OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2026-01-15 12:15:51 +00:00
Code Issues Actions 51 Packages Projects Releases Wiki Activity
97,544 Commits 8,304 Branches 150 Tags
bfa6f5e0730dead84017e779e02de6cea768ee33
Commit Graph

1132 Commits

Author SHA1 Message Date
Nikita Shulga
bfa6f5e073 [CUDA] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412) 
Opening this PR for testing...

Note that we are proposing 9.15 instead of 9.16 as we have not had sufficient signal on 9.16 internally

NS: Added hacky workaround to install 9.15.1 for torchbench testing

Pull Request resolved: https://github.com/pytorch/pytorch/pull/169412
Approved by: https://github.com/atalman, https://github.com/malfet

Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com>
Co-authored-by: Ting Lu <tingl@nvidia.com>
 2025-12-23 04:55:48 +00:00
PyTorch MergeBot
2a9a1cc632 Revert "[CUDA] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412)" 
This reverts commit aadd016020.

Reverted https://github.com/pytorch/pytorch/pull/169412 on behalf of https://github.com/huydhn due to Sorry for reverting the change but there seems to cause an import error running vLLM tests ([comment](https://github.com/pytorch/pytorch/pull/169412#issuecomment-3684592758))
 2025-12-23 00:07:05 +00:00
Ivan Zaitsev
49f600e864 Remove concurrency limits in workflows for workflow_dispatches (#171132) 
Autorevert can issue multiple dispatches without waiting for the last one to finish:
https://github.com/pytorch/pytorch/actions/workflows/pull.yml?query=branch%3Atrunk%2Faadd016020d718ae862361d23d98f61a5e6e3903
(this is expected behavior in certain cases, e.g. the specific job was already finished, but not the whole workflow)

But currently in pytorch workflows the concurrency policy cancels concurrent workflow runs, even if they are dispatches.

This PR:
1. removes the limit for dispatches (for the workflows that are monitored by autorevert). Note: there is still a hard cap for the total number of dispatches on autorevert side.

2. adds logging, so in the future we can change the concurrency to apply only to autorevert dispatches (we'll know what correct `actor` value to use)

3. removes garbage from the key in linux-aarch64.yml wf

----

Testing:

see my two manual concurrent dispatches here:
https://github.com/pytorch/pytorch/actions/workflows/pull.yml?query=branch%3Aunlimited-dispatches++
(also notice that concurrency correctly cancels wf on PR update)

new logging:
https://github.com/pytorch/pytorch/actions/runs/20444849087/job/58745963215#step:2:20
Pull Request resolved: https://github.com/pytorch/pytorch/pull/171132
Approved by: https://github.com/clee2000, https://github.com/jeanschmidt
 2025-12-22 22:24:02 +00:00
Nikita Shulga
aadd016020 [CUDA] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412) 
Opening this PR for testing...

Note that we are proposing 9.15 instead of 9.16 as we have not had sufficient signal on 9.16 internally

NS: Added hacky workaround to install 9.15.1 for torchbench testing

Pull Request resolved: https://github.com/pytorch/pytorch/pull/169412
Approved by: https://github.com/atalman, https://github.com/malfet

Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com>
Co-authored-by: Ting Lu <tingl@nvidia.com>
 2025-12-22 19:51:33 +00:00
PyTorch MergeBot
70e416db71 Revert "[CUDA] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412)" 
This reverts commit 83c187fe2f.

Reverted https://github.com/pytorch/pytorch/pull/169412 on behalf of https://github.com/yf225 due to Seems inductor tests still have error 83c187fe2f (58591620665-box), we might need to revert it again ([comment](https://github.com/pytorch/pytorch/pull/169412#issuecomment-3677616096))
 2025-12-20 08:55:16 +00:00
Eddie Yan
83c187fe2f [CUDA] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412) 
Opening this PR for testing...

Note that we are proposing 9.15 instead of 9.16 as we have not had sufficient signal on 9.16 internally

NS: Added hacky workaround to install 9.15.1 for torchbench testing

Pull Request resolved: https://github.com/pytorch/pytorch/pull/169412
Approved by: https://github.com/atalman, https://github.com/malfet

Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com>
 2025-12-20 00:13:28 +00:00
PyTorch MergeBot
ac9943d380 Revert "[CUDA][cuDNN] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412)" 
This reverts commit b383f6e269.

