Move system_instruction into LLMSettings

Add `system_instruction` field to `LLMSettings` so it is runtime-updatable via settings.
For Google (GoogleLLMService, GoogleVertexLLMService), deprecate the init-time arg since it was already shipped. For Anthropic, AWS Bedrock, and OpenAI, remove the init-time arg entirely since it was never shipped.

Add system instruction prepend logic to `build_chat_completion_params` overrides in Cerebras, SambaNova, Fireworks, Mistral, and Perplexity, which build params from scratch rather than calling `super()`.

Still need to handle realtime services (OpenAI Realtime, Grok Realtime, Gemini Live).

examples/foundational/21-tavus-transport.py

@@ -24,7 +24,7 @@ from pipecat.processors.aggregators.llm_response_universal import (
 )
 from pipecat.services.cartesia.tts import CartesiaTTSService, CartesiaTTSSettings
 from pipecat.services.deepgram.stt import DeepgramSTTService
-from pipecat.services.google.llm import GoogleLLMService
+from pipecat.services.google.llm import GoogleLLMService, GoogleLLMSettings
 from pipecat.transports.tavus.transport import TavusParams, TavusTransport
 
 load_dotenv(override=True)
@@ -58,7 +58,9 @@ async def main():
 
         llm = GoogleLLMService(
             api_key=os.getenv("GOOGLE_API_KEY"),
-            system_instruction="You are a helpful LLM in a WebRTC call. Your goal is to demonstrate your capabilities in a succinct way. Your output will be spoken aloud, so avoid special characters that can't easily be spoken, such as emojis or bullet points. Respond to what the user said in a creative and helpful way.",
+            settings=GoogleLLMSettings(
+                system_instruction="You are a helpful LLM in a WebRTC call. Your goal is to demonstrate your capabilities in a succinct way. Your output will be spoken aloud, so avoid special characters that can't easily be spoken, such as emojis or bullet points. Respond to what the user said in a creative and helpful way.",
+            ),
         )
 
         context = LLMContext()