def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_size = int(config['num_filters'])
self.win_size1 = int(config['window_size_left'])
self.win_size2 = int(config['window_size_right'])
self.dpool_size1 = int(config['dpool_size_left'])
self.dpool_size2 = int(config['dpool_size_right'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.emb_layer = layers.EmbeddingEnhancedLayer(self.vocab_size,
self.emb_size,
zero_pad=True,
scale=False)
self.cnn_layer = layers.CNNDynamicPoolingLayer(
self.seq_len1, self.seq_len2, self.win_size1, self.win_size2,
self.dpool_size1, self.dpool_size2, self.kernel_size)
self.relu_layer = layers.ReluLayer()
self.tanh_layer = layers.TanhLayer()
if 'match_mask' in config and config['match_mask'] != 0:
self.match_mask = True
else:
self.match_mask = False
self.fc1_layer = layers.FCLayer(
self.kernel_size * self.dpool_size1 * self.dpool_size2,
self.hidden_size)
self.n_class = int(config['n_class'])
self.fc2_layer = layers.FCLayer(self.hidden_size, self.n_class)
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.k_max_num = int(config['k_max_num'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.task_mode = config['training_mode']
self.emb_layer = layers.EmbeddingEnhancedLayer(self.vocab_size,
self.emb_size,
zero_pad=True,
scale=False)
self.fw_cell = tf.nn.rnn_cell.LSTMCell(num_units=self.hidden_size,
state_is_tuple=True)
self.bw_cell = tf.nn.rnn_cell.LSTMCell(num_units=self.hidden_size,
state_is_tuple=True)
if 'match_mask' in config and config['match_mask'] != 0:
self.match_mask = True
else:
self.match_mask = False
if self.task_mode == "pointwise":
self.n_class = int(config['n_class'])
self.fc2_layer = layers.FCLayer(self.k_max_num, self.n_class)
elif self.task_mode == "pairwise":
self.fc2_layer = layers.FCLayer(self.k_max_num, 1)
else:
logging.error("training mode not supported")
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.rnn_hidden_size = int(config['rnn_hidden_size'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.task_mode = config['training_mode']
self.emb_layer = layers.EmbeddingEnhancedLayer(self.vocab_size,
self.emb_size,
zero_pad=True,
scale=False)
self.rnn = layers.LSTMLayer(self.rnn_hidden_size)
self.extract = layers.ExtractLastLayer()
if self.task_mode == "pointwise":
self.n_class = int(config['n_class'])
self.fc1_layer = layers.FCLayer(self.rnn_hidden_size * 2,
self.hidden_size)
self.fc2_layer = layers.FCLayer(self.hidden_size, self.n_class)
elif self.task_mode == "pairwise":
self.fc1_layer = layers.FCLayer(self.rnn_hidden_size * 1,
self.hidden_size)
self.cos_layer = layers.CosineLayer()
else:
logging.error("training mode not supported")
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_size = int(config['num_filters'])
self.win_size1 = int(config['window_size_left'])
self.win_size2 = int(config['window_size_right'])
self.dpool_size1 = int(config['dpool_size_left'])
self.dpool_size2 = int(config['dpool_size_right'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.task_mode = config['training_mode']
self.emb_layer = layers.EmbeddingEnhancedLayer(self.vocab_size,
self.emb_size,
zero_pad=True,
scale=False) #加强版本的词向量
self.cnn_layer = layers.CNNDynamicPoolingLayer(
self.seq_len1, self.seq_len2, self.win_size1, self.win_size2,
self.dpool_size1, self.dpool_size2, self.kernel_size)
self.relu_layer = layers.ReluLayer()
self.tanh_layer = layers.TanhLayer()
if 'match_mask' in config and config['match_mask'] != 0:
self.match_mask = True #是否考虑填充区域
else:
self.match_mask = False
self.fc1_layer = layers.FCLayer(
self.kernel_size * self.dpool_size1 * self.dpool_size2,
self.hidden_size)
if self.task_mode == "pointwise":
self.n_class = int(config['n_class'])
self.fc2_layer = layers.FCLayer(self.hidden_size, self.n_class)
