Skip to content

Pull requests: pytorch/ao

New pull request New
308 Open 3,167 Closed
308 Open 3,167 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

fix fb-built pytorch cannot use fp16 2:4 sparsity (#4280) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported meta-exported module: inference quantize_ api inference flow
#4280 opened Apr 14, 2026 by ZihaoLiu0927 Loading…
5
Fix race condition in mxfp8 CUDA kernels CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4278 opened Apr 14, 2026 by lw Contributor Loading…
1
Add Sparse2x4CUSPARSELTFloat8Tensor (#4277) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported meta-exported module: rocm
#4277 opened Apr 14, 2026 by bbeckca Contributor Loading…
4
[ROCm] Add MXFP8 training support for gfx950 (MI355X) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: rocm
#4275 opened Apr 14, 2026 by indianspeedster Loading…
1
[xpu][mx][test] Enable mx serialization tests on xpu CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4272 opened Apr 13, 2026 by ugolowic Contributor Loading…
1
[xpu][mx] Fix NaN scale propagation in RCEIL triton kernel CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4271 opened Apr 13, 2026 by ugolowic Contributor Loading…
1
[optim] Add GrokAdamW optimizer with low-bit quantization support CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4270 opened Apr 13, 2026 by vaibhavhariram Loading…
1
3
Add torch.uint16, torch.uint32 CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4269 opened Apr 12, 2026 by Freed-Wu Loading…
1
fix(docs): correct image path CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: not user facing Use this tag if you don't want this PR to show up in release notes
#4268 opened Apr 12, 2026 by itsPronay Loading…
5
Fix DTensor errors in PruneOptimizer, add iterative reweighting CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: not user facing Use this tag if you don't want this PR to show up in release notes
#4267 opened Apr 10, 2026 by lisjin Contributor Loading…
12
Add reduce_range to avoid overflow in int8 tensor CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: not user facing Use this tag if you don't want this PR to show up in release notes
#4266 opened Apr 10, 2026 by cyxlily Contributor Loading…
20
Support 32x32 scaling for weights in MXFP8 weight quantization kernel CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: training quantize_ api training flow moe mx
#4254 opened Apr 9, 2026 by alexsamardzic Collaborator Loading…
1
21
[nvfp4 training] add autograd support for NVFP4 emulated grouped GEMM CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4252 opened Apr 9, 2026 by roycho96 Contributor Loading…
1
[xpu][mx] Enable mx matmul tests on xpu CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4251 opened Apr 8, 2026 by ugolowic Contributor Loading…
1
[ROCm] Switch CI to allowlist of validated tests CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: not user facing Use this tag if you don't want this PR to show up in release notes module: rocm topic: rocm
#4248 opened Apr 7, 2026 by brucechanglongxu Contributor Loading…
10
[nvfp4_training] Add Triton kernel for global amax of columnwise RHT (SM90+) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: training quantize_ api training flow nvfp4
#4247 opened Apr 7, 2026 by rdspring1 Loading…
2 tasks
23
Build GPU profile benchmark CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: inference quantize_ api inference flow topic: for developers Use this tag if this PR is mainly developer facing
#4244 opened Apr 6, 2026 by namgyu-youn Contributor Loading…
5 tasks
4
add gptq algorithm diagram CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: not user facing Use this tag if you don't want this PR to show up in release notes
#4237 opened Apr 3, 2026 by vkuzo Contributor Loading…
6
[test] Refactor test_quant_api to use shared ToyTwoLinearModel CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4236 opened Apr 3, 2026 by abhinav-TB Loading…
4
Add deprecation warning for BlockSparseWeightConfig (#4230) CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: deprecation Use this tag if this PR deprecates a feature
#4232 opened Apr 3, 2026 by jerryzh168 Contributor Loading…
1
fix blockwise FP8 scaled_mm scale layout in Float8BlockwiseLinear CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: training quantize_ api training flow
#4229 opened Apr 2, 2026 by iamzainhuda Contributor Loading… FP8 Blockwise Training
@iamzainhuda
6
Fix #4106: ./test/float8/test_everything_multi_gpu.sh failing in 4xH1... CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
#4222 opened Apr 2, 2026 by JiwaniZakir Loading…
1
6
Add aten.view.dtype dispatch to optim state subclasses CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: training quantize_ api training flow
#4216 opened Apr 1, 2026 by winglian Loading…
7
[xpu][mx][test] Enable test_mx_tensor on xpu CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: training quantize_ api training flow xpu Intel XPU related features
#4214 opened Apr 1, 2026 by ugolowic Contributor Loading…
4
[xpu][mx][test] Enable test_kernels on xpu CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. module: training quantize_ api training flow xpu Intel XPU related features
#4213 opened Apr 1, 2026 by ugolowic Contributor Loading…
4
ProTip! Adding no:label will show everything without a label.