ajbrock / boilerplate Goto Github PK
View Code? Open in Web Editor NEWOdds and Ends and Things I've implemented.
Odds and Ends and Things I've implemented.
Hi,
In fixup.py:
I noticed when I make the following change:
print(b.layer_index, math.sqrt(2. / n), self.layer_index **(-0.5))
to
print(b.layer_index, self.layer_index, math.sqrt(2. / n), self.layer_index **(-0.5))
That b.layer_index
and self.layer_index
differ...
Does this not imply that this line:
b.conv1.weight.data.normal_(0,(self.layer_index ** (-0.5)) * math.sqrt(2. / n))
should rather use (b.layer_index + 1)
instead of self.layer_index
?
Feel free to close this comment if you disagree :)
In the paper you can see the diagram of Fixup which applies the residual branch, and adds the output to the original input. Finally it applies a ReLU to this sum.
https://i.stack.imgur.com/T67F3.png
In the code I cannot see this being applied anywhere (once for each block) - is this an oversight or am i missing something?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.