flair.datasets.sequence_labeling.NER_NOISEBENCH#

class flair.datasets.sequence_labeling.NER_NOISEBENCH(noise='clean', base_path=None, in_memory=True, **corpusargs)View on GitHub #

__init__(noise='clean', base_path=None, in_memory=True, **corpusargs)View on GitHub #

Initialize the NoiseBench corpus.

Parameters:

noise (string) – Chooses the labelset for the data. clean (default): Clean labels crowd,crowdbest,expert,distant,weak,llm : Different kinds of noisy labelsets (details: …)
base_path (Optional[Union[str, Path]]) – Path to the data. Default is None, meaning the corpus gets automatically downloaded and saved. You can override this by passing a path to a directory containing the unprocessed files but typically this should not be necessary.
in_memory (bool) – If True the dataset is kept in memory achieving speedups in training.
**corpusargs – The arguments propagated to :meth:’flair.datasets.ColumnCorpus.__init__’.

Methods

`__init__`([noise, base_path, in_memory])	Initialize the NoiseBench corpus.
`add_label_noise`(label_type, labels[, ...])	Generates uniform label noise distribution in the chosen dataset split.
`downsample`([percentage, downsample_train, ...])	Randomly downsample the corpus to the given percentage (by removing data points).
`filter_empty_sentences`()	A method that filters all sentences consisting of 0 tokens.
`filter_long_sentences`(max_charlength)	A method that filters all sentences for which the plain text is longer than a specified number of characters.
`get_all_sentences`()	Returns all sentences (spanning all three splits) in the `Corpus`.
`get_label_distribution`()	Counts occurrences of each label in the corpus and returns them as a dictionary object.
`make_label_dictionary`(label_type[, ...])	Creates a dictionary of all labels assigned to the sentences in the corpus.
`make_tag_dictionary`(tag_type)	Create a tag dictionary of a given label type.
`make_vocab_dictionary`([max_tokens, min_freq])	Creates a `Dictionary` of all tokens contained in the corpus.
`obtain_statistics`([label_type, pretty_print])	Print statistics about the corpus, including the length of the sentences and the labels in the corpus.

Attributes

`SAVE_TRAINDEV_FILE`
`dev`	The dev split as a `torch.utils.data.Dataset` object.
`label_url`
`test`	The test split as a `torch.utils.data.Dataset` object.
`train`	The training split as a `torch.utils.data.Dataset` object.

label_url = 'https://raw.githubusercontent.com/elenamer/NoiseBench/main/data/annotations/'#