版权声明:本文为博主([email protected])原创文章,未经博主允许不得转载。 https://blog.csdn.net/z_feng12489/article/details/90033833
6.5 多输出感知机梯度
多输出感知机
x=tf.random.normal([1,3])
w=tf.ones([3,2])
b=tf.ones([2])
y = tf.constant([0, 1])
with tf.GradientTape() as tape:
tape.watch([w, b])
logits = tf.sigmoid(x@w+b)
loss = tf.reduce_mean(tf.losses.MSE(y, logits))
grads = tape.gradient(loss, [w, b])
print('w grad:', grads[0])
print('b grad:', grads[1])
w grad: tf.Tensor(
[[ 0.0929644 -0.01531836]
[ 0.01146083 -0.00188848]
[-0.02067768 0.0034072 ]], shape=(3, 2), dtype=float32)
b grad: tf.Tensor([ 0.10427201 -0.01718159], shape=(2,), dtype=float32)