Reverted https://github.com/pytorch/pytorch/pull/169412 on behalf of https://github.com/malfet due to Broke inductor tests, see 1d2512a56a/1 ([comment](https://github.com/pytorch/pytorch/pull/169412#issuecomment-3676924457))
 2025-12-19 22:50:43 +00:00
Eddie Yan
b383f6e269 [CUDA][cuDNN] Upgrade cuDNN to 9.15.1 for CUDA 13 builds (#169412) 
Opening this PR for testing...

Note that we are proposing 9.15 instead of 9.16 as we have not had sufficient signal on 9.16 internally

Pull Request resolved: https://github.com/pytorch/pytorch/pull/169412
Approved by: https://github.com/atalman, https://github.com/malfet
 2025-12-19 18:48:52 +00:00
Jithun Nair
19847256b4 [ROCm][CI] Remove ciflow/rocm and ciflow/inductor-rocm triggers (#170797) 
Use only the arch-specific one to prevent double triggering the workflows if both ciflow labels are added. For eg. `ciflow/rocm` was being added by default for any PRs matching "rocm" in the title (until that was replaced with `ciflow/rocm-mi300` in https://github.com/pytorch/test-infra/pull/7578). If a dev would add `ciflow/rocm-mi300` on the same PR, it would double-trigger the same workflow rocm-mi300.yml.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/170797
Approved by: https://github.com/jeffdaily
 2025-12-18 20:42:20 +00:00
albanD
efb64175dc assert removal in ci, github, numpy ref, wo, backends, nn and onnx (#170328) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/170328
Approved by: https://github.com/justinchuby, https://github.com/liangel-02
ghstack dependencies: #170327
 2025-12-16 22:41:29 +00:00
atalman
da032315a5 Unpin patch version of nvidia-cuda-runtime for CUDA 13.0 builds (#170331) 
Fixes https://github.com/pytorch/pytorch/issues/170286

CUDA 13.0.48 is the first one available on https://download.pytorch.org/whl/nightly/nvidia-cuda-runtime/:
<img width="783" height="207" alt="Screenshot 2025-12-12 at 4 13 23 PM" src="https://github.com/user-attachments/assets/4f505453-d75b-4c51-90fa-e36d31f6e05d" />

Pull Request resolved: https://github.com/pytorch/pytorch/pull/170331
Approved by: https://github.com/malfet, https://github.com/yangw-dev
 2025-12-13 00:30:43 +00:00
Ke Wen
47da33f68a Consolidate NCCL version for different CUDA; 2.28.9 (#169927) 
Upgrade NCCL to 2.28.9 (for both CUDA 12 and 13).
Pull Request resolved: https://github.com/pytorch/pytorch/pull/169927
Approved by: https://github.com/Skylion007, https://github.com/atalman
 2025-12-10 22:12:45 +00:00
atalman
06b03073b5 Rename pytorch-triton package to triton (#169888) 
Fixes https://github.com/pytorch/pytorch/issues/163963
Makes naming more consistent with official triton release

Pull Request resolved: https://github.com/pytorch/pytorch/pull/169888
Approved by: https://github.com/malfet, https://github.com/njriasan
 2025-12-09 20:57:29 +00:00
Huy Do
d3748222db [vLLM] Remove xformers in vLLM build workflow (#169914) 
This is not needed anymore after https://github.com/vllm-project/vllm/pull/29262.  So, we only need to build vLLM wheel on PyTorch CI now.

1. https://github.com/vllm-project/vllm/pull/29262 has removed xformers from vLLM, and this PR applies that change to our end
2. This vLLM nightly wheel will be used to power nightly benchmark runs on PyTorch CI.  It's good that we just need to rebuild vLLM now and none of its dependencies
3. I'm trying to get rid of the Dockerfile on PyTorch eventually, and just use the official one from vLLM instead.  This is a work in progress

Pull Request resolved: https://github.com/pytorch/pytorch/pull/169914
Approved by: https://github.com/zou3519
 2025-12-09 04:38:05 +00:00
Wang, Chuanqi
365a6c84db [BE] Upgrade XPU support package to 2025.3 (#166829) 
Follows #166723. Including below changes,

- Add XPU support package 2025.3 build and test in CI for both Linux and Windows
- Keep XPU support package 2025.2 build in CI to ensure no break issue until PyTorch 2.10 release
- Upgrade XPU support package from 2025.2 to 2025.3 in CD for both Linux and Windows
- Update XPU runtime pypi packages dependencies of CD wheels

