abstractvoice 0.4.1__py3-none-any.whl → 0.5.0__py3-none-any.whl

abstractvoice/voice_manager.py

@@ -38,8 +38,8 @@ class VoiceManager:
     # Smart language configuration - high quality stable defaults
     LANGUAGES = {
         'en': {
-            'default': 'tts_models/en/ljspeech/vits',          # High quality premium voice
-            'premium': 'tts_models/en/ljspeech/vits',          # Use same stable model
+            'default': 'tts_models/en/ljspeech/tacotron2-DDC', # Reliable, compatible voice
+            'premium': 'tts_models/en/ljspeech/vits',          # High quality (requires espeak)
             'name': 'English'
         },
         'fr': {
@@ -70,15 +70,39 @@ class VoiceManager:
     # Complete voice catalog with metadata
     VOICE_CATALOG = {
         'en': {
-            'vits_premium': {
-                'model': 'tts_models/en/ljspeech/vits',
-                'quality': 'premium',
+            'tacotron2': {
+                'model': 'tts_models/en/ljspeech/tacotron2-DDC',
+                'quality': 'good',
                 'gender': 'female',
                 'accent': 'US English',
                 'license': 'Open source (LJSpeech)',
-                'requires': 'espeak-ng'
+                'requires': 'none'
+            },
+            'jenny': {
+                'model': 'tts_models/en/jenny/jenny',
+                'quality': 'excellent',
+                'gender': 'female',
+                'accent': 'US English',
+                'license': 'Open source (Jenny)',
+                'requires': 'none'
+            },
+            'ek1': {
+                'model': 'tts_models/en/ek1/tacotron2',
+                'quality': 'excellent',
+                'gender': 'male',
+                'accent': 'British English',
+                'license': 'Open source (EK1)',
+                'requires': 'none'
+            },
+            'sam': {
+                'model': 'tts_models/en/sam/tacotron-DDC',
+                'quality': 'good',
+                'gender': 'male',
+                'accent': 'US English',
+                'license': 'Open source (Sam)',
+                'requires': 'none'
             },
-            'fast_pitch_reliable': {
+            'fast_pitch': {
                 'model': 'tts_models/en/ljspeech/fast_pitch',
                 'quality': 'good',
                 'gender': 'female',
@@ -86,12 +110,12 @@ class VoiceManager:
                 'license': 'Open source (LJSpeech)',
                 'requires': 'none'
             },
-            'vctk_multi': {
-                'model': 'tts_models/en/vctk/vits',
+            'vits': {
+                'model': 'tts_models/en/ljspeech/vits',
                 'quality': 'premium',
-                'gender': 'multiple',
-                'accent': 'British English',
-                'license': 'Open source (VCTK)',
+                'gender': 'female',
+                'accent': 'US English',
+                'license': 'Open source (LJSpeech)',
                 'requires': 'espeak-ng'
             }
         },
@@ -191,6 +215,20 @@ class VoiceManager:
                 lang_name = self.LANGUAGES[self.language]['name']
                 print(f"🌍 Using {lang_name} voice: {tts_model}")
 
+        # Initialize TTS engine with instant setup for new users
+        from .instant_setup import ensure_instant_tts, get_instant_model, is_model_cached
+
+        # If using default VITS model but it's not cached, use instant setup
+        if tts_model == "tts_models/en/ljspeech/vits" and not is_model_cached(tts_model):
+            if debug_mode:
+                print("🚀 First-time setup: ensuring instant TTS availability...")
+
+            # Try instant setup with lightweight model
+            if ensure_instant_tts():
+                tts_model = get_instant_model()  # Use fast_pitch instead
+                if debug_mode:
+                    print(f"✅ Using essential model: {tts_model}")
+
         # Initialize TTS engine using lazy import
         TTSEngine = _import_tts_engine()
         self.tts_engine = TTSEngine(
@@ -391,32 +429,68 @@ class VoiceManager:
         return self.speed
 
