tf.nn.relu 与 tf.contrib.layers.relu 对比? [英] tf.nn.relu vs. tf.contrib.layers.relu?

问题描述

我看到这里记录了这个tf.nn.relu":https://www.tensorflow.org/api_docs/python/tf/nn/relu

I see this "tf.nn.relu" documented here: https://www.tensorflow.org/api_docs/python/tf/nn/relu

但后来我也在model_fn"这个页面上看到了 tf.contrib.layers.relu 的用法:https://www.tensorflow.org/extend/estimators

But then I also see usage of tf.contrib.layers.relu on this page in "model_fn": https://www.tensorflow.org/extend/estimators

似乎后者不像第一个以类似 API 的方式描述，而只是在使用中呈现.

It seems like the latter isn't described like the first one in an API-like fashion, but only presented in use.

这是为什么?文档是否过时?为什么有两个 - 一个是旧的并且不再支持/将被删除?

Why is this? Are the docs out of date? Why have two - is one old and no longer supported/going to be removed?

推荐答案

它们不是一回事.

后者不是激活函数而是fully_connected layer，其激活函数预设为nn.relu:

The latter is not an activation function but a fully_connected layer that has its activation function preset as nn.relu:

relu = functools.partial(fully_connected, activation_fn=nn.relu)
# ^                                                     |<   >|
# |_ tf.contrib.layers.relu                     tf.nn.relu_|

<小时>

如果您阅读了 contrib.layers，你会发现:

---

If you read the docs for contrib.layers, you'll find:

fully_connected 的别名，它设置了默认的激活函数可用:relu、relu6 和 linear.

Aliases for fully_connected which set a default activation function are available: relu, relu6 and linear.

总而言之，tf.contrib.layers.relu 是 fully_connected 层具有 relu 激活，而 tf.nn.relu 是整流线性单元激活函数本身.

Summarily, tf.contrib.layers.relu is an alias for a fully_connected layer with relu activation while tf.nn.relu is the REctified Linear Unit activation function itself.

这篇关于tf.nn.relu 与 tf.contrib.layers.relu 对比?的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！