Pull Request resolved: https://github.com/pytorch/pytorch/pull/166829
Approved by: https://github.com/atalman
 2025-12-05 02:48:23 +00:00
atalman
a36e1d39eb Triton 3.6 pin update (#168096) 
Required for release 2.10

Rocm wheel build fix provided by: https://github.com/pytorch/pytorch/pull/169369

Pull Request resolved: https://github.com/pytorch/pytorch/pull/168096
Approved by: https://github.com/njriasan, https://github.com/malfet, https://github.com/huydhn
 2025-12-04 15:09:20 +00:00
PyTorch MergeBot
fdf863d5e1 Revert "Triton 3.6 pin update (#168096)" 
This reverts commit 93d0d6838c.

Reverted https://github.com/pytorch/pytorch/pull/168096 on behalf of https://github.com/atalman due to Causes timeouts https://github.com/pytorch/pytorch/issues/169492 ([comment](https://github.com/pytorch/pytorch/pull/168096#issuecomment-3609092057))
 2025-12-03 22:23:29 +00:00
atalman
93d0d6838c Triton 3.6 pin update (#168096) 
Required for release 2.10

Rocm wheel build fix provided by: https://github.com/pytorch/pytorch/pull/169369

Pull Request resolved: https://github.com/pytorch/pytorch/pull/168096
Approved by: https://github.com/njriasan, https://github.com/malfet
 2025-12-02 17:28:48 +00:00
Wei Wang
ce5e7e3bf1 [CI][CUDA][Distributed] Update NCCL to 2.28.9 for CUDA13 (#168091) 
This PR updates the NCCL version for CUDA13 from 2.27.7 to 2.28.9.

2.28.9 release notes: https://github.com/NVIDIA/nccl/releases/tag/v2.28.9-1
2.28.7 release notes: https://github.com/NVIDIA/nccl/releases/tag/v2.28.7-1
2.28.3 release notes: https://github.com/NVIDIA/nccl/releases/tag/v2.28.3-1

CUDA 12 remains at 2.27.5 and is untouched by this PR.

Reference PR: https://github.com/pytorch/pytorch/pull/166174
Pull Request resolved: https://github.com/pytorch/pytorch/pull/168091
Approved by: https://github.com/atalman
 2025-12-02 01:09:40 +00:00
Huy Do
cddec6562e [mergebot] Set header for workflow when calling Dr.CI (#169037) 
Same spirit as https://github.com/pytorch/test-infra/pull/7513
Pull Request resolved: https://github.com/pytorch/pytorch/pull/169037
Approved by: https://github.com/clee2000, https://github.com/yangw-dev
 2025-12-01 19:56:43 +00:00
Yuanyuan Chen
f47dd0ddef Enable SIM118 (#167399) 
This PR enables the `SIM118` rule of ruff, which checks for key-existence checks against dict.keys() calls.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/167399
Approved by: https://github.com/albanD
 2025-11-28 08:00:09 +00:00
Huy Do
6461548b4d [vLLM] Update xformers and remove flashinfer-python (#168141) 
A couple of changes:
* Update `xformers==0.0.33.post1`.  This is the latest version for 2.9 release
* Remove `flashinfer-python` build, we don't need to compile it anymore after https://github.com/vllm-project/vllm/pull/26443.  This is now a regular dependency for vLLM
* I also switch the base image to 12.9.1 to match what is vLLM is using nowadays

### Testing

https://github.com/pytorch/pytorch/actions/runs/19490188972/job/55780754518
Pull Request resolved: https://github.com/pytorch/pytorch/pull/168141
Approved by: https://github.com/yangw-dev
 2025-11-20 00:00:43 +00:00
Nikita Shulga
cea86781f2 [CD] Add cuda-bindings dependency to CUDA wheels (#167769) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/167769
Approved by: https://github.com/ngimel, https://github.com/leofang
 2025-11-19 04:31:00 +00:00
linhaifeng
cfb3a6b3da [2/N][BugFix][Refactor] fix several instances which use f = open(...) without a corresponding f.close() (#167628) 
continue in https://github.com/pytorch/pytorch/pull/167423

