https://github.com/pytorch/ao

brrr cuda dtypes float8 inference llama mx pytorch quantization sparsity training transformer

Score: 28.378166544026453

Last synced: about 11 hours ago
JSON representation

Repository metadata:

PyTorch native quantization for training and inference

Host: GitHub
URL: https://github.com/pytorch/ao
Owner: pytorch
License: other
Created: 2023-11-03T21:27:36.000Z (about 3 years ago)
Default Branch: main
Last Pushed: 2026-10-28T06:21:01.000Z (7 days ago)
Last Synced: 2026-10-28T08:06:51.636Z (7 days ago)
Topics: brrr, cuda, dtypes, float8, inference, llama, mx, pytorch, quantization, sparsity, training, transformer
Language: Python
Homepage: https://pytorch.org/ao/stable/index.html
Size: 168 MB
Stars: 3,018
Watchers: 44
Forks: 662
Open Issues: 827
Metadata Files:
- Readme: README.md
- Contributing: CONTRIBUTING.md
- License: LICENSE
- Code of conduct: CODE_OF_CONDUCT.md
- Citation: CITATION.cff
- Codeowners: CODEOWNERS
- Claude: CLAUDE.md

Owner metadata:

Name: pytorch
Login: pytorch
Email:
Kind: organization
Description:
Website: https://pytorch.org
Location: where the eigens are valued
Twitter:
Company:
Icon url: https://avatars.githubusercontent.com/u/21003710?v=4
Repositories: 70
Last Synced at: 2026-11-01T23:32:36.590Z
Profile URL: https://github.com/pytorch

Committers metadata

Last synced: 2 days ago

Total Commits: 2,815
Total Committers: 257
Avg Commits per committer: 10.953
Development Distribution Score (DDS): 0.865

Commits in past year: 841
Committers in past year: 143
Avg Commits per committer in past year: 5.881
Development Distribution Score (DDS) in past year: 0.819

Name	Email	Commits
Vasiliy Kuznetsov	v****o	380
Jerry Zhang	j**8@g**m	364
Daniel Vega-Myhre	d**m@m**m	259
andrewor14	a**4@g**m	159
Scott Roy	1****y	143
Apurva Jain	a**a@g**m	129
Mark Saroufim	m**m@m**m	104
Jesse Cai	j**i@m**m	93
HDCharles	3****s	82
Driss Guessous	3****g	82
cpuhrsch	c**h@g**m	62
Thien Tran	g**t@y**g	55
Xia Weiwen	w**a@i**m	47
Zeyu Song	8****s	38
liangel-02	l**l@m**m	29
Lisa Jin	l****n	28
iamzainhuda	3****a	28
xiangdong	4****6	24
Peter Yeh	p****x	21
Kimish Patel	k**l@m**m	21
Manuel Candales	4****s	21
Howard Zhang	3****v	20
Xuan Liao	x**o@i**m	16
Aleksandar Samardžić	1****c	16
namgyu-youn	y**2@g**m	16
Svetlana Karslioglu	s**s@m**m	14
Andrey Talman	a**n@f**m	14
shiyang-weng	s**g@i**m	13
Huy Do	h**n@g**m	13
supriyar	s**r@f**m	11
and 227 more...

Issue and Pull Request metadata

Last synced: 2 days ago

Total issues: 651
Total pull requests: 3,783
Average time to close issues: 6 months
Average time to close pull requests: 11 days
Total issue authors: 235
Total pull request authors: 364
Average comments per issue: 2.77
Average comments per pull request: 2.39
Merged pull request: 2,210
Bot issues: 20
Bot pull requests: 4

Past year issues: 97
Past year pull requests: 460
Past year average time to close issues: 2 months
Past year average time to close pull requests: 29 days
Past year issue authors: 46
Past year pull request authors: 167
Past year average comments per issue: 3.66
Past year average comments per pull request: 3.92
Past year merged pull request: 118
Past year bot issues: 20
Past year bot pull requests: 1

More stats: https://issues.ecosyste.ms/repositories/lookup?url=https://github.com/pytorch/ao

Top Issue Authors

vkuzo (59)
msaroufim (55)
jerryzh168 (38)
gau-nernst (25)
drisspg (22)
github-actions[bot] (20)
andrewor14 (20)
danielvegamyhre (18)
HDCharles (11)
metascroy (11)
jcaip (10)
felipemello1 (7)
goldhuang (7)
ebsmothers (6)
vayuda (5)

Top Pull Request Authors

jerryzh168 (400)
vkuzo (356)
jainapurva (313)
danielvegamyhre (235)
metascroy (228)
msaroufim (178)
andrewor14 (169)
drisspg (168)
HDCharles (117)
jcaip (95)
gau-nernst (83)
cpuhrsch (67)
petrex (47)
kimishpatel (43)
alexsamardzic (40)

