sparse_categorical_crossentropy和SparseCategoricalCrossentropy的用法区别

TensorFlow2 -基本函数解释

tf.keras.losses.SparseCategoricalCrossentropy
和
tf.keras.losses.sparse_categorical_crossentropy

从以下几个方面区分

函数

tf.keras.losses.sparse_categorical_crossentropy(y_true, y_pred)
函数,求交叉熵损失
例子:

import tensorflow as tf
import numpy as np
y_true = tf.constant([1, 2])
y_pred = tf.constant([[0.05, 0.95, 0], [0.1, 0.8, 0.1]])

tf.keras.losses.sparse_categorical_crossentropy(y_true, y_pred)

输出

<tf.Tensor: shape=(2,), dtype=float32, numpy=array([0.05129344, 2.3025851 ], dtype=float32)>

求平均

loss = tf.reduce_mean(tf.keras