Pull Request resolved: https://github.com/pytorch/pytorch/pull/167628
Approved by: https://github.com/cyyever, https://github.com/Skylion007
 2025-11-13 21:15:45 +00:00
albanD
52231a7974 show current env before running lint (#166860) 
There seems to be some discrepency between CI and local for Pyrefly so logging these to be able to check for different dependency versions
Pull Request resolved: https://github.com/pytorch/pytorch/pull/166860
Approved by: https://github.com/janeyx99
 2025-11-11 05:28:31 +00:00
Yuanyuan Chen
5135ace3a3 Enable ruff UP035 rule (#167307) 
This PR enables `UP035` rule of ruff.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/167307
Approved by: https://github.com/Lucaskabela
 2025-11-09 06:40:03 +00:00
atalman
cc8bfd1206 Docker release build: Use 13.0.0 nvidia docker (#166904) 
Forward fix for failing Docker release builds
Related to: https://github.com/pytorch/pytorch/issues/166897

Nightly Docker build failure https://github.com/pytorch/pytorch/actions/runs/18900508440/job/53946606434
Due to missing base image:
```
ERROR: failed to build: failed to solve: docker.io/nvidia/cuda:13.0.2-devel-ubuntu22.04: not found
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/166904
Approved by: https://github.com/tinglvv, https://github.com/malfet
 2025-11-04 13:58:10 +00:00
Jeff Daily
239e7b541a [ROCm][CI] upgrade nightly wheels to ROCm 7.1 (#166730) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/166730
Approved by: https://github.com/jeffdaily

Co-authored-by: Jeff Daily <jeff.daily@amd.com>
 2025-10-31 17:30:47 +00:00
Xuehai Pan
69be99ee51 Remove manually synced arch versions in tools/nightly.py (#166616) 
Discussed with @atalman offline. To reduce duplicate changes and reduce the number of files to change when updating arch versions.

------

Pull Request resolved: https://github.com/pytorch/pytorch/pull/166616
Approved by: https://github.com/ezyang
 2025-10-31 15:11:28 +00:00
Aaron Gokaslan
96b61844a7 [BE]: Update nvshmem to 3.4.5 (#164046) 
Release notes can be found here: https://docs.nvidia.com/nvshmem/release-notes-install-guide/release-notes/release-3405.html main difference is the addition of a CPU assisted IBGDA fallback which should allow NVSHMEM IBGDA to work on way more systems without admin intervention and without using GDRCopy.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164046
Approved by: https://github.com/ezyang, https://github.com/kwen2501
 2025-10-29 07:32:05 +00:00
Ting Lu
544b443ea1 [CD] Upgrade to CUDA 13.0.2 for nightly binaries (#165470) 
13.0.U2 is posted, adding to nightlies
Why we want to upgrade: CUDA 13.0.U2 included a new release from cuBLAS that
1. Enabled opt-in fixed-point emulation for FP64 matmuls (D/ZGEMM) which improves performance and power-efficiency.
2. Improved performance on NVIDIA [DGX Spark](https://www.nvidia.com/en-us/products/workstations/dgx-spark/) for FP16/BF16 and FP8 GEMMs.
3. adds BF16x9 FP32 emulation support for SYRK and HERK routines.
Reference: https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html#cublas-release-13-0-update-2

Pull Request resolved: https://github.com/pytorch/pytorch/pull/165470
Approved by: https://github.com/atalman
 2025-10-28 15:14:43 +00:00
PyTorch MergeBot
74336f8c77 Revert "[CD] Upgrade to CUDA 13.0.2 for nightly binaries (#165470)" 
This reverts commit 5e769ff867.

Reverted https://github.com/pytorch/pytorch/pull/165470 on behalf of https://github.com/atalman due to Sorry reverting for now, to restore trunk health ([comment](https://github.com/pytorch/pytorch/pull/165470#issuecomment-3454166879))
 2025-10-28 02:21:48 +00:00
Ting Lu
5e769ff867 [CD] Upgrade to CUDA 13.0.2 for nightly binaries (#165470) 
13.0.U2 is posted, adding to nightlies
Why we want to upgrade: CUDA 13.0.U2 included a new release from cuBLAS that
1. Enabled opt-in fixed-point emulation for FP64 matmuls (D/ZGEMM) which improves performance and power-efficiency.
2. Improved performance on NVIDIA [DGX Spark](https://www.nvidia.com/en-us/products/workstations/dgx-spark/) for FP16/BF16 and FP8 GEMMs.
3. adds BF16x9 FP32 emulation support for SYRK and HERK routines.