elif self.task_mode == "pairwise":
self.fc2_layer = layers.FCLayer(self.hidden_size, 1)
else:
logging.error("training mode not supported")
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_size = int(config['num_filters'])
self.win_size = int(config['window_size'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len = config['left_slots'][0] #左文本长度
self.right_name, self.seq_len = config['right_slots'][0] #右文本长度
self.task_mode = config['training_mode']
self.emb_layer = layers.EmbeddingLayer(self.vocab_size,
self.emb_size) #embedding矩阵
self.cnn_layer = layers.CNNLayer(self.seq_len, self.emb_size,
self.win_size,
self.kernel_size) #卷积池化层
self.relu_layer = layers.ReluLayer() #激活函数层
self.concat_layer = layers.ConcatLayer() #concat连接层
if self.task_mode == "pointwise":
self.n_class = int(config['n_class'])
self.fc1_layer = layers.FCLayer(
2 * self.kernel_size,
self.hidden_size) #全连接层1 2*self.kernel_size hidden_size
self.fc2_layer = layers.FCLayer(
self.hidden_size, self.n_class) #全连接层2 hidden_size * n_class
elif self.task_mode == "pairwise":
self.fc1_layer = layers.FCLayer(self.kernel_size, self.hidden_size)
self.cos_layer = layers.CosineLayer()
else:
logging.error("training mode not supported")
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.bow_size = int(config["bow_size"])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len = config["left_slots"][0]
self.right_name, self.seq_len = config["right_slots"][0]
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.seq_pool_layer = layers.SequencePoolingLayer()
self.softsign_layer = layers.SoftsignLayer()
self.bow_layer = layers.FCLayer(self.emb_size, self.bow_size)
self.n_class = int(config['n_class'])
self.fc_layer = layers.FCLayer(2 * self.hidden_size, self.n_class)
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.rnn_hidden_size = int(config['rnn_hidden_size'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.emb_layer = layers.EmbeddingEnhancedLayer(self.vocab_size,
self.emb_size,
zero_pad=True,
scale=False)
self.rnn = layers.LSTMLayer(self.rnn_hidden_size)
self.extract = layers.ExtractLastLayer()
self.n_class = int(config['n_class'])
self.fc1_layer = layers.FCLayer(self.rnn_hidden_size * 2,
self.hidden_size)
self.fc2_layer = layers.FCLayer(self.hidden_size, self.n_class)
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_size = int(config['num_filters'])
self.win_size = int(config['window_size'])
self.hidden_size = int(config['hidden_size'])
self.left_name, self.seq_len = config['left_slots'][0]
self.right_name, self.seq_len = config['right_slots'][0]
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.cnn_layer = layers.CNNLayer(self.seq_len, self.emb_size,
self.win_size, self.kernel_size)
self.relu_layer = layers.ReluLayer()
self.tanh_layer = layers.TanhLayer()
self.concat_layer = layers.ConcatLayer()
self.fc1_layer = layers.FCLayer(self.kernel_size, self.hidden_size)
self.n_class = int(config['n_class'])
self.fc2_layer = layers.FCLayer(2 * self.hidden_size, self.n_class)
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_num = int(config['kernel_num'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.lamb = float(config['lamb'])
self.task_mode = config['training_mode']
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.sim_mat_layer = layers.SimilarityMatrixLayer()
self.kernel_pool_layer = layers.KernelPoolingLayer(self.kernel_num, self.lamb)
self.tanh_layer = layers.TanhLayer()
if self.task_mode == "pointwise":
self.n_class = int(config['n_class'])
self.fc_layer = layers.FCLayer(self.kernel_num, self.n_class)
elif self.task_mode == "pairwise":
self.fc_layer = layers.FCLayer(self.kernel_num, 1)
else:
logging.error("training mode not supported")
