langml.baselines.contrastive.simcse.model

Module Contents

Classes

SimCSE

Functions

simcse_loss(y_true, y_pred)

langml.baselines.contrastive.simcse.model.simcse_loss(y_true, y_pred)

class langml.baselines.contrastive.simcse.model.SimCSE(config_path: str, ckpt_path: str, params: langml.baselines.Parameters, backbone: str = 'roberta')

Bases: langml.baselines.BaselineModel

get_pooling_output(self, model: langml.tensor_typing.Models, output_index: int, pooling_strategy: str = 'cls') → langml.tensor_typing.Tensors

get pooling output :param model: keras.Model, BERT model :param output_index: int, specify output index of feedforward layer. :param pooling_strategy: str, specify pooling strategy from [‘cls’, ‘first-last-avg’, ‘last-avg’], default cls

build_model(self, pooling_strategy: str = 'cls', lazy_restore: bool = False) → langml.tensor_typing.Models