Adopted from Timothy Dozat https://github.com/tdozat/Parser/blob/master/lib/models/nn.py Args: parse_probs(NDArray): seq_len x seq_len, the

Adopted from Timothy Dozat https://github.com/tdozat/Parser/blob/master/lib/models/nn.py Args: parse_probs(NDArray): seq_len x seq_len, the

Text style transfer aims to change the style of the input text to the target style while preserving its content. @param text Source text. @par

Split a document into sentences and tokenize them. @param text A document. @param coarse Whether to perform coarse-grained or fine-grained tokeniza

A set of all tokens in this vocab.

This method is overrode to honor the zero indexed token used in custom dict. Although for a tokenizer, cls_is_bos = sep_is_eos = True

Translate tags for each annotation. This is an inplace operation. .. Attention:: Note that the translated document might not print w

convert a type object to class path in str format Args: type_object: type Returns: class path

Given a tensor of embeddings with shape (d1 * ... * dn, sequence_length, embedding_dim) and the original shape (d1, ..., dn, sequence_leng

The index of ``UNK`` token.

Unlock each vocab.