Adaptive Methods for Nonconvex Optimization.
Implementation of YOGI optimizer as mentioned in the paper https://papers.nips.cc/paper/8186-adaptive-methods-for-nonconvex-optimization.
Getting Started
Yogi is an optimizer proposed to overcome the disadvantages of Adam Optimizer.
Implementation
This is the first implementaion of Yogi Optimizer in PyTorch.
How to Run
Refer to PyTorch folder README for running the code.
To-Do
-
Implement Yogi in
a) MxNet
b) Tensorflow
-
Add more experiments such as:
a) Auto Encoders
b) Machine Translation