✨ Deki ✨
Deki is a library is to create beautiful
And the best part, it is extremely simple to use! Just a few steps and you'll find your visuals ready to be played with.
Table of Contents
Installation
$ pip install dekiUsage
Deki works by taking in torch.optim.Optimizer classes and a configuration dictionary for the hyperparameters and other keyword arguments, like in the example below
from torch.optim import SGD
from deki import Rosenbrock
optimiser = SGD
optimiser_state = {"lr":1E-3}
steps = 1000
Rosenbrock.plot((optimiser, optimiser_state), steps)
Custom Optimizer
Just like its mentioned above, any PyTorch optimizer class can be used as long as its required keyword args can be passed in a dict. In conclusion, Deki can handle your custom optimizers, so build away!
from torch.optim import Optimizer
from deki import Rosenbrock
class MySpecialOptimizer(Optimizer):
...
optimizer = MySpecialOptimizer
optimiser_state = {"lr":1E-3}
steps = 1000
Rosenbrock.plot((optimiser, optimiser_state), steps)Custom Test Function
Not only this, Deki also manages to make its Visualizer general so you can implement your custom test functions super easily, just like in the following manner
import torch
from torch.optim import SGD
from deki import Visualizer
# you can define any function as long as it takes a X
# wherein X will be torch Tensor tuple of x,y
def my_special_test_function (X : Tuple[torch.Tensor, torch.Tensor]):
x, y = X
# as an example we will return the sphere function
# use only torch ops to maintain differentiability
return torch.pow(x, 2) + torch.pow(y, 2)
# Define the object for the Visualizer class
# For supported functions, such objects come pre-defined in deki
visualizer = Visualizer( my_special_test_function,
x_range = (-2,2),
y_range = (-2,2),
minima = (0, 0),
start = (-1.5, 1)
)
# Now call the plot function with the optimizer to get the plot
visualizer.plot((SGD, {"lr":1E-3}), steps=1000)
Supported Functions
| Function | Example |
|---|---|
| Rosenbrock | |
| Sphere |
Acknowledgement
A special thanks to all the contributors on this project for making it what it is today!
If you found this repository helpful, please leave a star!
Citation
If you are using this repository for research, please cite it in the following manner:
@misc{minhas_2023,
url={https://github.com/Dawn-Of-Eve/deki},
title={Deki: Library for Visualizing Optimization Algorithms},
publisher={Dawn Of Eve},
author={Minhas, Bhavnick Singh},
year={2023},
month={Feb}} 
