`services.openai_client`#

Module Contents#

`AsyncOpenAIClient`	A wrapper around OpenAI’s Python async client for querying models
`OpenAIClient`	A wrapper around OpenAI’s Python client for querying models

class services.openai_client.AsyncOpenAIClient(async_openai_client: openai.AsyncOpenAI)#

A wrapper around OpenAI’s Python async client for querying models

Initialization

async query_model( *, messages: collections.abc.Iterable, model: str, conversation_formatter: nemo_curator.services.conversation_formatter.ConversationFormatter | None = None, max_tokens: int | None | openai._types.NotGiven = NOT_GIVEN, n: int | None | openai._types.NotGiven = NOT_GIVEN, seed: int | None | openai._types.NotGiven = NOT_GIVEN, stop: str | None | list[str] | openai._types.NotGiven = NOT_GIVEN, stream: bool | None | openai._types.NotGiven = False, temperature: float | None | openai._types.NotGiven = NOT_GIVEN, top_k: int | None = None, top_p: float | None | openai._types.NotGiven = NOT_GIVEN, ) → list[str]#

async query_reward_model( *, messages: collections.abc.Iterable, model: str, ) → dict#: Prompts an LLM Reward model to score a conversation between a user and assistant Args: messages: The conversation to calculate a score for. Should be formatted like: [{“role”: “user”, “content”: “Write a sentence”}, {“role”: “assistant”, “content”: “This is a sentence”}, …] model: The name of the model that should be used to calculate the reward. Must be a reward model, cannot be a regular LLM. Returns: A mapping of score_name -> score

class services.openai_client.OpenAIClient(openai_client: openai.OpenAI)#

A wrapper around OpenAI’s Python client for querying models

Initialization

query_model( *, messages: collections.abc.Iterable, model: str, conversation_formatter: nemo_curator.services.conversation_formatter.ConversationFormatter | None = None, max_tokens: int | None | openai._types.NotGiven = NOT_GIVEN, n: int | None | openai._types.NotGiven = NOT_GIVEN, seed: int | None | openai._types.NotGiven = NOT_GIVEN, stop: str | None | list[str] | openai._types.NotGiven = NOT_GIVEN, stream: bool | None | openai._types.NotGiven = False, temperature: float | None | openai._types.NotGiven = NOT_GIVEN, top_k: int | None = None, top_p: float | None | openai._types.NotGiven = NOT_GIVEN, ) → list[str]#

query_reward_model( *, messages: collections.abc.Iterable, model: str, ) → dict#: Prompts an LLM Reward model to score a conversation between a user and assistant Args: messages: The conversation to calculate a score for. Should be formatted like: [{“role”: “user”, “content”: “Write a sentence”}, {“role”: “assistant”, “content”: “This is a sentence”}, …] model: The name of the model that should be used to calculate the reward. Must be a reward model, cannot be a regular LLM. Returns: A mapping of score_name -> score