google / google/learned_optimization
Understanding the differences compared to your earlier work & library
Open
- Dominant language
- Python
- Stars
- 814
- Forks
- 73
- Avg merge
- 9h 40m
- Merged PRs (30d)
- 2
Description
What is the situation when comparing this work to your earlier work of [opt_list](https://github.com/google-research/google-research/tree/master/opt_list)? Is that obsolete? What are the practical differences?
Also, any plans to add [AdaBound](https://arxiv.org/abs/1902.09843) and [Adam-HD](https://github.com/gbaydin/hypergradient-descent) to your benchmarks? Both have fairly high citations/year.
Contributor guide
Assessment
This issue has not been assessed yet.