从meta

Question

我想还原模型而无需手动构建图形。我只想将预测变量用于新数据样本。

所以我发现这是通常的方式：

model_path = "my_saved_model"
inference_graph = tf.Graph()
with tf.Session(graph= inference_graph) as sess:
# Load the graph with the trained states
  loader = tf.train.import_meta_graph(model_path+'.meta')
  loader.restore(sess, model_path)

获取tnesors（这是我的问题）

pred = inference_graph.get_tensor_by_name('predictions?:0')
_x  = inference_graph get_tensor_by_name('x1:0')
_y  = inference_graph.get_tensor_by_name('x2:0')

然后执行sess.run（_accuracy，feed_dict = {_ x：...，_y：...}

因此，您需要从模型（从元文件）中获得所需结果的名称

但是我不知道我的预测用哪个名字。这是我的模型：

    self.x1 = tf.placeholder(dtype=tf.int32, shape=[None, max_sequence_len],name="x1")
    self.x2 = tf.placeholder(dtype=tf.int32, shape=[None, max_sequence_len],name="x2")
    self.is_training = tf.placeholder(dtype=tf.bool)
    self.labels = tf.placeholder(dtype=tf.int32, shape=[None, 1])
    self.sentences_lengths = tf.placeholder(dtype=tf.int32, shape=[None])

    self.debug = None

    self.embedding_size = main_cfg['PARAMS'].getint('embedding_size')
    self.learning_rate = main_cfg['TRAINING'].getfloat('learning_rate')

    self.embedding_mat=embedding_mat

    with tf.variable_scope('embeddings'):
        word_embeddings = tf.constant(self.embedding_mat, dtype=tf.float32, name="embedding")
        self.embedded_x1 = tf.nn.embedding_lookup(word_embeddings, self.x1)
        self.embedded_x2 = tf.nn.embedding_lookup(word_embeddings, self.x2)

       # word_embeddings = tf.get_variable('word_embeddings', [vocabulary_size, self.embedding_size])
        #self.embedded_x1 = tf.gather(word_embeddings, self.x1)
        #self.embedded_x2 = tf.gather(word_embeddings, self.x2)

    with tf.variable_scope('siamese'):
        self.predictions = self.siamese_layer(max_sequence_len, model_cfg)

    with tf.variable_scope('loss'):
        self.loss = loss_function(self.labels, self.predictions)
        self.opt = optimize(self.loss, self.learning_rate)

    with tf.variable_scope('metrics'):
        self.temp_sim = tf.rint(self.predictions)
        self.correct_predictions = tf.equal(self.temp_sim, tf.to_float(self.labels))
        self.accuracy = tf.reduce_mean(tf.to_float(self.correct_predictions))

    with tf.variable_scope('summary'):
        tf.summary.scalar("loss", self.loss)
        tf.summary.scalar("accuracy", self.accuracy)
        self.summary_op = tf.summary.merge_all()



def siamese_layer(self, sequence_len, model_cfg):
    num_filters = parse_list(model_cfg['PARAMS']['num_filters'])
    filter_sizes = parse_list(model_cfg['PARAMS']['filter_sizes'])

   # print(self.embedded_x1)
    embedded_x1=self.embedded_x1
    out1 = cnn_layers(self.embedded_x1,
                      sequence_len,
                      num_filters=num_filters,
                      filter_sizes=filter_sizes)

    out2 = cnn_layers(self.embedded_x2,
                      sequence_len,
                      num_filters=num_filters,
                      filter_sizes=filter_sizes,
                      reuse=True)

    out1 = dropout(out1, self.is_training)
    out2 = dropout(out2, self.is_training)

    return manhattan_similarity(out1, out2)

因此，对于新的预测，我只需要输入两个名称分别为x1和x2的sel.prediction即可。但是如何从上面的代码示例中按名称来命名呢？