feat: Introduce comprehensive text preprocessing, structured Hugging …

README.md

@@ -12,6 +12,8 @@
 
 Kitten TTS is an open-source realistic text-to-speech model with just 15 million parameters, designed for lightweight deployment and high-quality voice synthesis.
 
+> **UI and enhancement made by Muhammad Umer**
+
 *Currently in developer preview*
 
 [Join our discord](https://discord.com/invite/VJ86W4SURW) 
@@ -59,7 +61,28 @@ pip install https://github.com/KittenML/KittenTTS/releases/download/0.8/kittentt
 
 
 
- ### Basic Usage 
+```
+
+### Graphical Interface (Windows)
+
+We have provided a native, modern UI designed specifically for Windows! 
+To use the UI:
+1. Double-click **`Run_KittenTTS.bat`** from your File Explorer.
+2. It will automatically build an environment if needed, and open the KittenTTS Desktop app.
+3. Type your text, select your voice and model, and click **Generate & Play ▶**!
+
+### Terminal CLI Usage
+
+You can also use KittenTTS directly from your terminal/command line!
+```bash
+# Generate immediate audio
+kittentts "Hello! I am speaking to you directly from the terminal" --voice Jasper --output out.wav
+
+# View all dynamically loaded voices
+kittentts --list-voices
+```
+
+ ### Basic Python Usage 
 
 ```
 from kittentts import KittenTTS

Run_KittenTTS.bat

@@ -0,0 +1,38 @@
+@echo off
+echo ==============================================
+echo        KittenTTS Graphical Interface
+echo ==============================================
+echo.
+
+:: Check if virtual environment exists
+if not exist "venv\Scripts\python.exe" (
+    echo [INFO] Virtual environment not found. Setting it up now...
+    python -m venv venv
+    if errorlevel 1 (
+        echo [ERROR] Failed to create virtual environment. Ensure Python is installed and in PATH.
+        pause
+        exit /b 1
+    )
+)
+
+:: Activate the environment and install/update dependencies quietly
+echo [INFO] Updating dependencies...
+call "venv\Scripts\activate.bat"
+if errorlevel 1 (
+    echo [ERROR] Failed to activate virtual environment.
+    pause
+    exit /b 1
+)
+
+venv\Scripts\python.exe -m pip install -e . -q
+if errorlevel 1 (
+    echo [ERROR] Initialization failed while installing dependencies.
+    pause
+    exit /b 1
+)
+
+:: Run the UI application
+echo [INFO] Launching UI...
+venv\Scripts\python.exe -m kittentts.ui
+
+pause

kittentts/cli.py

@@ -0,0 +1,45 @@
+import argparse
+import sys
+from kittentts import KittenTTS
+
+def main():
+    parser = argparse.ArgumentParser(description="KittenTTS: High-quality, ultra-lightweight Text-to-Speech")
+    parser.add_argument("text", nargs="?", help="The text to synthesize into speech")
+    parser.add_argument("--voice", default="Jasper", help="Voice to use (e.g., Bella, Jasper, Luna). Default: Jasper")
+    parser.add_argument("--model", default="KittenML/kitten-tts-nano-0.1", help="Model ID to use. Default: KittenML/kitten-tts-nano-0.1")
+    parser.add_argument("--output", default="output.wav", help="Output audio file path. Default: output.wav")
+    parser.add_argument("--speed", type=float, default=1.0, help="Speech speed (1.0 = normal). Default: 1.0")
+    parser.add_argument("--list-voices", action="store_true", help="List available voices for the model and exit")
+
+    args = parser.parse_args()
+
+    if not args.list_voices and not args.text:
+        parser.error("The 'text' argument is required unless --list-voices is specified.")
+
+    # Initialize model
+    print(f"Loading model: {args.model}...")
+    model = KittenTTS(model_name=args.model)
+
+    if model.model is None:
+        sys.exit(1)
+
+    if args.list_voices:
+        print("\n=== Available Voices ===")
+        print("Aliases:")
+        for alias in model.model.voice_aliases.keys():
+            print(f"  - {alias}")
+        print("\nNative Voices:")
+        for voice in model.available_voices:
+            if voice not in model.model.voice_aliases.values():
+                print(f"  - {voice}")
+        sys.exit(0)
+
+
+    print(f"Generating audio with voice '{args.voice}'...")
+    try:
+        model.generate_to_file(args.text, args.output, voice=args.voice, speed=args.speed)
+    except ValueError as e:
+        sys.exit(1)
+
+if __name__ == "__main__":
+    main()

kittentts/get_model.py

@@ -64,37 +64,46 @@ def download_from_huggingface(repo_id="KittenML/kitten-tts-nano-0.1", cache_dir=
     Returns:
         KittenTTS_1_Onnx: Instantiated model ready for use
     """
-    # Download config file first
-    config_path = hf_hub_download(
-        repo_id=repo_id,
-        filename="config.json",
-        cache_dir=cache_dir
-    )
-
-    # Load config
