facebookresearch/higher

torch.optim.AdamW not in the list of supported optimizers

Open

#88 opened on Nov 8, 2020

View on GitHub
 (2 comments) (0 reactions) (0 assignees)Python (1,556 stars) (123 forks)batch import
enhancementgood first issue

Description

I know that it's possible to define our own custom Differentiable Optimizers and get higher to register them. But I was wondering why optim.AdamW is not supported, since from the doc it appears that optimizers from torch.optim are available out of the box.

Contributor guide

Tech stack
python
Domain
machine learning
Issue type
feature
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
3
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
fresh
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
mostly clear
Prerequisites
basic PyTorch knowledgeunderstanding of higher's optimizer registration
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
50
Research direction
Investigate how existing optimizers like SGD and Adam are registered in higher's source code (e.g., in `higher/optim.py`). Add `torch.optim.AdamW` by following the same pattern, ensuring weight decay and other parameters are correctly handled. Check if there are any existing tests for optimizer support and add a test for AdamW.