Top Issue Labels

triaged (48)
bug (41)
good first issue (35)
float8 (32)
ci-failure (20)
enhancement (19)
tracker (16)
ci (15)
question (14)
optimizer (12)
mx (10)
topic: documentation (9)
quantize (9)
module: rocm (8)
multibackend (8)
binaries (7)
performance (7)
CLA Signed (6)
rfc (6)
autoquant (4)
moe (4)
build (4)
cpu (4)
sparsity (4)
module: training (3)
quantize_ (3)
distributed (2)
pt2e_quant (2)
reproduction needed (2)
topic: performance (2)

Top Pull Request Labels

CLA Signed (3,213)
topic: not user facing (876)
fb-exported (294)
topic: improvement (220)
topic: bug fix (152)
topic: new feature (135)
module: not user facing (109)
topic: documentation (98)
module: rocm (97)
ciflow/rocm (78)
topic: for developers (73)
topic: performance (65)
module: training (52)
meta-exported (45)
topic: bc-breaking (43)
ciflow/xpu (42)
ci-no-td (42)
xpu (42)
mx (41)
float8 (35)
ci (31)
module: inference (29)
sparsity (21)
module: qat (20)
moe (19)
build (16)
topic: deprecation (16)
Merged (16)
module: pt2e_quant (14)
cpu (12)

Package metadata

Total packages: 5
Total downloads:
- conda: 370 total
- pypi: 2,135,868 last-month
Total dependent packages: 1 (may contain duplicates)
Total dependent repositories: 0 (may contain duplicates)
Total versions: 98
Total maintainers: 9

proxy.golang.org: github.com/pytorch/ao

Homepage:
Documentation: https://pkg.go.dev/github.com/pytorch/ao#section-documentation
Licenses: other
Latest release: v0.18.0 (published 3 months ago)
Last Synced: 2026-11-02T03:36:49.526Z (2 days ago)
Versions: 68
Dependent Packages: 0
Dependent Repositories: 0
Rankings:
- Dependent packages count: 6.515%
- Average: 6.733%
- Dependent repos count: 6.952%

pypi.org: torchao

Package for applying ao techniques to GPU models

Homepage: https://github.com/pytorch/ao
Documentation: https://torchao.readthedocs.io/
Licenses: other
Latest release: 0.18.0 (published 3 months ago)
Last Synced: 2026-11-02T03:36:47.399Z (2 days ago)
Versions: 24
Dependent Packages: 1
Dependent Repositories: 0
Downloads: 2,135,868 Last month
Rankings:
- Dependent packages count: 9.917%
- Forks count: 29.845%
- Average: 36.595%
- Stargazers count: 38.86%
- Dependent repos count: 67.758%
Maintainers (7)
- marksaroufim
- seemethere
- cpuhrsch
- rvarm1
- atalman
- HDCharles
- jerryzh168

anaconda.org: torchao

TorchAO is an easy-to-use quantization and optimization library built on native PyTorch. It provides quantization (INT8, INT4, FP8), sparsity, and low-bit optimizer support, and works out-of-the-box with torch.compile() and FSDP2. TorchAO is the official PyTorch quantization backend for vLLM, HuggingFace Transformers, and TorchTune.

Homepage: https://github.com/pytorch/ao
Licenses: BSD-3-Clause
Latest release: 0.17.0 (published 5 months ago)
Last Synced: 2026-07-15T20:17:07.208Z (4 months ago)
Versions: 2
Dependent Packages: 0
Dependent Repositories: 0
Downloads: 370 Total
Rankings:
- Dependent packages count: 35.312%
- Average: 36.908%
- Dependent repos count: 38.503%

nixpkgs-unstable: python313Packages.torchao

PyTorch native quantization and sparsity for training and inference

Homepage: https://github.com/pytorch/ao
Documentation: https://github.com/NixOS/nixpkgs/blob/nixos-unstable/pkgs/development/python-modules/torchao/default.nix#L329
Licenses: BSD-3-Clause
Latest release: 0.18.0 (published 2 days ago)
Last Synced: 2026-11-01T11:15:11.896Z (2 days ago)
Versions: 2
Dependent Packages: 0
Dependent Repositories: 0
Rankings:
- Dependent repos count: 0.0%
- Dependent packages count: 0.0%
- Average: 100%
Maintainers (2)
- GaetanLepage
- sarahec

nixpkgs-unstable: python314Packages.torchao

PyTorch native quantization and sparsity for training and inference

Homepage: https://github.com/pytorch/ao
Documentation: https://github.com/NixOS/nixpkgs/blob/nixos-unstable/pkgs/development/python-modules/torchao/default.nix#L329
Licenses: BSD-3-Clause
Latest release: 0.18.0 (published 2 days ago)
Last Synced: 2026-11-01T11:49:12.808Z (2 days ago)
Versions: 2
Dependent Packages: 0
Dependent Repositories: 0
Rankings:
- Dependent repos count: 0.0%
- Dependent packages count: 0.0%
- Average: 100%
Maintainers (2)
- GaetanLepage
- sarahec

