Enhance Gradio interface and audio conversion capabilities

- Added audio format conversion functionality using pydub, supporting WAV, MP3, and AAC formats.
- Improved error handling for voice directory access and audio conversion processes.
- Updated README to reflect new web interface features and installation requirements, including FFmpeg.
- Enhanced the TTS generation function to utilize the correct Python interpreter across platforms.
- Documented new features in the README, including real-time progress monitoring and network sharing capabilities.

README.md

@@ -1,6 +1,6 @@
 # Kokoro TTS Local
 
-A local implementation of the Kokoro Text-to-Speech model, featuring dynamic module loading and automatic dependency management.
+A local implementation of the Kokoro Text-to-Speech model, featuring dynamic module loading, automatic dependency management, and a web interface.
 
 ## Current Status
 
@@ -12,6 +12,7 @@ The project has been updated with:
 - Improved error handling and debugging
 - Interactive CLI interface
 - Cross-platform setup scripts
+- Web interface with Gradio
 
 ## Features
 
@@ -24,12 +25,24 @@ The project has been updated with:
 - Dynamic module loading from Hugging Face
 - Comprehensive error handling and logging
 - Cross-platform support (Windows, Linux, macOS)
+- **NEW: Web Interface Features**
+  - Modern, user-friendly UI
+  - Real-time generation progress
+  - Multiple output formats (WAV, MP3, AAC)
+  - Network sharing capabilities
+  - Audio playback and download
+  - Voice selection dropdown
+  - Detailed process logging
 
 ## Prerequisites
 
 - Python 3.8 or higher
 - Git (for cloning the repository)
 - Internet connection (for initial model download)
+- FFmpeg (required for MP3/AAC conversion):
+  - Windows: Automatically installed with pydub
+  - Linux: `sudo apt-get install ffmpeg`
+  - macOS: `brew install ffmpeg`
 
 ## Dependencies
 
@@ -42,21 +55,37 @@ munch
 soundfile
 huggingface-hub
 espeakng-loader
+gradio>=4.0.0
+pydub  # For audio format conversion
 ```
 
 ## Setup
 
 ### Windows
-Run the PowerShell setup script:
 ```powershell
+# Clone the repository
+git clone https://github.com/PierrunoYT/Kokoro-TTS-Local.git
+cd Kokoro-TTS-Local
+
+# Run the setup script
 .\setup.ps1
 ```
 
 ### Linux/macOS
-Run the bash setup script:
 ```bash
+# Clone the repository
+git clone https://github.com/PierrunoYT/Kokoro-TTS-Local.git
+cd Kokoro-TTS-Local
+
+# Run the setup script
 chmod +x setup.sh
 ./setup.sh
+
+# Install FFmpeg (if needed)
+# Linux:
+sudo apt-get install ffmpeg
+# macOS:
+brew install ffmpeg
 ```
 
 ### Manual Setup
@@ -79,34 +108,37 @@ python -m pip install --upgrade pip
 pip install -r requirements.txt
 ```
 
+3. Install system dependencies:
+```bash
+# Windows
+# FFmpeg is automatically installed with pydub
+
+# Linux
+sudo apt-get update
+sudo apt-get install espeak-ng ffmpeg
+
+# macOS
+brew install espeak ffmpeg
+```
+
 ## Usage
 
-### List Available Voices
-To see all available voices from the Hugging Face repository:
+### Web Interface
 ```bash
-python tts_demo.py --list-voices
+# Start the web interface
+python gradio_interface.py
 ```
+This will:
+1. Launch a web interface at http://localhost:7860
+2. Create a public share link (optional)
+3. Allow you to:
+   - Input text to synthesize
+   - Select from available voices
+   - Choose output format (WAV/MP3/AAC)
+   - Monitor generation progress
+   - Play or download generated audio
 
-### Basic Usage
-Run the demo script with default text and voice:
-```bash
-python tts_demo.py
-```
-
-### Custom Text
-Specify your own text:
-```bash
-python tts_demo.py --text "Your custom text here"
-```
-
-### Voice Selection
-Choose a different voice (use --list-voices to see available options):
-```bash
-python tts_demo.py --voice "af" --text "Custom text with specific voice"
-```
-
-### Interactive Mode
-If you run without any arguments, you'll be prompted to enter text interactively:
+### Command Line Interface
 ```bash
 python tts_demo.py
 ```
@@ -133,6 +165,11 @@ The script will:
   - Interactive text input mode
   - Voice selection and listing
   - Error handling and user feedback
+- `gradio_interface.py`: Web interface implementation
+  - Modern, responsive UI
+  - Real-time progress monitoring
+  - Multiple output formats
+  - Network sharing capabilities
 - `setup.ps1`: Windows PowerShell setup script
   - Environment creation
   - Dependency installation
@@ -156,7 +193,7 @@ The project uses the Kokoro-82M model from Hugging Face:
 
