函数原型:tf.placeholder(dtype, shape=None, name=None)
参数解释:
dtype: The type of elements in the tensor to be fed.(参数类型)
shape: The shape of the tensor to be fed (optional). If the shape is not
specified, you can feed a tensor of any shape.(参数的shape)
name: A name for the operation (optional).
placeholder,中文意思是占位符,在tensorflow中类似于函数参数,运行时必须传入值。
下面复制官方文档中的示例:
x = tf.placeholder(tf.float32, shape=(1024, 1024))
y = tf.matmul(x, x)
with tf.Session() as sess:
print(sess.run(y)) # ERROR: will fail because x was not fed.
rand_array = np.random.rand(1024, 1024)
print(sess.run(y, feed_dict={x: rand_array})) # Will succeed.
从官方示例来看,placeholder一定要在运行时候feed值。
下面看另一个示例:
x = tf.placeholder(tf.float32, [None, 784])
这个None,代表可以动态放入任意维数.