xinference.client.Client.describe_model#

Client.describe_model(model_uid: str)[fuente]#

Get model information via RESTful APIs.

Parámetros:

model_uid (str) – The unique id that identify the model.

Devuelve:

A dictionary containing the following keys:

  • »model_type»: str

    the type of the model determined by its function, e.g. «LLM» (Large Language Model)

  • »model_name»: str

    the name of the specific LLM model family

  • »model_lang»: List[str]

    the languages supported by the LLM model

  • »model_ability»: List[str]

    the ability or capabilities of the LLM model

  • »model_description»: str

    a detailed description of the LLM model

  • »model_format»: str

    the format specification of the LLM model

  • »model_size_in_billions»: int

    the size of the LLM model in billions

  • »quantization»: str

    the quantization applied to the model

  • »revision»: str

    the revision number of the LLM model specification

  • »context_length»: int

    the maximum text length the LLM model can accommodate (include all input & output)

Tipo del valor devuelto:

dict

Muestra:

RuntimeError – Report failure to get the wanted model with given model_uid. Provide details of failure through error message.