Useful for performing evaluation on the slow weights (which typically generalize better)

Reads a tab separated value file.

update parameters and rate :return:

适合bert类型模型，适合pytorch_transformer模块 :param state: :param current: :return:

计算损失 参数: logits: [B, L, out_size] targets: [B, L] lengths: [B]

计算双向LSTM-CRF模型的损失 该损失函数的计算可以参考:https://arxiv.org/pdf/1603.01360.pdf

Calculate Chebyshev polynomials up to order k. Return a list of sparse matrices (tuple representation).

batch should be a list of (sequence, target, length) tuples... Returns a padded tensor of sequences sorted from longest to shortest,

batch should be a list of (sequence, target, length) tuples... Returns a padded tensor of sequences sorted from longest to shortest,

Convert labeled data export from doccano for classification task.

example: { title prompt content result_list }

Loads a data file into a list of `InputBatch`s `cls_token_at_end` define the location of the CLS token: - False (Default, BERT/XL

Convert labeled data export from doccano for extraction and aspect-level classification task.

:return: number of days in the dataloader

正常模型 :param state: 需要保存的信息 :param current: 当前判断指标 :return:

Compute the conditional log likelihood of a sequence of tags given emission scores. Args: emissions (`~torch.Tensor`): Emission sc

input: [N, C] target: [N, ]

Create a schedule with a constant learning rate.

Create a schedule with a constant learning rate preceded by a warmup period during which the learning rate increases linearly between 0 and 1.

Create a schedule with a learning rate that decreases following the values of the cosine function between 0 and `pi * cycles` after a warmup

Create a schedule with a learning rate that decreases following the values of the cosine function with several hard restarts, after a warmup

Gets a collection of `InputExample`s for the dev set.

See base class.

See base class.

See base class.

Gets the list of labels for this data set.

See base class.

See base class.

See base class.

See base class.

See base class.