Instructions to use Motif-Technologies/optimizer with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Kernels
How to use Motif-Technologies/optimizer with Kernels:
# !pip install kernels from kernels import get_kernel # a version (or an explicit revision) is required; see the "Files and versions" tab for the available ones kernel = get_kernel("Motif-Technologies/optimizer", version=1) - Notebooks
- Google Colab
- Kaggle
Commit History
-
fix init
Verified signaturefix muon to recieve model parameter, not model6d75af0
-
fix(muon): add update_p stage and dealloc tensors properly
99e7c0c
-
chore: add .gitignore
79fc8ba
-
applied lint
db36e39
-
fix(optimizer): resolve bug where weight decay was multiplied by wrong lr value (#5)
Verified signature- fix(optimizer): resolve bug where weight decay was multiplied by wrong lr value (33568272536def07cf6095dadff78ca2d4b182b5)671b033
-
Support HSDP (#4)
Verified signature- feat(muon): support HSDP (dd04b0152725e7b55970171706d5d16f59352cda)8447fd1
-
fix(muon): handle un-distributed env
1f13dae -
refactor(muon): change argument adam_wd to weight_decay and handle params' type
02ac540
-
fix(muon): free tensors that are no longer needed
64757cb -
chore(muon): update comment
036642a -
chore(muon): clean build and update doc
febdf5b
-
fix(muon): delete intermediate tensors immediately to lower peak mem usage
bdd2678 -
chore: initial commit
8535e80