diff --git a/matrix_pygmalion_bot/bot/ai/langchain.py b/matrix_pygmalion_bot/bot/ai/langchain.py index c6a095b..fd69611 100644 --- a/matrix_pygmalion_bot/bot/ai/langchain.py +++ b/matrix_pygmalion_bot/bot/ai/langchain.py @@ -93,8 +93,8 @@ class AI(object): self.max_context = 2048 from ..wrappers.langchain_koboldcpp import KoboldCpp - self.llm_chat = KoboldCpp(temperature=self.bot.temperature, endpoint_url="http://172.16.85.10:5001/api/latest/generate", max_context=self.max_context, stop=['<|endoftext|>'], verbose=True) - self.llm_summary = KoboldCpp(temperature=0.7, repeat_penalty=1.15, top_k = 20, top_p= 0.9, endpoint_url="http://172.16.85.10:5001/api/latest/generate", max_context=self.max_context, stop=['<|endoftext|>'], max_tokens=512, verbose=True) + self.llm_chat = KoboldCpp(temperature=self.bot.temperature, endpoint_url="http://172.16.33.10:5001/api/latest/generate", max_context=self.max_context, stop=['<|endoftext|>'], verbose=True) + self.llm_summary = KoboldCpp(temperature=0.7, repeat_penalty=1.15, top_k = 20, top_p= 0.9, endpoint_url="http://172.16.33.10:5001/api/latest/generate", max_context=self.max_context, stop=['<|endoftext|>'], max_tokens=512, verbose=True) self.llm_chat_model = "pygmalion-7b" self.llm_summary_model = "vicuna-13b" self.text_wrapper = text_wrapper diff --git a/matrix_pygmalion_bot/bot/wrappers/langchain_koboldcpp.py b/matrix_pygmalion_bot/bot/wrappers/langchain_koboldcpp.py index 8eb3b9d..2b79731 100644 --- a/matrix_pygmalion_bot/bot/wrappers/langchain_koboldcpp.py +++ b/matrix_pygmalion_bot/bot/wrappers/langchain_koboldcpp.py @@ -18,7 +18,7 @@ logger = logging.getLogger(__name__) class KoboldCpp(LLM): """KoboldCpp LLM wrapper for testing purposes.""" - endpoint_url: str = "http://172.16.85.10:5001/api/latest/generate" + endpoint_url: str = "http://172.16.33.10:5001/api/latest/generate" temperature: Optional[float] = 0.7 """The temperature to use for sampling."""