generator for line :param fname: :return:

inputs: (batch_size, seq_len, emb_dim)

write the data

resize

Reshape the last dimension of inpunt tensor x so that it becomes two dimensions and then transpose. Specifically, input a tensor with

Reshape the last dimension of inpunt tensor x so that it becomes two dimensions and then transpose. Specifically, input a tensor with

Strips accents from a piece of text.

Args: db (TYPE): NULL col (TYPE): NULL Returns: TODO Raises: NULL

convert text to vocab ids

Compute bleu and rouge scores.

get data batch

get data batch

Generate data for train, dev or test. Args: batch_size: int. The batch size of generated data. phase: string

data_generator

Generate data for train, dev or test. Args: examples: list. Train, dev or test data. batch_size: int. The ba

drop path op Args: input: tensor with arbitrary shape drop_prob: float number of drop path probability, default: 0.0

Args: Returns:

Args: Returns:

Obtain model loss or decoding output

f1_score

file_reader

Get the first feature of each sequence for classification

Get the first feature of each sequence for classification

gets the id of a token, returns the id of unk token if token is not in vocab Args: key: a string indicating the word

get_num_examples

get startup program.

Desc: 判断一个对象是否为空对象 Args: self : self Return: None Raise: None

len to mask

load data from disk Args: save_dir (TYPE): NULL Returns: TODO Raises: NULL

Desc: 计算region的mbr Args: self : self Return: [Point(minx, miny), Point(maxx, maxy)] Raise:

:param s: :return:

Pad the instances to the max sequence length in batch, and generate the corresponding position data and input mask.

parser char for text

parser char for text

Add word embeddings and position encodings. The output tensor has a shape of: [batch_size, max_src_length_in_batch, d_model]. This module

Add word embeddings and position encodings. The output tensor has a shape of: [batch_size, max_src_length_in_batch, d_model]. This module

Returns text encoded in a way suitable for print or `tf.logging`.

run frame, default train

TBD

doc

convert to LoDTensor

to tensor

tokenize the text

tokenize the text

update one instance

getitem

Desc: 计算指定点的grid, 可能为多个 Args: self : self x : x y : y grid_size : grid_size

init of class Args: cfg (TYPE): NULL

Registers 'flag_name' with 'default_value' and 'docstring'.

Registers 'flag_name' with 'default_value' and 'docstring'.

extract condition and/or