     def set_tts_model(self, model_name):
-        """Change the TTS model.
-        
+        """Change the TTS model safely without memory conflicts.
+
         Available models (all pure Python, cross-platform):
         - "tts_models/en/ljspeech/fast_pitch" (default, recommended)
         - "tts_models/en/ljspeech/glow-tts" (alternative)
         - "tts_models/en/ljspeech/tacotron2-DDC" (legacy)
-        
+
         Args:
             model_name: TTS model name to use
-            
+
         Returns:
             True if successful
-            
+
         Example:
             vm.set_tts_model("tts_models/en/ljspeech/glow-tts")
         """
         # Stop any current speech
         self.stop_speaking()
-        
+
+        # CRITICAL: Crash-safe cleanup of old TTS engine
+        if hasattr(self, 'tts_engine') and self.tts_engine:
+            try:
+                # Stop all audio and cleanup player
+                if hasattr(self.tts_engine, 'audio_player') and self.tts_engine.audio_player:
+                    # Try stop method if available
+                    if hasattr(self.tts_engine.audio_player, 'stop'):
+                        self.tts_engine.audio_player.stop()
+                    self.tts_engine.audio_player.cleanup()
+
+                # Force cleanup of TTS object and release GPU memory
+                if hasattr(self.tts_engine, 'tts') and self.tts_engine.tts:
+                    # Clear CUDA cache if using GPU
+                    try:
+                        import torch
+                        if torch.cuda.is_available():
+                            torch.cuda.empty_cache()
+                    except:
+                        pass
+
+                    del self.tts_engine.tts
+
+                # Clear the engine itself
+                del self.tts_engine
+                self.tts_engine = None
+
+                # Force garbage collection to prevent memory leaks
+                import gc
+                gc.collect()
+
+            except Exception as e:
+                if self.debug_mode:
+                    print(f"Warning: TTS cleanup issue: {e}")
+                # Force clear even if cleanup failed
+                self.tts_engine = None
+
         # Reinitialize TTS engine with new model using lazy import
         TTSEngine = _import_tts_engine()
         self.tts_engine = TTSEngine(
             model_name=model_name,
             debug_mode=self.debug_mode
         )
-        
+
         # Restore callbacks
         self.tts_engine.on_playback_start = self._on_tts_start
         self.tts_engine.on_playback_end = self._on_tts_end
@@ -474,13 +548,37 @@ class VoiceManager:
 
         # Select best model for this language
         selected_model = self._select_best_model(language)
-        models_to_try = [selected_model, self.SAFE_FALLBACK]
+
+        # CRITICAL FIX: Check if model is available, download if not
+        from .instant_setup import is_model_cached
+        from .simple_model_manager import download_model
+
+        if not is_model_cached(selected_model):
+            if self.debug_mode:
+                print(f"📥 Model {selected_model} not cached, downloading...")
+
+            # Try to download the model
+            success = download_model(selected_model)
+            if not success:
+                if self.debug_mode:
+                    print(f"❌ Failed to download {selected_model}")
+                # If download fails and it's not English, we have a problem
+                if language != 'en':
+                    print(f"❌ Cannot switch to {self.LANGUAGES[language]['name']}: Model download failed")
+                    print(f"   Try: abstractvoice download-models --language {language}")
+                    return False
+
+        models_to_try = [selected_model]
+
+        # Only add fallback if it's different from selected
+        if selected_model != self.SAFE_FALLBACK:
+            models_to_try.append(self.SAFE_FALLBACK)
 
         for model_name in models_to_try:
             try:
                 if self.debug_mode:
                     lang_name = self.LANGUAGES[language]['name']
-                    print(f"🌍 Switching to {lang_name} voice: {model_name}")
+                    print(f"🌍 Loading {lang_name} voice: {model_name}")
 
                 # Reinitialize TTS engine
                 TTSEngine = _import_tts_engine()
@@ -505,12 +603,16 @@ class VoiceManager:
 
             except Exception as e:
                 if self.debug_mode:
-                    print(f"⚠️ Model {model_name} failed: {e}")
+                    print(f"⚠️ Model {model_name} failed to load: {e}")
+                # Don't silently continue - report the failure
+                if model_name == selected_model and language != 'en':
+                    print(f"❌ Failed to load {lang_name} voice model")
+                    print(f"   The model might be corrupted. Try:")
+                    print(f"   abstractvoice download-models --language {language}")
                 continue
 