Reference: https://docs.nvidia.com/cuda/cuda-toolkit-release-notes/index.html#cublas-release-13-0-update-2

Pull Request resolved: https://github.com/pytorch/pytorch/pull/165470
Approved by: https://github.com/atalman
 2025-10-28 00:21:47 +00:00
Huy Do
9095a9dfae [CD] Apply the fix from #162455 to aarch64+cu129 build (#165794) 
When trying to bring cu129 back in https://github.com/pytorch/pytorch/pull/163029, I mainly looked at https://github.com/pytorch/pytorch/pull/163029 and missed another tweak coming from https://github.com/pytorch/pytorch/pull/162455

I discover this issue when testing aarch64+cu129 builds in https://github.com/pytorch/test-infra/actions/runs/18603342105/job/53046883322?pr=7373.  Surprisingly, there is no test running for aarch64 CUDA build from what I see in 79a37055e7.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165794
Approved by: https://github.com/malfet
 2025-10-18 04:16:24 +00:00
Yuanyuan Chen
e925dfcc6b Enable all SIM rules except disabled ones (#164645) 
`SIM` rules are useful for simplifying boolean expressions and enhances code readability.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164645
Approved by: https://github.com/ezyang, https://github.com/mlazos
 2025-10-17 07:27:11 +00:00
Huy Do
6dedd34c31 [CD] Skip 12.9 build on Windows (#165665) 
Per title

Pull Request resolved: https://github.com/pytorch/pytorch/pull/165665
Approved by: https://github.com/Camyll, https://github.com/malfet
 2025-10-16 19:11:27 +00:00
Catherine Lee
64699b8042 [trymerge] Do not check for rules when reverting (#165342) 
Why do we need to check for merge rules when reverting?
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165342
Approved by: https://github.com/malfet
 2025-10-13 19:07:00 +00:00
Catherine Lee
684df93975 [CI] Default keep-going true for tags of form ciflow/something/commitsha (#165180) 
Tags of the form `ciflow/something/commitsha` are usually created by running the workflow from HUD

Pull Request resolved: https://github.com/pytorch/pytorch/pull/165180
Approved by: https://github.com/huydhn
 2025-10-13 16:12:37 +00:00
Huy Do
4400c5d31e Continue to build nightly CUDA 12.9 for internal (#163029) 
Revert part of https://github.com/pytorch/pytorch/pull/161916 to continue building CUDA 12.9 nightly

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163029
Approved by: https://github.com/malfet
 2025-10-11 08:26:47 +00:00
Nikita Shulga
90c0825e2d [GHF] Allow reverts from pytorch-auto-revert app (#164911) 
This is a bit weird, but author_login is not a unique field, but author_url is.

Explicitly allow https://github.com/apps/pytorch-auto-revert to issue revert commands

Update mocks by running
```
sed -i -e s/8e262b0495bd934d39dda198d4c09144311c5ddd6cca6a227194bd48dbfe7201/47860a8f57a214a426d1150c29893cbc2aa49507f12b731483b1a1254bca3428/ gql_mocks.json
```

Test plan: Run
```python
from trymerge import GitHubPR
pr=GitHubPR("pytorch", "pytorch", 164660)
print(pr.get_last_comment().author_url, pr.get_comment_by_id(3375785595).author_url)
```
that should produce
```
https://github.com/pytorch-auto-revert https://github.com/apps/pytorch-auto-revert
```
Plus added a regression test that checks two particular comments for revert validity

`pytorch-auto-revert` user is my alter ego :)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164911
Approved by: https://github.com/jeanschmidt
 2025-10-08 15:15:45 +00:00
Wei Wang
773c6762b8 [CD][CUDA13][NCCL] Fix nccl version typo for cu13 (#164383) 
https://pypi.org/project/nvidia-nccl-cu13/#history does not have 2.27.5 but 2.27.7+.
Companion PR: https://github.com/pytorch/pytorch/pull/164352

Fixes a potential binary breakage due to non-existence of referenced NCCL cu13 version.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164383
Approved by: https://github.com/tinglvv, https://github.com/Skylion007, https://github.com/atalman
 2025-10-01 21:32:25 +00:00
Ivan Zaitsev
1288c6d8bb Enable keep-going for trunk tags (#164307) 
Tags like `trunk/{sha}` are used to re-run signals by [autorevert project](https://github.com/pytorch/test-infra/blob/main/aws/lambda/pytorch-auto-revert/README.md).

