在训练期间打印中间张量

时间:2020-10-14 07:32:04

标签: python tensorflow keras keras-layer tf.keras

我有一个自定义层,我想通过自定义层的call()方法打印未链接到返回的张量(如代码所示)的中间张量。我使用的代码是:

class Similarity(Layer):
    
    def __init__(self, num1, num2):    
        super(Similarity, self).__init__()
        self.num1 = num1
        self.num2 = num2
#         self.total = tf.Variable(initial_value=tf.zeros((16,self.num1, 1)), trainable=False)    
        
    def build(self, input_shape):
        super(Similarity, self).build((None, self.num1, 1))
            
    
    def compute_mask(self, inputs, mask=None):
        # Just pass the received mask from previous layer, to the next layer or 
        # manipulate it if this layer changes the shape of the input
        return mask
        
    def call(self, inputs, mask=None):
        print(">>", type(inputs), inputs.shape, inputs)

        normalized = tf.nn.l2_normalize(inputs, axis = 2)
        print("norm", normalized)
        # multiply row i with row j using transpose
        # element wise product
        similarity = tf.matmul(normalized, normalized,
                         adjoint_b = True # transpose second matrix
                         )
    
        print("SIM", similarity)
        
        z=tf.linalg.band_part(similarity, 0, -1)*3 + tf.linalg.band_part(similarity, -1, 0)*2 - tf.linalg.band_part(similarity,0,0)*6 + tf.linalg.band_part(similarity,0,0)
#         z = K.print_tensor(tf.reduce_sum(z, 2, keepdims=True))
        z = tf.reduce_sum(z, 2, keepdims=True)
    
        z = tf.argsort(z)                # <----------- METHOD2: Reassigned the Z to the tensor I want to print temporarily
        z = K.print_tensor(z)
        print(z)
    
        z=tf.linalg.band_part(similarity, 0, -1)*3 + tf.linalg.band_part(similarity, -1, 0)*2 - tf.linalg.band_part(similarity,0,0)*6 + tf.linalg.band_part(similarity,0,0)

        z = K.print_tensor(tf.reduce_sum(z, 2, keepdims=True)) #<------------- THIS LINE WORKS/PRINTS AS Z is returned
        # z = tf.reduce_sum(z, 2, keepdims=True)
        
        
        @tf.function                             
                                              #<------------- METHOD1: Want to print RANKT tensor but this DID NOT WORKED
        def f(z):
            rankt = K.print_tensor(tf.argsort(z))
#             rankt = tf.reshape(rankt, (-1, self.num1))
#             rankt = K.print_tensor(rankt)
            return rankt
        
        pt = f(z)
        
        return z               # <--------- The returned tensor
    
    def compute_output_shape(self, input_shape):
        print("IS", (None, self.num1, 1))
        return (None, self.num1, 1)

更清楚

我使用了method1,其中我使用了@tf.function来打印rankt张量,但是没有用。

第二,在method2中,我暂时重新分配了z(在{{1}之后返回张量),以便在call()中执行它,并得到打印的值。之后,我将backprop重新分配给原始操作符

总而言之,我不需要z的值,但是我想打印一些依赖于z的变量的值,但是我不能打印除z

2 个答案:

答案 0 :(得分:0)

tf.print功能。

在急切模式下,它不返回任何内容,仅打印张量。在计算图构建期间使用时,它会返回TF运算符,这些运算符可以标识并打印张量值作为副作用。

答案 1 :(得分:0)

我已经晒了很多东西,但是找不到打印中间张量的任何东西。事实证明,我们只能打印链接到张量张量的张量(此处为z)。所以我所做的是,我使用z打印了K.print_tensor(),然后使用那个张量(现在显然是列表形式)来执行我的计算(是侧面计算,而不是在逻辑中实现) )