inworld: removal of unnecessary default assignment since already done

Co-authored-by: Mark Backman <m.backman@gmail.com>

src/pipecat/services/inworld/tts.py

@@ -227,7 +227,7 @@ class InworldTTSService(TTSService):
         # This will be sent as JSON payload in each TTS request
         # Note: Language is automatically inferred from text by Inworld's models
         self._settings = {
-            "voiceId": params.voice_id or "Ashley",  # Voice selection from params
+            "voiceId": params.voice_id,  # Voice selection from params
             "modelId": params.model or "inworld-tts-1",  # TTS model selection from params
             "audio_config": {  # Audio format configuration
                 "audio_encoding": encoding,  # Format: LINEAR16, MP3, etc.