neuro-simulator 0.4.4__py3-none-any.whl → 0.5.0__py3-none-any.whl

neuro_simulator/services/audience.py

@@ -1,102 +0,0 @@
-# neuro_simulator/services/audience.py
-import logging
-
-from google import genai
-from google.genai import types
-from openai import AsyncOpenAI
-
-from ..core.config import config_manager, AppSettings
-from ..utils.state import app_state
-
-logger = logging.getLogger(__name__.replace("neuro_simulator", "server", 1))
-
-class AudienceLLMClient:
-    async def generate_chat_messages(self, prompt: str, max_tokens: int) -> str:
-        raise NotImplementedError
-
-class GeminiAudienceLLM(AudienceLLMClient):
-    def __init__(self, api_key: str, model_name: str):
-        if not api_key:
-            raise ValueError("Gemini API Key is not provided for GeminiAudienceLLM.")
-        self.client = genai.Client(api_key=api_key)
-        self.model_name = model_name
-        logger.info(f"Initialized GeminiAudienceLLM (new SDK), model: {self.model_name}")
-
-    async def generate_chat_messages(self, prompt: str, max_tokens: int) -> str:
-        response = await self.client.aio.models.generate_content(
-            model=self.model_name,
-            contents=prompt,
-            config=types.GenerateContentConfig(
-                temperature=config_manager.settings.audience_simulation.llm_temperature,
-                max_output_tokens=max_tokens
-            )
-        )
-        raw_chat_text = ""
-        if hasattr(response, 'text') and response.text:
-            raw_chat_text = response.text
-        elif response.candidates and response.candidates[0].content and response.candidates[0].content.parts:
-            for part in response.candidates[0].content.parts:
-                if hasattr(part, 'text') and part.text:
-                    raw_chat_text += part.text
-        return raw_chat_text
-
-class OpenAIAudienceLLM(AudienceLLMClient):
-    def __init__(self, api_key: str, model_name: str, base_url: str | None):
-        if not api_key:
-            raise ValueError("OpenAI API Key is not provided for OpenAIAudienceLLM.")
-        self.client = AsyncOpenAI(api_key=api_key, base_url=base_url)
-        self.model_name = model_name
-        logger.info(f"Initialized OpenAIAudienceLLM, model: {self.model_name}, API Base: {base_url}")
-
-    async def generate_chat_messages(self, prompt: str, max_tokens: int) -> str:
-        response = await self.client.chat.completions.create(
-            model=self.model_name,
-            messages=[{"role": "user", "content": prompt}],
-            temperature=config_manager.settings.audience_simulation.llm_temperature,
-            max_tokens=max_tokens,
-        )
-        if response.choices and response.choices[0].message and response.choices[0].message.content:
-            return response.choices[0].message.content.strip()
-        return ""
-
-async def get_dynamic_audience_prompt() -> str:
-    current_neuro_speech = ""
-    async with app_state.neuro_last_speech_lock:
-        current_neuro_speech = app_state.neuro_last_speech
-
-    prompt = config_manager.settings.audience_simulation.prompt_template.format(
-        neuro_speech=current_neuro_speech,
-        num_chats_to_generate=config_manager.settings.audience_simulation.chats_per_batch
-    )
-    return prompt
-
-class AudienceChatbotManager:
-    def __init__(self):
-        self.client: AudienceLLMClient = self._create_client(config_manager.settings)
-        self._last_checked_settings: dict = config_manager.settings.audience_simulation.model_dump()
-        logger.info("AudienceChatbotManager initialized.")
-
-    def _create_client(self, settings: AppSettings) -> AudienceLLMClient:
-        provider = settings.audience_simulation.llm_provider
-        logger.info(f"Creating new audience LLM client for provider: {provider}")
-        if provider.lower() == "gemini":
-            if not settings.api_keys.gemini_api_key:
-                raise ValueError("GEMINI_API_KEY not set in config")
-            return GeminiAudienceLLM(api_key=settings.api_keys.gemini_api_key, model_name=settings.audience_simulation.gemini_model)
-        elif provider.lower() == "openai":
-            if not settings.api_keys.openai_api_key:
-                raise ValueError("OPENAI_API_KEY not set in config")
-            return OpenAIAudienceLLM(api_key=settings.api_keys.openai_api_key, model_name=settings.audience_simulation.openai_model, base_url=settings.api_keys.openai_api_base_url)
-        else:
-            raise ValueError(f"Unsupported AUDIENCE_LLM_PROVIDER: {provider}")
-
-    def handle_config_update(self, new_settings: AppSettings):
-        new_audience_settings = new_settings.audience_simulation.model_dump()
-        if new_audience_settings != self._last_checked_settings:
-            logger.info("Audience simulation settings changed, re-initializing LLM client...")
-            try:
-                self.client = self._create_client(new_settings)
-                self._last_checked_settings = new_audience_settings
-                logger.info("LLM client hot-reloaded successfully.")
-            except Exception as e:
-                logger.error(f"Error hot-reloading LLM client: {e}", exc_info=True)