langspace.models package

Module contents

langspace.models.select_models(encoder: str = None, decoder: str = None, latent_dim: int = 0, annotations: List[str] = None, conditional: bool = False) → List[str][source]

Selects a list of LM-VAE models available from the neuro-symbolic-ai repository, according to the specified criteria.

Parameters:

  • encoder – The name of the encoder model (e.g., bert-base-cased, flan-t5-base)

  • decoder – The name of the decoder model (e.g., gpt2, Llama-3.2-3B)

  • latent_dim – The latent dimension of the LM-VAE (e.g., 64, 128)

  • annotations – Annotations the model was trained on (e.g., srl)

  • conditional – If it is a conditional variable model

Returns:

A list of available modes