Hugging Face
Models
Datasets
Spaces
Buckets
new
Docs
Enterprise
Pricing
Log In
Sign Up
Motif-Technologies
/
optimizer
like
55
Follow
Motif Technologies
179
Kernels
arxiv:
2511.07464
License:
apache-2.0
Model card
Files
Files and versions
xet
Community
8
Use this model
2a8631f
optimizer
18.1 MB
Ctrl+K
Ctrl+K
15 contributors
History:
26 commits
ca1207
use COMM_DTYPE instead of hardcoded dtype
2a8631f
7 months ago
.github
chore: add nix build workflow (#8)
7 months ago
build
feat(muon_clip) : add muon clip (#6)
7 months ago
docs
chore(muon): clean build and update doc
10 months ago
optimizer
applied lint
8 months ago
test
apply all2all scatter gather
7 months ago
torch-ext
use COMM_DTYPE instead of hardcoded dtype
7 months ago
.gitattributes
Safe
1.64 kB
chore: initial commit
10 months ago
.gitignore
Safe
233 Bytes
fix(muon): add update_p stage and dealloc tensors properly
8 months ago
.pre-commit-config.yaml
Safe
850 Bytes
add more exclude files
8 months ago
README.md
Safe
2.24 kB
feat: update muon to receive paramgroups, not model (#4)
8 months ago
build.toml
Safe
323 Bytes
chore: initial commit
10 months ago
flake.nix
Safe
315 Bytes
chore: initial commit
10 months ago