Gets the list of labels for this data set.

Gets the list of labels for this data set.

Gets a collection of `InputExample`s for prediction.

Gets a collection of `InputExample`s for the train set.

:param sentence_representation: [batch_size,num_sentences,hidden_size*2] :return:forward hidden state: a list,length is num_sentences

:param sentence_representation: [batch_size,num_sentences,hidden_size*2] :return:forward hidden state: a list,length is num_sentences

:param embedded_words:[batch_size*num_sentences,sentence_length,embed_size] :return: backward hidden state:a list.length is sentenc

:param embedded_words:[batch_size*num_sentences,sentence_length,embed_size] :return: backward hidden state:a list.length is sentenc

:param sentence_representation: [batch_size,num_sentences,hidden_size*2] :return:forward hidden state: a list,length is num_sentences

:param sentence_representation: [batch_size,num_sentences,hidden_size*2] :return:forward hidden state: a list,length is num_sentences

:param embedded_words:[batch_size*num_sentences,sentence_length,embed_size] :return:forward hidden state: a list.length is sentence_l

:param embedded_words:[batch_size*num_sentences,sentence_length,embed_size] :return:forward hidden state: a list.length is sentence_l

main computation graph here: 1.Word Encoder. 2.Word Attention. 3.Sentence Encoder 4.Sentence Attention 5.linear classifier

main computation graph here: 1.Word Encoder. 2.Word Attention. 3.Sentence Encoder 4.Sentence Attention 5.linear classifier

main computation graph here: 1.Word Encoder. 2.Word Attention. 3.Sentence Encoder 4.Sentence Attention 5.dropout 6.transform for each task 7.linear cl

main computation graph here: a.Input Module,b.Question Module,c.Episodic Memory Module,d.Answer Module

building blocks: encoder:6 layers.each layers has two sub-layers. the first is multi-head self-attention mechanism; the second is position

building blocks: encoder:6 layers.each layers has two sub-layers. the first is multi-head self-attention mechanism; the second is position

main computation graph here: 1.embedding-->2.average-->3.linear classifier

main computation graph here: 1.embedding-->2.average-->3.linear classifier

main computation graph here: 1. embeddding layers, 2.convolutional layer, 3.max-pooling, 4.softmax layer.

main computation graph here: 1.embedding-->2.CONV-BN-RELU-MAX_POOLING-->3.linear classifier

main computation graph here: 1.embedding-->2.CONV-RELU-MAX_POOLING-->3.linear classifier

main computation graph here: 1. embeddding layer, 2.Bi-LSTM layer, 3.mean pooling, 4.FC layer, 5.softmax

main computation graph here: 1. embeddding layer, 2.Bi-LSTM layer-->dropout,3.LSTM layer-->dropout 4.FC layer 6.softmax layer

main computation graph here: 1. embeddding layer, 2.Bi-LSTM layer, 3.concat, 4.FC layer 5.softmax

main computation graph here: 1.input encoder 2.dynamic emeory 3.output layer

main computation graph here: #1.Word embedding. 2.Encoder with GRU 3.Decoder using GRU(optional with attention).

main computation graph here: 1. embeddding layer, 2.Bi-LSTM layer, 3.max pooling, 4.FC layer 5.softmax

main computation graph here: 1. embeddding layer, 2.Bi-LSTM layer, 3.max pooling, 4.FC layer 5.softmax

main computation graph here: 1.embedding-->2.average-->3.linear classifier

main computation graph here: 1. embeddding layer, 2.Bi-LSTM layer, 3.max pooling, 4.FC layer 5.softmax

use bi-directional lstm to encode query_embedding:[batch_size,sequence_length,embed_size] and story_embeddin

encode raw texts into vector representation

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

define all weights here

Runs layer normalization followed by dropout.

x should be:[batch_size,sequence_length,d_model] :return:[batch_size,sequence_length,d_model]

x should be:[batch_size,sequence_length,d_model] :return:

load data from h5py and pickle cache files, which is generate by take step by step of pre-processing.ipynb :param cache_file_h5py: :param

load data from h5py and pickle cache files, which is generate by take step by step of pre-processing.ipynb :param cache_file_h5py: :param

load data from h5py and pickle cache files, which is generate by take step by step of pre-processing.ipynb :param cache_file_h5py: :param

load data from h5py and pickle cache files, which is generate by take step by step of pre-processing.ipynb :param cache_file_h5py: :param

input: a file path :return: train, test, valid. where train=(trainX, trainY). where trainX: is a list of list.each list repre

input: a file path :return: train, test, valid. where train=(trainX, trainY). where trainX: is a list of list.each list re

input: a file path :return: train, test, valid. where train=(trainX, trainY). where trainX: is a list of list.each list repre

Loads a vocabulary file into a dictionary.

calculate loss using (NCE)cross entropy here

calculate loss using (NCE)cross entropy here