feat: Add Lyria enum for music generation mode for vocalization

PiperOrigin-RevId: 789847872

Author: matthew29tang

google/genai/_live_converters.py

@@ -1098,6 +1098,13 @@ def _LiveMusicGenerationConfig_to_mldev(
         getv(from_object, ['only_bass_and_drums']),
     )
 
+  if getv(from_object, ['music_generation_mode']) is not None:
+    setv(
+        to_object,
+        ['musicGenerationMode'],
+        getv(from_object, ['music_generation_mode']),
+    )
+
   return to_object
 
 
@@ -2871,6 +2878,13 @@ def _LiveMusicGenerationConfig_from_mldev(
         getv(from_object, ['onlyBassAndDrums']),
     )
 
+  if getv(from_object, ['musicGenerationMode']) is not None:
+    setv(
+        to_object,
+        ['music_generation_mode'],
+        getv(from_object, ['musicGenerationMode']),
+    )
+
   return to_object
 
 

google/genai/tests/live/test_live_music.py

@@ -186,9 +186,19 @@ async def test_async_session_send_config(
   )
   if vertexai:
     with pytest.raises(NotImplementedError):
-      await session.set_music_generation_config(config=types.LiveMusicGenerationConfig(bpm=140))
+      await session.set_music_generation_config(
+          config=types.LiveMusicGenerationConfig(
+              bpm=140,
+              music_generation_mode=types.MusicGenerationMode.VOCALIZATION,
+          )
+      )
     return
-  await session.set_music_generation_config(config=types.LiveMusicGenerationConfig(bpm=140))
+  await session.set_music_generation_config(
+      config=types.LiveMusicGenerationConfig(
+          bpm=140,
+          music_generation_mode=types.MusicGenerationMode.VOCALIZATION,
+      )
+  )
   mock_websocket.send.assert_called_once()
   sent_data = json.loads(mock_websocket.send.call_args[0][0])
   assert 'musicGenerationConfig' in sent_data

google/genai/types.py

@@ -672,6 +672,22 @@ class Scale(_common.CaseInSensitiveEnum):
   """B major or Ab minor."""
 
 
+class MusicGenerationMode(_common.CaseInSensitiveEnum):
+  """The mode of music generation."""
+
+  MUSIC_GENERATION_MODE_UNSPECIFIED = 'MUSIC_GENERATION_MODE_UNSPECIFIED'
+  """Rely on the server default generation mode."""
+  QUALITY = 'QUALITY'
+  """Steer text prompts to regions of latent space with higher quality
+      music."""
+  DIVERSITY = 'DIVERSITY'
+  """Steer text prompts to regions of latent space with a larger
+      diversity of music."""
+  VOCALIZATION = 'VOCALIZATION'
+  """Steer text prompts to regions of latent space more likely to
+      generate music with vocals."""
+
+
 class LiveMusicPlaybackControl(_common.CaseInSensitiveEnum):
   """The playback control signal to apply to the music generation."""
 
@@ -13332,6 +13348,10 @@ class LiveMusicGenerationConfig(_common.BaseModel):
       default=None,
       description="""Whether the audio output should contain only bass and drums.""",
   )
+  music_generation_mode: Optional[MusicGenerationMode] = Field(
+      default=None,
+      description="""The mode of music generation. Default mode is QUALITY.""",
+  )
 
 
 class LiveMusicGenerationConfigDict(TypedDict, total=False):
@@ -13375,6 +13395,9 @@ class LiveMusicGenerationConfigDict(TypedDict, total=False):
   only_bass_and_drums: Optional[bool]
   """Whether the audio output should contain only bass and drums."""
 
+  music_generation_mode: Optional[MusicGenerationMode]
+  """The mode of music generation. Default mode is QUALITY."""
+
 
 LiveMusicGenerationConfigOrDict = Union[
     LiveMusicGenerationConfig, LiveMusicGenerationConfigDict