Tensorflow 2.1：获取 "Tensor("StatefulPartitionedCall:0", shape=(), dtype=float64)" 作为损失

Question

我得到的不是 Dataset 张量，而是这个，我不知道如何处理：

Tensor("StatefulPartitionedCall:0", shape=(), dtype=float64)

一切顺利（我认为），但这是我尝试打印出损失时得到的结果。这是我正在玩的代码：

import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
from sklearn.datasets import load_breast_cancer
import tensorflow as tf
from tensorflow.keras.models import Model
from tensorflow.keras.layers import Dense
tf.keras.backend.set_floatx('float64')

x, y = load_breast_cancer(return_X_y=True)

data = tf.data.Dataset.from_tensors((x, y)).shuffle(len(x))

train_data = data.take(int(8e-1*len(x))).batch(32)
test_data = data.skip(int(8e-1*len(x)))


class DenseNet(Model):
    def __init__(self):
        super(DenseNet, self).__init__()
        self.D1 = Dense(8, activation=tf.keras.activations.selu)
        self.D2 = Dense(16, activation=tf.keras.activations.elu)
        self.D3 = Dense(32, activation=tf.keras.activations.relu)
        self.D4 = Dense(1)

    def __call__(self, x):
        x = self.D1(x)
        x = self.D2(x)
        x = self.D3(x)
        out = self.D4(x)
        return out


network = DenseNet()

optimizer = tf.keras.optimizers.Adam()


@tf.function
def compute_loss(labels, logits):
    labels = tf.cast(tf.one_hot(labels, depth=1), tf.float64)
    return tf.reduce_mean(
        tf.nn.sigmoid_cross_entropy_with_logits(labels=labels, logits=logits))


@tf.function
def compute_accuracy(labels, logits):
    labels = tf.cast(tf.one_hot(labels, depth=2), tf.float64)
    return tf.reduce_mean(tf.cast(tf.equal(logits, labels), tf.float32))


@tf.function
def train_step(inputs, targets):
    with tf.GradientTape() as tape:
        logits = network(inputs)
        loss = compute_loss(labels=targets, logits=logits)

    gradients = tape.gradient(loss, network.trainable_variables)
    optimizer.apply_gradients(zip(gradients, network.trainable_variables))
    accuracy = compute_accuracy(labels=targets, logits=logits)
    return loss, accuracy


@tf.function
def train():
    for inputs, labels in train_data:
        loss, acc = train_step(inputs, labels)
        print(loss, acc)

def main(epochs=5):
    for i in range(1, epochs + 1):
        train()


if __name__ == '__main__':
    main(epochs=10)

Answer 1

因为你用 @tf.function 修饰 train 函数并且它包含一个 print 调用，那么这意味着它有副作用，因此它只会执行一次（当函数被跟踪）。来自 tf.function 文档：

Important: Any Python side-effects (appending to a list, printing with print, etc) will only happen once, when func is traced. To have side-effects executed into your tf.function they need to be written as TF ops.

因此，要么删除此函数的 @tf.function 装饰（即 train），要么使用 tf.print，这将是 TensorFlow 中等效的打印操作：

    #...
    tf.print("Loss:", loss, "Acc:", acc)

Tensorflow 2.1：获取 "Tensor("StatefulPartitionedCall:0", shape=(), dtype=float64)" 作为损失

Tensorflow 2.1: getting "Tensor("StatefulPartitionedCall:0", shape=(), dtype=float64)" as loss

python

keras

tensorflow

tf.keras

tensorflow2.0