We need to have `keep-going` enabled for those reruns, so that they surface all test failures, not just the first one.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164307
Approved by: https://github.com/clee2000
 2025-10-01 17:21:43 +00:00
albanD
2610746375 Revert nccl upgrade back to 2.27.5 (#164352) 
Revert https://github.com/pytorch/pytorch/pull/162351 as it breaks H100
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164352
Approved by: https://github.com/atalman, https://github.com/malfet
 2025-10-01 15:27:40 +00:00
Nikita Shulga
5a93f00c79 [CI] Delete binary smoke workflows (#164260) 
Those were very useful in the past, because:
- CI builder jobs did not generates wheels, but rather run `python setup.py develop` and shared docker layers, which is no longer the case, all CI jobs produce wheels
- CD jobs were targeting pre-CXX11 ABI, but this is no longer the case after manylinux2_28 migration

Existing, but acceptable gaps:
 - Windows libtorch debug builds sometimes might fail, but IMO it's ok not to be able to produce those for a few days, as number of libtorch users are somewhat small
 - All CD jobs are based on AlmaLinux, while CI are based on Ubuntu, but this could be adjusted if needed, besides AlmaLinux-9 and Ubuntu-22.04 are pretty close in terms of glibc and gcc versions
 - CD jobs build for all GPU architectures, while CI only for the one being tested, but there are now periodic H100 and B200 jobs, and not a lot of development happens for Voltas or Pascals

Besides there are better tools to alert about the nightly failures

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164260
Approved by: https://github.com/seemethere, https://github.com/atalman
 2025-09-30 20:00:07 +00:00
Aaron Gokaslan
5504a06e01 [BE]: Update NCCL to 2.28.3 (#162351) 
@eqy New NCCL has some a bunch of bugfixes for features including reducing the number SMs needed by NVLINK collectives as well as some very useful new APIs for SymmetricMemory.  Also allows FP8 support for non-reductive operations on pre-sm90 devices.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/162351
Approved by: https://github.com/ezyang, https://github.com/malfet, https://github.com/atalman
 2025-09-28 01:38:59 +00:00
Jeff Daily
f1260c9b9a [ROCm][CI/CD] upgrade nightly wheels to ROCm 7.0 (#163937) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/163937
Approved by: https://github.com/jeffdaily

Co-authored-by: Jeff Daily <jeff.daily@amd.com>
 2025-09-26 21:42:09 +00:00
Nikita Shulga
00f96dd84d [CI] Run CUDA-13 binary builds on trunk (#163787) 
There are numerous other workflows that could be used to catch CUDA-12
build regression (our CI builds are almost identical to CD ones), but not many CUDA-13 builds around, so https://github.com/pytorch/pytorch/issues/163342 are really hard to detect in CI
Pull Request resolved: https://github.com/pytorch/pytorch/pull/163787
Approved by: https://github.com/atalman, https://github.com/huydhn
 2025-09-25 00:58:17 +00:00
Nikita Shulga
52dd7a898c Move ROCM trunk wheel builds to 3.10 (#163339) 
This code is a delicious spaghetti: Sometimes python version is defined in jinja template (see https://github.com/pytorch/pytorch/pull/162297 ) sometimes in shell script (see https://github.com/pytorch/pytorch/pull/162877 ), but this time around it's in a python file (and there is another one called `generate_binary_build_matrix.py` that defines `FULL_PYTHON_VERSIONS`)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163339
Approved by: https://github.com/clee2000
 2025-09-19 18:52:00 +00:00
Huy Do
66133b1ab7 Build vLLM aarch64 nightly wheels (#162664) 
PyTorch has published its aarch64 nightly wheels for all CUDA version after https://github.com/pytorch/pytorch/pull/162364
Pull Request resolved: https://github.com/pytorch/pytorch/pull/162664
Approved by: https://github.com/atalman
 2025-09-13 03:43:55 +00:00
Svetlana Karslioglu
e15686b40d Remove actionable label from docathon label sync script (#155713) 
Make sure we don't propagate actionable label in docathon sync label script.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/155713
Approved by: https://github.com/clee2000
 2025-09-12 15:36:50 +00:00