         # All models failed
-        if self.debug_mode:
-            print(f"❌ All models failed for language '{language}'")
+        print(f"❌ Cannot switch to {self.LANGUAGES[language]['name']}: No working models")
         return False
 
     def get_language(self):
@@ -782,15 +884,29 @@ class VoiceManager:
             return False
 
         voice_info = self.VOICE_CATALOG[language][voice_id]
+        model_name = voice_info['model']
+
+        # CRITICAL FIX: Download model if not cached
+        from .instant_setup import is_model_cached
+        from .simple_model_manager import download_model
 
-        # Check compatibility
-        if voice_info['requires'] == 'espeak-ng' and not self._test_model_compatibility(voice_info['model']):
+        if not is_model_cached(model_name):
+            print(f"📥 Voice model '{voice_id}' not cached, downloading...")
+            success = download_model(model_name)
+            if not success:
+                print(f"❌ Failed to download voice '{voice_id}'")
+                print(f"   Check your internet connection and try again")
+                return False
+            print(f"✅ Voice model '{voice_id}' downloaded successfully")
+
+        # Check compatibility after download
+        if voice_info['requires'] == 'espeak-ng' and not self._test_model_compatibility(model_name):
             if self.debug_mode:
                 print(f"⚠️ Voice '{voice_id}' requires espeak-ng. Install it for premium quality.")
-            return False
+            # Don't fail - try to load anyway
+            # return False
 
         # Set the specific voice
-        model_name = voice_info['model']
         if self.debug_mode:
             print(f"🎭 Setting {language} voice to: {voice_id}")
             print(f"   Model: {model_name}")

{abstractvoice-0.4.1.dist-info → abstractvoice-0.5.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: abstractvoice
-Version: 0.4.1
+Version: 0.5.0
 Summary: A modular Python library for voice interactions with AI systems
 Author-email: Laurent-Philippe Albou <contact@abstractcore.ai>
 License-Expression: MIT
@@ -19,6 +19,14 @@ Description-Content-Type: text/markdown
 License-File: LICENSE
 Requires-Dist: numpy>=1.24.0
 Requires-Dist: requests>=2.31.0
+Requires-Dist: appdirs>=1.4.0
+Requires-Dist: coqui-tts<0.30.0,>=0.27.0
+Requires-Dist: torch<2.4.0,>=2.0.0
+Requires-Dist: torchvision<0.19.0,>=0.15.0
+Requires-Dist: torchaudio<2.4.0,>=2.0.0
+Requires-Dist: librosa>=0.10.0
+Requires-Dist: sounddevice>=0.4.6
+Requires-Dist: soundfile>=0.12.1
 Provides-Extra: voice
 Requires-Dist: sounddevice>=0.4.6; extra == "voice"
 Requires-Dist: webrtcvad>=2.0.10; extra == "voice"
@@ -164,38 +172,51 @@ AbstractVoice automatically detects espeak-ng and upgrades to premium quality vo
 
 ## Quick Start
 
-### ⚡ Instant TTS (v0.4.0+)
+### ⚡ Instant TTS (v0.5.0+)
 
 ```python
 from abstractvoice import VoiceManager
 
-# Initialize voice manager - automatically downloads essential model if needed
+# Initialize voice manager - works immediately with included dependencies
 vm = VoiceManager()
 
-# Text-to-speech works immediately!
+# Text-to-speech works right away!
 vm.speak("Hello! TTS works out of the box!")
+
+# Language switching with automatic model download
+vm.set_language('fr')
+vm.speak("Bonjour! Le français fonctionne aussi!")
 ```
 
-**That's it!** AbstractVoice v0.4.0+ automatically:
-- ✅ Downloads essential English model (107MB) on first use
-- ✅ Caches models permanently for offline use
-- ✅ Works immediately after first setup
+**That's it!** AbstractVoice v0.5.0+ automatically:
+- ✅ Includes essential TTS dependencies in base installation
+- ✅ Downloads models automatically when switching languages/voices
+- ✅ Works immediately after `pip install abstractvoice`
+- ✅ No silent failures - clear error messages if download fails
 - ✅ No complex configuration needed
 
