Shulou(Shulou.com)06/01 Report--

In this article, the editor introduces in detail "how to use the torch.nn.identity () method in python". The content is detailed, the steps are clear, and the details are handled properly. I hope that this article "how to use the torch.nn.identity () method in python" can help you solve your doubts.

First look at the code m = nn.Identity (54, unused_argument1=0.1, unused_argument2=False) input = torch.randn (128,20) output = m (input) > > print (output.size ()) torch.Size ([128,20])

This is the code given in the official documentation, obviously, there is no change, the input is torch, so is the output, and the given parameters do not seem to have the effect of change.

Look at the source class Identity (Module): r "A placeholder identity operator that is argument-insensitive. Args: args: any argument (unused) kwargs: any keyword argument (unused) Examples:: > > m = nn.Identity (54, unused_argument1=0.1, unused_argument2=False) > input = torch.randn (128,20) > > output = m (input) > print (output.size ()) torch.Size ([128,20]) "" def _ init__ (self, * args) * * kwargs): super (Identity, self). _ _ init__ () def forward (self, input: Tensor)-> Tensor: return input

This is quite simple and clear, ah, what is the input, directly to the output, do not make any changes. Let's look at a sentence in the document: A placeholder identity operator that is argument-insensitive.

A placeholder identification operator that does not distinguish between parameters. Baidu translation, actually means that the design of this network layer is used to occupy space, that is, do not work, but there is such a layer, put into the residual network is to skip the connection of the place to use this layer, appears not so empty!

Application

For example, at this time: if we use se_layer at this time, then SELayer (dim), otherwise output what you enter (do nothing)

After reading this, the article "how to use the torch.nn.identity () method in python" has been introduced. If you want to master the knowledge of this article, you still need to practice and use it yourself to understand it. If you want to know more about related articles, you are welcome to follow the industry information channel.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.