Copy the first 6 columns from gold file to pred file Args: gold_file: pred_file: copied_pred_file: keep_comments: (De

Creates an optimizer with learning rate schedule. Args: init_lr: num_train_steps: num_warmup_steps: weight_decay_rate:

Decode the highest scoring sequence of tags in TensorFlow. This is a function for tensor. Args: potentials: A [batch_size, max_seq

Computes the normalization for a CRF. Args: inputs: A [batch_size, max_seq_len, num_tags] tensor of unary potentials to use as input to

Tokenize and prepare for the model a sequence or a pair of sequences. .. warning:: This method is deprecated, ``__call__

Given a sentence, return all token spans within the sentence. Spans are `inclusive`. Additionally, you can provide a maximum and minimum span

Feature extraction for windowed approaches See Also https://github.com/chqiwang/convseg/ Parameters ---------- sentence bigra

Generates a lemma rule to transform the source to the target Args: form: lemma: allow_copy: (Default value = False) Ret

Build sinusoidal embeddings. This matches the implementation in tensor2tensor, but differs slightly from the description in Section 3.

Get the precisest pos for this word. Args: main_pos: Use the main pos (cpos or upos) or the minor pos (pos or xpos).

Get subj/obj position sequence. Args: start_idx: end_idx: length: Returns:

Args: filepath: The path to pretrained embedding. vocab: The vocabulary from training set. extend_vocab: Unlock vocabula

r""" Checks if a dependency tree is projective. This also works for partial annotation. Besides the obvious crossing arcs, the examples b

r""" Checks if the arcs form an valid dependency tree. Args: sequence (list[int]): A list of head indices. proj (

Load a component from a ``meta.json`` (legacy TensorFlow component) or a ``config.json`` file. Args: save_dir: The identifier.

Load from path Args: path(str): file path Returns:

Finds the minimum edit script to transform the source to the target Args: source: target: allow_copy: (Default value = False

Use GPU nicely.

Called when config is ready, either during ``fit`` or ``load``. Subclass can perform extra initialization tasks in this callback. Arg

Keywords lookup which takes a piece of text as input, and lookup all occurrences of keywords in it. These occurrences can overlap with each ot

Creates a pipeline of components. It's made for bundling `KerasComponents`. For `TorchComponent`, use :class:`~hanlp.components.mtl.multi_task_lea

Args: source: dereify: Restore reverted relations remove_wiki: Returns:

Enumerate all possible tokenizations of a text. Args: text: A text. Returns: All possible tokenizations.