Merge pull request #3667 from ianbbqzy/ian/fix-auto-mode-space

[inworld] aggregate_sentence mode needs trailing space

changelog/3667.fixed.md

@@ -0,0 +1 @@
+- Fixed an issue in `InworldTTSService` where punctuation was pronounced. Now, the `InworldTTSService` ensures proper spacing between sentences, resolving pronunciation issues.

src/pipecat/services/inworld/tts.py

@@ -439,6 +439,7 @@ class InworldTTSService(AudioContextWordTTSService):
         encoding: str = "LINEAR16",
         params: InputParams = None,
         aggregate_sentences: bool = True,
+        append_trailing_space: bool = True,
         **kwargs: Any,
     ):
         """Initialize the Inworld WebSocket TTS service.
@@ -452,6 +453,7 @@ class InworldTTSService(AudioContextWordTTSService):
             encoding: Audio encoding format.
             params: Input parameters for Inworld WebSocket TTS configuration.
             aggregate_sentences: Whether to aggregate sentences before synthesis.
+            append_trailing_space: Whether to append a trailing space to text before sending to TTS.
             **kwargs: Additional arguments passed to the parent class.
         """
         super().__init__(
@@ -460,6 +462,7 @@ class InworldTTSService(AudioContextWordTTSService):
             pause_frame_processing=True,
             sample_rate=sample_rate,
             aggregate_sentences=aggregate_sentences,
+            append_trailing_space=append_trailing_space,
             **kwargs,
         )