lilipads / gradient_descent_viz Goto Github PK
View Code? Open in Web Editor NEWinteractive visualization of 5 popular gradient descent methods with step-by-step illustration and hyperparameter tuning UI
License: MIT License
interactive visualization of 5 popular gradient descent methods with step-by-step illustration and hyperparameter tuning UI
License: MIT License
I didn't see one in the repo. I'd recommend MIT or Unlicense if you're down!
@lilipads Thanks for sharing the interesting work. I am a researcher from the Robotics field with a special focus on state estimation. Do you have any idea that how can we visualize the shape of the optimization when the size of the state to be optimized is more than 3 dimensions?
Thanks a lot in advance,
Best,
Weisong,
如题,您好,请问是镜像问题还是我下载的问题?可否更新一下镜像?谢谢。Nice work。
Cannot run the .exe file on Windows 10 due to error:
The code execution cannot proceed because libgcc_s_seh-1.dll was not found. Reinstalling the program may fix this problem.
Reinstalling the program did not fix the problem.
good works !
some enhance you can reference:
surface visualization like https://github.com/tomgoldstein/loss-landscape
training loss animation visualization
https://github.com/joshuagruenstein/atlas
https://github.com/spencer-hann/Loss_Surface_Visualizer
https://www.deeplearning.ai/ai-notes/optimization
Hi,
Sry I'm new to this website and i don't know if that's the right section to post in for this, but i have a presentation for my first year in master of mathematics, and i wanted to ask for permission to show images/videos of this tool during a presentation on gradient descent methods.
Thank you for helping me understand these better anyway,
Cordially,
Guillaume Bauman.
double GradientDescent::f(double x, double z){
switch (function_name){
case Function::local_minimum:{
z *= 1.4;
return -2 * exp(-((x - 1) * (x - 1) + z * z) / .2) -
6. * exp(-((x + 1) * (x + 1) + z * z) / .2) +
x * x + z * z;
}
case Function::global_minimum:{
return x * x + z * z;
}
case Function::saddle_point:{
return sin(x) + z * z;
}
case Function::ecliptic_bowl:{
x /= 2.;
z /= 2.;
return -exp(-(x * x + 5 * z * z)) + x * x + 0.5 * z * z;
}
case Function::hills:{
z *= 1.4;
return 2 * exp(-((x - 1) * (x - 1) + z * z) / .2) +
6. * exp(-((x + 1) * (x + 1) + z * z) / .2) -
2 * exp(-((x - 1) * (x - 1) + (z + 1) * (z + 1)) / .2) +
x * x + z * z;
}
case Function::plateau:{
x *= 10;
z *= 10;
double r = sqrt(z * z + x * x) + 0.01;
return -sin(r) / r + 0.01 * r * r;
}
}
return 0.;
}
sometimes z will be multiplied by 1.4, why?
Edit: fixed
Just wonder if someone went through these issues
I´m running on Windows 10 Single Home, 64 bits
How to change the surface function? I want to use my own objective function to generate animation. Thanks. Nice work!
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 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. 📊📈🎉
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 ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.