Update app.py

app.py

@@ -2,21 +2,22 @@ import os
 import sys
 import subprocess
 
-## restart
+# 1. Clone the repo if it doesn't exist
 if not os.path.exists("LuxTTS"):
     subprocess.run(["git", "clone", "https://github.com/ysharma3501/LuxTTS.git"])
 
-
+# 2. Install requirements from the cloned folder
+# This ensures all dependencies (transformers, librosa, etc.) are present
 subprocess.run([sys.executable, "-m", "pip", "install", "-r", "LuxTTS/requirements.txt"])
 
-
+# 3. Add to path so the 'zipvoice' module is importable
 sys.path.append(os.path.abspath("LuxTTS"))
 
 import gradio as gr
 import torch
-from zipvoice.luxvoice import LuxTTS
+from zipvoice.luxtts import LuxTTS
 
-# Standard setup continues...
+# Init Model
 device = "cuda" if torch.cuda.is_available() else "cpu"
 lux_tts = LuxTTS('YatharthS/LuxTTS', device=device, threads=2)
 
@@ -24,7 +25,10 @@ def infer(text, audio_prompt, rms, t_shift, num_steps, speed, return_smooth):
     if audio_prompt is None or not text:
         return None
     
+    # Encode reference
     encoded_prompt = lux_tts.encode_prompt(audio_prompt, rms=rms)
+    
+    # Generate speech with ALL params
     final_wav = lux_tts.generate_speech(
         text, 
         encoded_prompt, 
@@ -37,19 +41,32 @@ def infer(text, audio_prompt, rms, t_shift, num_steps, speed, return_smooth):
 
 # Gradio UI
 with gr.Blocks() as demo:
-    gr.Markdown("# LuxTTS (Automatic Setup)")
+    gr.Markdown("# 🎙️ LuxTTS Voice Cloning")
+    
     with gr.Row():
         with gr.Column():
-            input_text = gr.Textbox(label="Text", value="Hello there!")
-            input_audio = gr.Audio(label="Reference Voice", type="filepath")
+            input_text = gr.Textbox(label="Text to Synthesize", value="Hey, what's up? I'm feeling really great!")
+            input_audio = gr.Audio(label="Reference Audio (.wav)", type="filepath")
+            
             with gr.Row():
-                rms_val = gr.Number(value=0.01, label="RMS")
+                rms_val = gr.Number(value=0.01, label="RMS (Loudness)")
                 t_shift_val = gr.Number(value=0.9, label="T-Shift")
-                steps_val = gr.Slider(1, 10, value=4, step=1, label="Steps")
-            btn = gr.Button("Synthesize")
+                steps_val = gr.Slider(1, 10, value=4, step=1, label="Num Steps")
+            
+            with gr.Row():
+                speed_val = gr.Slider(0.5, 2.0, value=1.0, step=0.1, label="Speed (Lower = Faster)")
+                smooth_val = gr.Checkbox(label="Return Smooth", value=False)
+            
+            btn = gr.Button("Generate Speech", variant="primary")
+            
         with gr.Column():
-            audio_out = gr.Audio(label="Output")
+            audio_out = gr.Audio(label="Result")
 
-    btn.click(infer, [input_text, input_audio, rms_val, t_shift_val, steps_val], audio_out)
+    # Fixed: Passing all inputs to match the infer function signature
+    btn.click(
+        fn=infer, 
+        inputs=[input_text, input_audio, rms_val, t_shift_val, steps_val, speed_val, smooth_val], 
+        outputs=audio_out
+    )
 
 demo.launch()