 - Sample rate: 22050Hz
 - Input: Text in any language (English recommended)
-- Output: WAV audio file
+- Output: WAV/MP3/AAC audio file
 - Dependencies are automatically managed
 - Modules are dynamically loaded from Hugging Face
 - Error handling includes stack traces for debugging

gradio_interface.py

@@ -14,18 +14,20 @@ Key Features:
 Dependencies:
 - gradio: Web interface framework
 - soundfile: Audio file handling
+- pydub: Audio format conversion
 - models: Custom module for voice model management
 """
 
 import gradio as gr
 import subprocess
 import os
+import sys
 import platform
 from datetime import datetime
 import shutil
-import json
-import soundfile as sf
 from pathlib import Path
+import soundfile as sf
+from pydub import AudioSegment
 
 # Global configuration
 CONFIG_FILE = "tts_config.json"  # Stores user preferences and paths
@@ -42,51 +44,52 @@ def get_default_voices_path():
 
 def get_available_voices():
     """Get list of available voice models by checking the directory."""
-    voices_path = get_default_voices_path()  # Use platform-agnostic path
+    voices_path = get_default_voices_path()
     try:
-        # List all files in the directory and filter by .pt extension
+        if not os.path.exists(voices_path):
+            print(f"Voices directory not found: {voices_path}")
+            return []
         voices = [os.path.splitext(f)[0] for f in os.listdir(voices_path) if f.endswith('.pt')]
-        print("Available voices:", voices)  # Debugging log
+        print("Available voices:", voices)
         return voices
     except Exception as e:
         print(f"Error retrieving voices: {e}")
-        return []  # Return an empty list if there's an error
+        return []
+
+def convert_audio(input_path: str, output_path: str, format: str):
+    """Convert audio to specified format using pydub."""
+    try:
+        audio = AudioSegment.from_wav(input_path)
+        if format == "mp3":
+            audio.export(output_path, format="mp3", bitrate="192k")
+        elif format == "aac":
+            audio.export(output_path, format="aac", bitrate="192k")
+        else:  # wav
+            shutil.copy2(input_path, output_path)
+        return True
+    except Exception as e:
+        print(f"Error converting audio: {e}")
+        return False
 
 def generate_tts_with_logs(voice, text, format):
-    """Generate TTS audio with real-time logging and format conversion.
-    
-    This function:
-    1. Validates input text
-    2. Runs TTS generation subprocess
-    3. Streams progress logs in real-time
-    4. Converts output to requested format
-    5. Saves with timestamp in output directory
-    
-    Args:
-        voice (str): Selected voice model identifier (e.g., "af", "af_bella")
-        text (str): Input text to synthesize
-        format (str): Output audio format ("wav", "mp3", or "aac")
-    
-    Yields:
-        tuple: (log_text, output_path)
-            - log_text (str): Accumulated process logs
-            - output_path (str): Path to generated audio file, or None on error
-    
-    Notes:
-        - Temporary WAV file is created and deleted after conversion
-        - Output filename includes timestamp to prevent overwrites
-        - Errors are caught and reported in logs
-    """
+    """Generate TTS audio with real-time logging and format conversion."""
     if not text.strip():
         return "❌ Error: Text required", None
     
     logs_text = ""
     try:
+        # Use sys.executable to ensure correct Python interpreter
+        cmd = [sys.executable, "tts_demo.py", "--text", text, "--voice", voice]
+        
+        # Use shell=True on Windows
+        shell = platform.system().lower() == "windows"
+        
         process = subprocess.Popen(
-            ["python", "tts_demo.py", "--text", text, "--voice", voice],
+            cmd,
             stdout=subprocess.PIPE,
             stderr=subprocess.STDOUT,
-            universal_newlines=True
+            universal_newlines=True,
+            shell=shell
         )
         
         while True:
@@ -112,15 +115,14 @@ def generate_tts_with_logs(voice, text, format):
         os.makedirs(DEFAULT_OUTPUT_DIR, exist_ok=True)
         output_path = Path(DEFAULT_OUTPUT_DIR) / filename
         
-        if format == "wav":
-            shutil.copy2("output.wav", output_path)
-        else:
-            data, samplerate = sf.read("output.wav")
-            sf.write(str(output_path), data, samplerate)
-        
-        os.remove("output.wav")
+        # Convert audio using pydub
+        if convert_audio("output.wav", str(output_path), format):
             logs_text += f"✅ Saved: {output_path}\n"
+            os.remove("output.wav")
             yield logs_text, str(output_path)
+        else:
+            logs_text += "❌ Audio conversion failed\n"
+            yield logs_text, None
 
     except Exception as e:
         logs_text += f"❌ Error: {str(e)}\n"

requirements.txt

@@ -6,3 +6,5 @@ munch
 soundfile
 huggingface-hub
 espeakng-loader
+gradio>=4.0.0
+pydub  # For audio format conversion