-import pyttsx3
-import os
+from TTS.api import TTS
 import uuid
-from pydub import AudioSegment
 
-# Initialize pyttsx3 engine
-engine = pyttsx3.init()
+# Load TTS model (offline)
+tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC", progress_bar=False, gpu=False)
 
-def text_to_speech_offline(text, voice, speed):
+def tts_generate(text, speed):
     if not text.strip():
         return "Please enter some text.", None
 
     try:
-        # Unique filename for each session
-        temp_id = str(uuid.uuid4())
-        wav_path = f"audio_{temp_id}.wav"
-        mp3_path = f"audio_{temp_id}.mp3"
-
-        # Set voice
-        voices = engine.getProperty('voices')
-        engine.setProperty('voice', voice)
-
-        # Set speed
-        engine.setProperty('rate', int(200 * speed))  # 200 is average speed
-
-        # Save to .wav
-        engine.save_to_file(text, wav_path)
-        engine.runAndWait()
-
-        # Convert to mp3 using pydub
-        audio = AudioSegment.from_wav(wav_path)
-        audio.export(mp3_path, format="mp3")
-
-        # Cleanup .wav
-        os.remove(wav_path)
-
-        return "Speech generated successfully!", mp3_path
+        # Generate filename
+        filename = f"output_{uuid.uuid4().hex}.wav"
+        tts.tts_to_file(text=text, file_path=filename, speaker_wav=None, speed=speed)
+        return "Speech generated successfully!", filename
 
     except Exception as e:
         return f"Error: {e}", None
 
-# Get available voices
-available_voices = pyttsx3.init().getProperty('voices')
-voice_options = {v.name: v.id for v in available_voices}
-
 iface = gr.Interface(
-    fn=text_to_speech_offline,
+    fn=tts_generate,
     inputs=[
         gr.Textbox(label="Enter Text", lines=6, placeholder="Type or paste text here..."),
-        gr.Dropdown(choices=list(voice_options.values()), label="Select Voice"),
-        gr.Slider(minimum=0.5, maximum=2.0, value=1.0, step=0.25, label="Speech Speed")
+        gr.Slider(minimum=0.5, maximum=1.5, value=1.0, step=0.1, label="Speech Speed")
     ],
     outputs=[
         gr.Text(label="Status"),
-        gr.Audio(label="Generated Audio")
+        gr.Audio(label="Generated Speech")
     ],
-    title="🔊 Offline Text-to-Speech (No Rate Limits)",
-    description="Convert any paragraph of text into speech using local TTS. No API limits. Adjust voice and speed."
+    title="🎙️ Offline TTS with Coqui",
+    description="Generate speech from text using an offline TTS model (Tacotron2-DDC). No API keys. No limits."
 )
 
 if __name__ == "__main__":