-### 🌍 Multi-Language Support
+### 🌍 Multi-Language Support (Auto-Download in v0.5.0+)
 
 ```python
-# Download and use French voice
-vm.download_model('fr.css10_vits')  # Downloads automatically
+# Simply switch language - downloads model automatically if needed!
 vm.set_language('fr')
 vm.speak("Bonjour! Je parle français maintenant.")
 
-# Download and use German voice
-vm.download_model('de.thorsten_vits')
+# Switch to German - no manual download needed
 vm.set_language('de')
 vm.speak("Hallo! Ich spreche jetzt Deutsch.")
+
+# Spanish, Italian also supported
+vm.set_language('es')
+vm.speak("¡Hola! Hablo español ahora.")
+
+# If download fails, you'll get clear error messages with instructions
+# Example: "❌ Cannot switch to French: Model download failed"
+#          "   Try: abstractvoice download-models --language fr"
 ```
 
+**New in v0.5.0:** No more manual `download_model()` calls! Language switching handles downloads automatically.
+
 ### 🔧 Check System Status
 
 ```python
@@ -1363,20 +1384,22 @@ abstractvoice check-deps
 
 ### CLI Voice Commands
 
-In the CLI REPL, use these commands:
+In the CLI REPL, use these commands (v0.5.0+):
 
 ```bash
 # List all available voices with download status
 /setvoice
 
-# Download and set specific voice
-/setvoice fr.css10_vits      # French CSS10 VITS
-/setvoice de.thorsten_vits   # German Thorsten
-/setvoice it.mai_male_vits   # Italian Male
+# Automatically download and set specific voice (NEW in v0.5.0!)
+/setvoice fr.css10_vits      # Downloads French CSS10 if needed
+/setvoice de.thorsten_vits   # Downloads German Thorsten if needed
+/setvoice it.mai_male_vits   # Downloads Italian Male if needed
+/setvoice en.jenny           # Downloads Jenny voice if needed
 
-# Change language
-/language fr
-/language de
+# Change language (automatically downloads models if needed - NEW!)
+/language fr                 # Switches to French, downloads if needed
+/language de                 # Switches to German, downloads if needed
+/language es                 # Switches to Spanish, downloads if needed
 
