Given the name of a package on PyPI and a version (both strings), checks if the given version is the latest version of the package available.

Applies the chu-liu-edmonds algorithm recursively to a graph with edge weights defined by score_matrix. Note that this function operates

r""" The implementation of `Cocke-Kasami-Younger`_ (CKY) algorithm to parse constituency trees. References: - Yu Zhang, Houquan Zhou

Closes the file handler.

Close this dataloader and terminates internal processes and queue. It's recommended to call this method to ensure a program can gracefully

Compacts a string of json data by collapsing whitespace after the specified indent level NOTE: will not produce correct results when inde

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

r""" Coreference resolution is the task of clustering mentions in text that refer to the same underlying real world entities. Args:

Count number of sentences in this document. Returns: Number of sentences.

Create dataset related values,

Computes the binary scores of tag sequences. Args: tag_indices: A [batch_size, max_seq_len] matrix of tag indices. sequence_lengths:

Computes backward decoding in a linear-chain CRF. Args: inputs: A [batch_size, num_tags] matrix of backpointer of next step (in time or

Computes forward decoding in a linear-chain CRF. Args: inputs: A [batch_size, num_tags] matrix of unary potentials. state: A [batch_s

Computes the alpha values in a linear-chain CRF. See http://www.cs.columbia.edu/~mcollins/fb.pdf for reference. Args: inputs: A [b

Computes the log-likelihood of tag sequences in a CRF. Args: inputs: A [batch_size, max_seq_len, num_tags] tensor of unary potentials t

Computes the unnormalized score for a tag sequence. Args: inputs: A [batch_size, max_seq_len, num_tags] tensor of unary potentials to u

Computes the unary scores of tag sequences. Args: tag_indices: A [batch_size, max_seq_len] matrix of tag indices. sequence_lengths: A

adapted from arXiv:1911.07471

An iterator over CONLL formatted files which yields documents, regardless of the number of document annotations in a particular file.

An iterator over CONLL formatted files which yields documents, regardless of the number of document annotations in a particular file. This is

An iterator returning file_paths in a directory containing CONLL-formatted files.

r""" Args: s_arc (~torch.Tensor): ``[batch_size, seq_len, seq_len]``. Scores of all possible arcs. s_s

r""" Args: s_arc (~torch.Tensor): ``[batch_size, seq_len, seq_len]``. Scores of all possible arcs. mask (~torch.BoolTe

Note: Counter to typical intuition, this function decodes the _maximum_ spanning tree. Decode the optimal MST tree with the Chu-Liu-Edmon

If you actually passed gradient-tracking Tensors to a Metric, there will be a huge memory leak, because it will prevent garbage colle

If you actually passed gradient-tracking Tensors to a Metric, there will be a huge memory leak, because it will prevent garbage colle

r""" Second-order Eisner algorithm for projective decoding. This is an extension of the first-order one that further incorporates sibling scor

Embedding sentences (list of words) with contextualized string embedding Args: texts: List of words, not chars texts: Lis

Checks if a chunk ended between the previous and current word. Args: prev_tag: previous chunk tag. tag: current chunk tag. prev

Evaluate on a dataloader. Args: data: Dataloader which can build from any data source. criterion: Loss function.

Evalute a field of a record using the type of the field_spec as a guide. Args: record: field_spec: Returns:

Args: input: input is the sentence lengths tensor. exclusive: exclude the last sentence length (Default value = True)

Implement this to run training loop. Args: trn: Training set. dev: Development set. epochs: Number of epo