Dependencies

setup.py pypi

torch *

.github/workflows/1xL4_tests.yml actions

.github/workflows/ci-failure-issue.yml actions

actions/checkout v4 composite
actions/download-artifact v4 composite
actions/upload-artifact v4 composite
anthropics/claude-code-action 593d7a5c4e0073569f74772c2b7b64c30ec14707 composite
aws-actions/configure-aws-credentials 7474bc4690e29a8392af63c5b98e7449536d5c3a composite
pytorch/test-infra/.github/actions/upload-claude-usage main composite

.github/workflows/regression_test_rocm.yml actions

.github/workflows/ruff_linter.yml actions

actions/checkout v4 composite
actions/setup-python v3 composite

.github/workflows/4xH100_tests.yml actions

.github/workflows/doc_build.yml actions

actions/checkout v4 composite
actions/download-artifact v4 composite
actions/github-script v7 composite
actions/upload-artifact v4 composite
conda-incubator/setup-miniconda v2 composite
seemethere/upload-artifact-s3 v5 composite

.github/workflows/nightly_smoke_test.yml actions

.github/workflows/pr-label-check.yml actions

.github/workflows/regression_test_pt2e_cpu.yml actions

.github/workflows/validate-binaries.yml actions

.github/workflows/1xH100_tests.yml actions

.github/workflows/build_wheels_linux_x86.yml actions

dawidd6/action-send-mail v4 composite

.github/workflows/claude-code.yml actions

.github/workflows/run_tutorials.yml actions

actions/checkout v4 composite
pytorch/test-infra/.github/actions/setup-miniconda main composite

.github/workflows/xpu_test.yml actions

actions/checkout v4 composite
actions/upload-artifact ea165f8d65b6e75b540449e92b4886f43607fa02 composite
pytorch/pytorch/.github/actions/setup-xpu main composite
pytorch/pytorch/.github/actions/teardown-xpu main composite
pytorch/test-infra/.github/actions/calculate-docker-image main composite
pytorch/test-infra/.github/actions/pull-docker-image main composite

docs/requirements.txt pypi

docutils ==0.20
myst-parser ==4.0.1
pytorch-sphinx-theme2 ==0.4.3
sphinx ==7.2.6
sphinx-copybutton ==0.5.2
sphinx-design ==0.6.1
sphinx-gallery ==0.11.1
sphinx-sitemap ==2.6.0
sphinx-tabs ==3.4.7
sphinxcontrib-mermaid ==1.0.0
sphinxcontrib.katex ==0.8.6
sphinxext-opengraph ==0.9.1

.github/workflows/build_wheels_linux_aarch64.yml actions

dawidd6/action-send-mail v4 composite

.github/workflows/metal_test.yml actions

.github/workflows/regression_test.yml actions

.github/workflows/regression_test_aarch64.yml actions

actions/checkout v3 composite
conda-incubator/setup-miniconda v3 composite

.github/workflows/trymerge.yml actions

actions/checkout v4 composite
actions/setup-python v4 composite
aws-actions/configure-aws-credentials v3 composite
seemethere/upload-artifact-s3 v5 composite

dev-requirements.txt pypi

bitsandbytes * development
blobfile * development
cmake <4.0.0,>=3.19.0 development
diskcache * development
expecttest * development
fire * development
hypothesis * development
importlib_metadata * development
lm_eval * development
matplotlib * development
ninja * development
packaging * development
pandas * development
parameterized * development
pre-commit * development
pycocotools * development
pytest ==9.0.3 development
pytest-xdist ==3.8.0 development
pyyaml * development
ruff ==0.11.6 development
sentencepiece * development
tabulate * development
tiktoken * development
tqdm * development
transformers * development
unittest-xml-reporting * development

pyproject.toml pypi

.github/workflows/regression_test_pt2e_gpu.yml actions

torchao/prototype/attention/fp8_fa3/setup.py pypi

torchao/prototype/attention/shared_utils/setup.py pypi

ecosyste.ms

Data

Tools

Indexes

Applications

Experiments

Summary