我可以在多类分类问题中使用tf.metrics.BinaryAccuracy吗？

import tensorflow as tf

# 假设有3个类别
num_classes = 3

# 将标签转换为One-hot编码
labels = tf.constant([0, 1, 2, 1, 0])
one_hot_labels = tf.one_hot(labels, num_classes)

# 假设有3个预测值
predictions = tf.constant([[0.9, 0.05, 0.05],
                           [0.2, 0.7, 0.1],
                           [0.1, 0.2, 0.7],
                           [0.3, 0.6, 0.1],
                           [0.8, 0.1, 0.1]])

# 计算准确率
accuracy, update_op = tf.metrics.BinaryAccuracy(one_hot_labels, predictions)

# 初始化变量
init_op = tf.global_variables_initializer()
local_init_op = tf.local_variables_initializer()

with tf.Session() as sess:
    sess.run(init_op)
    sess.run(local_init_op)
    
    # 更新指标
    sess.run(update_op)
    
    # 获取准确率
    acc = sess.run(accuracy)
    print("Accuracy:", acc)