-    with open(config_path, 'r') as f:
-        config = json.load(f)
+    try:
+        # Download config file first
+        config_path = hf_hub_download(
+            repo_id=repo_id,
+            filename="config.json",
+            cache_dir=cache_dir
+        )
+
+        # Load config
+        with open(config_path, 'r') as f:
+            config = json.load(f)
 
-    if config.get("type") not in ["ONNX1", "ONNX2"]:
-        raise ValueError("Unsupported model type.")
+        if config.get("type") not in ["ONNX1", "ONNX2"]:
+            print(f"Error: Unsupported model type '{config.get('type')}'.")
+            return None
 
-    # Download model and voices files based on config
-    model_path = hf_hub_download(
-        repo_id=repo_id,
-        filename=config["model_file"],
-        cache_dir=cache_dir
-    )
-
-    voices_path = hf_hub_download(
-        repo_id=repo_id,
-        filename=config["voices"],
-        cache_dir=cache_dir
-    )
-
-    # Instantiate and return model
-    model = KittenTTS_1_Onnx(model_path=model_path, voices_path=voices_path, speed_priors=config.get("speed_priors", {}) , voice_aliases=config.get("voice_aliases", {}))
-
-    return model
+        # Download model and voices files based on config
+        model_path = hf_hub_download(
+            repo_id=repo_id,
+            filename=config["model_file"],
+            cache_dir=cache_dir
+        )
+
+        voices_path = hf_hub_download(
+            repo_id=repo_id,
+            filename=config["voices"],
+            cache_dir=cache_dir
+        )
+
+        # Instantiate and return model
+        model = KittenTTS_1_Onnx(model_path=model_path, voices_path=voices_path, speed_priors=config.get("speed_priors", {}) , voice_aliases=config.get("voice_aliases", {}))
+
+        return model
+
+    except Exception as e:
+        print(f"\n❌ Failed to load model '{repo_id}'.")
+        print(f"Error details: {str(e)}")
+        print("Please check your internet connection or verify the model name is correct.")
+        print("Example valid models: 'KittenML/kitten-tts-nano-0.1', 'KittenML/kitten-tts-mini-0.8'\n")
+        return None
 
 
 def get_model(repo_id="KittenML/kitten-tts-nano-0.1", cache_dir=None):

kittentts/onnx_model.py

@@ -95,23 +95,42 @@ def __init__(self, model_path="kitten_tts_nano_preview.onnx", voices_path="voice
         self.text_cleaner = TextCleaner()
         self.speed_priors = speed_priors
 
-        # Available voices
-        self.available_voices = [
-            'expr-voice-2-m', 'expr-voice-2-f', 'expr-voice-3-m', 'expr-voice-3-f', 
-            'expr-voice-4-m', 'expr-voice-4-f', 'expr-voice-5-m', 'expr-voice-5-f'
-        ]
-        self.voice_aliases = voice_aliases
+        # Available voices dynamically loaded from the .npz file
+        self.available_voices = list(self.voices.keys())
+
+        # Default fallback aliases, can be extended by config
+        default_aliases = {
+            'Bella': 'expr-voice-2-f',
+            'Jasper': 'expr-voice-2-m',
+            'Luna': 'expr-voice-3-f',
+            'Bruno': 'expr-voice-3-m',
+            'Rosie': 'expr-voice-4-f',
+            'Hugo': 'expr-voice-4-m',
+            'Kiki': 'expr-voice-5-f',
+            'Leo': 'expr-voice-5-m'
+        }
+        self.voice_aliases = default_aliases
+        self.voice_aliases.update(voice_aliases)
+
 
         self.preprocessor = TextPreprocessor()
 
     def _prepare_inputs(self, text: str, voice: str, speed: float = 1.0) -> dict:
         """Prepare ONNX model inputs from text and voice parameters."""
+        # Try to resolve alias if necessary
         if voice in self.voice_aliases:
             voice = self.voice_aliases[voice]
 
-        if voice not in self.available_voices:
-            raise ValueError(f"Voice '{voice}' not available. Choose from: {self.available_voices}")
-
+        # Check if the requested voice exists in the dynamic voice dictionary
+        if voice not in self.voices:
+            fallback = list(self.aliases.keys())[0] if hasattr(self, 'aliases') and len(self.aliases) > 0 else (self.available_voices[0] if self.available_voices else None)
+            error_msg = f"\n❌ Voice '{voice}' not found."
+            error_msg += f"\n👉 Available native voices: {self.available_voices}"
+            error_msg += f"\n👉 Available voice aliases: {list(self.voice_aliases.keys())}"
+            if fallback:
+                error_msg += f"\nPlease try using a valid voice like '{fallback}'."
+            print(error_msg)
+            raise ValueError(f"Voice '{voice}' not available.")
         if voice in self.speed_priors:
             speed = speed * self.speed_priors[voice]