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.bow_size = int(config["bow_size"])
self.left_name, self.seq_len = config["left_slots"][0]
self.right_name, self.seq_len = config["right_slots"][0]
self.task_mode = config['training_mode']
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.seq_pool_layer = layers.SequencePoolingLayer()#将词向量作简单的平均
self.softsign_layer = layers.SoftsignLayer()#激活函数层
if self.task_mode == "pointwise":
self.n_class = int(config['n_class'])
self.bow_layer = layers.FCLayer(self.emb_size * 2, self.bow_size)
self.relu_layer = layers.ReluLayer()
self.fc_layer = layers.FCLayer(self.bow_size, self.n_class)
elif self.task_mode == "pairwise":
self.bow_layer = layers.FCLayer(self.emb_size, self.bow_size)
self.cos_layer = layers.CosineLayer()#余弦相似度计算
else:
logging.error("training mode not supported")
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_num = int(config['num_filters'])
self.win_size = int(config['window_size'])
self.pool_size = int(config['pool_size'])
self.hidden_size = int(config['hidden_size'])
self.dropout_rate = float(config['dropout_rate'])
self.drop_out = layers.DropoutLayer(drop_rate=self.dropout_rate)
self.tanh_layer = layers.TanhLayer()
self.left_name, self.seq_len = config['left_slots'][0]
self.right_name, self.seq_len = config['right_slots'][0]
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.cnn_layer = layers.CNNLayerConfig(seq_len=self.seq_len, emb_dim=self.emb_size, win_height=1,
win_width=self.win_size, kernel_num=self.kernel_num, same_conv=True,
activate=False, pooling=True, pool_size=self.pool_size)
self.concat_layer = layers.ConcatLayer()
self.fc1_layer = layers.FCLayer(self.kernel_size, self.hidden_size)
self.n_class = int(config['n_class'])
self.fc2_layer = layers.FCLayer(2 * self.hidden_size, self.n_class)
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.kernel_num = int(config['kernel_num'])
self.left_name, self.seq_len1 = config['left_slots'][0]
self.right_name, self.seq_len2 = config['right_slots'][0]
self.lamb = float(config['lamb'])
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.sim_mat_layer = layers.SimilarityMatrixLayer()
self.kernel_pool_layer = layers.KernelPoolingLayer(
self.kernel_num, self.lamb)
self.tanh_layer = layers.TanhLayer()
self.n_class = int(config['n_class'])
self.fc_layer = layers.FCLayer(self.kernel_num, self.n_class)
def __init__(self, config):
self.vocab_size = int(config['vocabulary_size'])
self.emb_size = int(config['embedding_dim'])
self.dropout_rate = float(config['dropout_rate'])
self.left_name, self.seq_len = config['left_slots'][0]
self.right_name, self.seq_len = config['right_slots'][0]
self.kernel_num1 = int(config['num_filters_1'])
self.win_size1 = int(config['window_size_1'])
self.kernel_num2 = int(config['num_filters_2'])
self.win_size2 = int(config['window_size_2'])
self.pool_size2 = int(config['pool_size2'])
self.drop_out = layers.DropoutLayer(drop_rate=self.dropout_rate)
self.emb_layer = layers.EmbeddingLayer(self.vocab_size, self.emb_size)
self.cnn1_layer = layers.CNNLayerConfig(seq_len=self.seq_len,
emb_dim=self.emb_size,
win_height=self.win_size1,
win_width=self.emb_size,
kernel_num=self.kernel_num1,
same_conv=False,
pooling=False,
activate=False)
self.cnn_out_len = self.seq_len - self.win_size1 + 1
self.cnn2_layer = layers.CNNLayerConfig(seq_len=self.seq_len,
emb_dim=self.seq_len,
win_height=self.win_size2,
win_width=self.win_size2,
kernel_num=self.kernel_num2,
same_conv=True,
pooling=True,
activate=True,
pool_size=self.pool_size2)
self.cnn2_out_size = (self.cnn_out_len - self.pool_size2 + 1) * (
self.cnn_out_len - self.pool_size2 + 1) * self.kernel_num2
self.n_class = int(config['n_class'])
self.fc_layer = layers.FCLayer(self.cnn2_out_size, self.n_class)