 # Voice controls
 /pause                       # Pause current speech
@@ -1387,6 +1410,8 @@ In the CLI REPL, use these commands:
 /exit
 ```
 
+**New in v0.5.0:** Language and voice commands now automatically download missing models with progress indicators. No more silent failures!
+
 ## Perspectives
 
 This is a test project that I designed with examples to work with Ollama, but I will adapt the examples and abstractvoice to work with any LLM provider (anthropic, openai, etc).

{abstractvoice-0.4.1.dist-info → abstractvoice-0.5.0.dist-info}/RECORD

@@ -1,23 +1,23 @@
-abstractvoice/__init__.py,sha256=HZYSCQ-xztoj7gWr5dVLBsGh4AYrViTYe8-ze4b-ynY,1011
+abstractvoice/__init__.py,sha256=EM_gfVmMvSOGeYPfMd8-AFThLrKwWh8TN4JFqdn0deU,1011
 abstractvoice/__main__.py,sha256=e6jhoONg3uwwPUCdnr68bSRTT1RrpWy2DrOJ6ozMJVc,4775
 abstractvoice/dependency_check.py,sha256=BUUADz4un4_FCZzNpgwk1qpJ6yqVi5Pvjfd3JLS8hAI,10045
-abstractvoice/model_manager.py,sha256=hnN3PTaY109mjTjgBuOB8yfAYVlMpqtMVBljLASRUi4,14275
+abstractvoice/instant_setup.py,sha256=_Q8T6tcMSor--1XPlgdOya3lvC-VtClHz4FSgDOXFNI,2667
 abstractvoice/recognition.py,sha256=4KtDUDFixEYuBUMDH2fWaD9csKlwA9tqXkMAkyQMSMo,11259
-abstractvoice/simple_model_manager.py,sha256=DTvEBEPtfu9zJA6V3S8SaWQ_pDYFlK_SoOMlnnRjBtk,13801
-abstractvoice/voice_manager.py,sha256=n7QHZPR1LWh3RjEBQ3LVrBKoOr5zccc3soKah5CBrac,32584
+abstractvoice/simple_model_manager.py,sha256=DfmrF3t3-V6rIWs4IvJmB_0Cck0LBY3YMvMzHGsHg4Q,19423
+abstractvoice/voice_manager.py,sha256=iw8Lu4VhsSMk0HQYlUU2NWAQR2C4dDiCAsiFjaeldiU,37664
 abstractvoice/examples/__init__.py,sha256=94vpKJDlfOrEBIUETg-57Q5Z7fYDidg6v4UzV7V_lZA,60
 abstractvoice/examples/cli_repl.py,sha256=kIgvgrGfyejX8-VFeFhvAVqrp3X-s-K3Ul861aM4Bh8,44220
-abstractvoice/examples/voice_cli.py,sha256=SYnFkz9KWWTISLgS2beJzb2tzLoz4dXpHQBWpKgS0sc,11585
+abstractvoice/examples/voice_cli.py,sha256=VdgDT01wly8HjWF53t_hDLkJoZc9FWQq2I-nxcSIAp8,11592
 abstractvoice/examples/web_api.py,sha256=0g5LKJpl7fZepPQJL25AcdaevV-xv34VqqyWGYYchPk,6376
 abstractvoice/stt/__init__.py,sha256=PFc6la3tTkxT4TJYwb0PnMIahM_hFtU4pNQdeKmbooo,120
 abstractvoice/stt/transcriber.py,sha256=GdaH1OsCHu4Vu9rUsQlzH6X9bfcnoiK5tGz1AW_uj6Q,5481
 abstractvoice/tts/__init__.py,sha256=WgJrxqdc_qaRyfFt1jbgMQD9S757jYuBpDzMRB02TFs,122
-abstractvoice/tts/tts_engine.py,sha256=HstJMwxTbZJx87Q-CY4mWeKHKbj17DhdvDdlch3xUNQ,49725
+abstractvoice/tts/tts_engine.py,sha256=iTa9eBH9vPH8VR2qoJX6nmNi6yERvA4Uz1jPu4OXrTA,55074
 abstractvoice/vad/__init__.py,sha256=RIIbFw25jNHgel06E4VvTWJnXjwjeFZ98m1Vx9hVjuo,119
 abstractvoice/vad/voice_detector.py,sha256=ghrhpDFlIR5TsMB2gpigXY6t5c_1yZ7vEX1imAMgWjc,3166
-abstractvoice-0.4.1.dist-info/licenses/LICENSE,sha256=TiDPM5WcFRQPoC5e46jGMeMppZ-eu0eFx_HytjE49bk,1105
-abstractvoice-0.4.1.dist-info/METADATA,sha256=AN_KjRcI2ZaetOIuAb9JOf4dvFWJI7AckQpV25wq2tI,40713
-abstractvoice-0.4.1.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
-abstractvoice-0.4.1.dist-info/entry_points.txt,sha256=3bDX2dNOGvrsTx1wZ_o_hVgmM_a2zbcHc1ZkL154rN4,72
-abstractvoice-0.4.1.dist-info/top_level.txt,sha256=a1qyxqgF1O8cJtPKpcJuImGZ_uXqPNghbLZ9gp-UiOo,14
-abstractvoice-0.4.1.dist-info/RECORD,,
+abstractvoice-0.5.0.dist-info/licenses/LICENSE,sha256=TiDPM5WcFRQPoC5e46jGMeMppZ-eu0eFx_HytjE49bk,1105
+abstractvoice-0.5.0.dist-info/METADATA,sha256=cH2FfruHztr3vENO6zo93yuKHkf6KmqI0-C1GnHCYA0,42154
+abstractvoice-0.5.0.dist-info/WHEEL,sha256=_zCd3N1l69ArxyTb8rzEoP9TpbYXkqRFSNOD5OuxnTs,91
+abstractvoice-0.5.0.dist-info/entry_points.txt,sha256=3bDX2dNOGvrsTx1wZ_o_hVgmM_a2zbcHc1ZkL154rN4,72
+abstractvoice-0.5.0.dist-info/top_level.txt,sha256=a1qyxqgF1O8cJtPKpcJuImGZ_uXqPNghbLZ9gp-UiOo,14
+abstractvoice-0.5.0.dist-info/RECORD,,