class Dict[str,] = None, use_language_specific_pos: bool = False, tokenizer: = None, lazy: bool = False)[source]


Reads a file in the conllu Universal Dependencies format.

token_indexersDict[str, TokenIndexer], optional (default=``{“tokens”: SingleIdTokenIndexer()}``)

The token indexers to be applied to the words TextField.

use_language_specific_posbool, optional (default = False)

Whether to use UD POS tags, or to use the language specific POS tags provided in the conllu format.

tokenizerTokenizer, optional, default = None

A tokenizer to use to split the text. This is useful when the tokens that you pass into the model need to have some particular attribute. Typically it is not necessary.

text_to_instance(self, words: List[str], upos_tags: List[str], dependencies: List[Tuple[str, int]] = None) →[source]
wordsList[str], required.

The words in the sentence to be encoded.

upos_tagsList[str], required.

The universal dependencies POS tags for each word.

dependenciesList[Tuple[str, int]], optional (default = None)

A list of (head tag, head index) tuples. Indices are 1 indexed, meaning an index of 0 corresponds to that word being the root of the dependency tree.

An instance containing words, upos tags, dependency head tags and head
indices as fields.