layers.py

# -*- coding: utf-8 -*-
"""
Created on Sun Dec 22 21:59:52 2019

@author: Deep
"""
import numpy as np
from keras.layers import *
from keras.activations import softmax
from keras.models import Model
from keras.optimizers import Nadam, Adam
from keras.regularizers import l2
import keras.backend as K

def create_pretrained_embedding(pretrained_weights_path, trainable=False, **kwargs):
    "Create embedding layer from a pretrained weights array"
    pretrained_weights = np.load(pretrained_weights_path)
    in_dim, out_dim = pretrained_weights.shape
    embedding = Embedding(in_dim, out_dim, weights=[pretrained_weights], trainable=False, **kwargs)
    return embedding

"""
双向LSTM 获取Char embedding
"""
def create_char_embedding(charsize, maxlen, max_char_len, char_embedding_dim):
    char_embedding =  Embedding(input_dim=charsize, output_dim=char_embedding_dim, \
                                embeddings_initializer='lecun_uniform', \
                                mask_zero=False)
         
#    char_embedding = Bidirectional(GRU(16, return_sequences=True), merge_mode="concat")
    return char_embedding

def soft_attention_alignment(input_1, input_2):
    "Align text representation with neural soft attention"
    attention = Dot(axes=-1)([input_1, input_2])
    w_att_1 = Lambda(lambda x: softmax(x, axis=1))(attention)
    w_att_2 = Permute((2,1))(Lambda(lambda x: softmax(x, axis=2))(attention))
    in1_aligned = Dot(axes=1)([w_att_1, input_1])
    in2_aligned = Dot(axes=1)([w_att_2, input_2])
    return in1_aligned, in2_aligned

def submult(input_1, input_2):
    "Get multiplication and subtraction then concatenate results"
    mult = Multiply()([input_1, input_2])
    sub = substract(input_1, input_2)
    out_= Concatenate()([sub, mult])
    return out_

def substract(input_1, input_2):
    "Substract element-wise"
    neg_input_2 = Lambda(lambda x: -x)(input_2)
    out_ = Add()([input_1, neg_input_2])
    return out_

def apply_multiple(input_, layers):
    "Apply layers to input then concatenate result"
    if not len(layers) > 1:
        raise ValueError('Layers list should contain more than 1 layer')
    else:
        agg_ = []
        for layer in layers:
            agg_.append(layer(input_))
        out_ = Concatenate()(agg_)
    return out_