feat: Granular Per Agent Speech Configuration

Merge https://github.com/google/adk-python/pull/3170

Addresses Feature Request: #3116

This PR adds a `speech_config` to the **LLM Agent configuration** for the **live use case**. When an **asynchronous LLM** call is made to the **Gemini Live API**, it prioritizes the most specific agent configuration's speech_config. If that is null, it then uses the run configuration's speech_config. Unit tests have been added to verify this behavior.

COPYBARA_INTEGRATE_REVIEW=https://github.com/google/adk-python/pull/3170 from qyuo:bidi_agent_speech_config af1bd277d4f95c4a7d9aa0b16828ba3de826ce08
PiperOrigin-RevId: 822305427

src/google/adk/agents/run_config.py

@@ -35,7 +35,10 @@ class StreamingMode(Enum):
 
 
 class RunConfig(BaseModel):
-  """Configs for runtime behavior of agents."""
+  """Configs for runtime behavior of agents.
+
+  The configs here will be overriden by agent-specific configurations.
+  """
 
   model_config = ConfigDict(
       extra='forbid',

src/google/adk/models/google_llm.py

@@ -60,6 +60,8 @@ class Gemini(BaseLlm):
 
   model: str = 'gemini-2.5-flash'
 
+  speech_config: Optional[types.SpeechConfig] = None
+
   retry_options: Optional[types.HttpRetryOptions] = None
   """Allow Gemini to retry failed responses.
 
@@ -269,6 +271,9 @@ class Gemini(BaseLlm):
           self._live_api_version
       )
 
+    if self.speech_config is not None:
+      llm_request.live_connect_config.speech_config = self.speech_config
+
     llm_request.live_connect_config.system_instruction = types.Content(
         role='system',
         parts=[