OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-06 12:20:52 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity
66,419 Commits 8,304 Branches 150 Tags 1.5 GiB
25fb88cf23
Commit Graph

64 Commits

Author SHA1 Message Date
albanD
25fb88cf23 Add all 3.12 binary build for wheel. Let's see how it goes. V2 (#112882) 
Fixes #ISSUE_NUMBER

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112882
Approved by: https://github.com/malfet, https://github.com/sammcj
 2023-11-16 18:20:12 +00:00
Huy Do
dd957138ec Pin Docker images to main (#112692) 
This will help prevent a commit like 77901321d9 pushing to release branch from overwrite the Docker images used in main.  In addition, the `DEFAULT_TAG` can be easily updated to `2.1` for example when doing branch cut release.  This basically pins the Docker images like https://github.com/pytorch/pytorch/pull/111971

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112692
Approved by: https://github.com/malfet
 2023-11-02 17:39:45 +00:00
Nikita Shulga
0ce8cf7c7a Update small wheel nccl-version to 2.19.3 (#112293) 
To keep it in sync with https://github.com/pytorch/pytorch/pull/110827

Added check to `scripts/generate_binary_build_matrix.py` to validate submodule and small wheel nccl versions are the same

Step one in addressing https://github.com/pytorch/pytorch/issues/112285
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112293
Approved by: https://github.com/huydhn
 2023-10-31 01:20:01 +00:00
atalman
f9053877b4 Add pypi required metadata to all wheels except linux (#111042) 
Will fix package after publishing https://github.com/pytorch/pytorch/issues/100974
Poetry install requires all wheels on pypi to have same metadata. Hence including linux dependencies in all non-linux wheels

Pull Request resolved: https://github.com/pytorch/pytorch/pull/111042
Approved by: https://github.com/malfet
 2023-10-12 17:40:13 +00:00
Jithun Nair
86a9534165 Upgrade nightly wheels to rocm5.7 (#109571) 
Follow-up to https://github.com/pytorch/builder/pull/1541

Pull Request resolved: https://github.com/pytorch/pytorch/pull/109571
Approved by: https://github.com/ezyang
 2023-09-21 13:41:23 +00:00
Mike Schneider
861ae39938 [aarch64] Add PT Docker build image for aarch64 (#106881) 
# Changes
* Update `generate_binary_build_matrix.py` for aarch64 to use `pytorch/manylinuxaarch64-builder:cpu` when it is created
* Executed `generate_binary_build_matrix.py` to update `generated-linux-aarch64-binary-manywheel-nightly.yml`

Aarch64 build/test will fail till the new docker image is available for consmption.

Builder PR to build docker image : https://github.com/pytorch/builder/pull/1472

This switches nightly to use the docker build : https://hud.pytorch.org/hud/pytorch/pytorch/nightly/1?per_page=50&name_filter=aarch64
Pull Request resolved: https://github.com/pytorch/pytorch/pull/106881
Approved by: https://github.com/atalman
 2023-08-09 20:28:04 +00:00
blorange-amd
88f119775d Upgrade nightly wheels to rocm5.6 (#105076) 
Tests https://github.com/pytorch/builder/pull/1442

Fixes #104419

Pull Request resolved: https://github.com/pytorch/pytorch/pull/105076
Approved by: https://github.com/atalman
 2023-07-19 13:47:58 +00:00
Nikita Shulga
348dfc1cf3 Update cuDNN to 8.9.2.26 (#104757) 
Companion PR for https://github.com/pytorch/builder/pull/1436
Should fix `cuDNN version incompatibility: PyTorch was compiled  against (8, 9, 2) but found runtime version (8, 8, 1). PyTorch already comes bundled with cuDNN. One option to resolving this error is to ensure PyTorch can find the bundled cuDNN.one possibility is that there is a conflicting cuDNN in LD_LIBRARY_PATH.` error in [manywheel-py3_8-cuda12_1-with-pypi-cudnn-test](https://github.com/pytorch/pytorch/actions/runs/5480628146/jobs/9986843286#step:16:2347)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/104757
Approved by: https://github.com/drisspg, https://github.com/xw285cornell, https://github.com/r-barnes
 2023-07-07 05:46:46 +00:00
Mike Schneider
ec85ab6157 Adding aarch64 wheel CI workflows (#104109) 
Adding Workflows for building aarch64 Linux PyTorch PIP wheels

Updates:
* Created aarch64 template for generated workflows
* Updated generate_ci_workflows.py to include aarch64
* Generated the aarch64 wheel workflow
* added _binary-build-aarch64.yml for building aarch64 wheel
* added _binary-test-aarch64.yml for sanity check of aarch64 wheel
* Updated binary_linux_test.sh to use --extra-index-url for aarch64 till needed aarch64 dependencies are available at https://download.pytorch.org/whl/nightly/cpu

NOTES:
* The build and test workflows are using arm64v8/alpine and quay.io/pypa/manylinux2014_aarch64:latest docker images at this time.
* Conda generated workflow not included at this time and being worked on.

Workflows were successfully tested at https://github.com/xncqr/pytorch/actions/runs/5351891068
Pull Request resolved: https://github.com/pytorch/pytorch/pull/104109
Approved by: https://github.com/malfet, https://github.com/atalman
 2023-06-29 18:58:43 +00:00
pbialecki
72f09faf10 remove CUDA 11.7 builds (#103904) 
CC @atalman @malfet

Pull Request resolved: https://github.com/pytorch/pytorch/pull/103904
Approved by: https://github.com/malfet, https://github.com/atalman
 2023-06-21 18:16:34 +00:00
blorange-amd
7360d0f904 Upgraded nightly wheels to rocm5.5 (#102242) 
Upgraded nightly wheels to rocm5.5

Follow-up to https://github.com/pytorch/builder/pull/1407

Pull Request resolved: https://github.com/pytorch/pytorch/pull/102242
Approved by: https://github.com/malfet
 2023-06-13 01:34:10 +00:00
pbialecki
de354bf53e Replace CUDA 11.7 small pip wheels with 12.1 (#103091) 
CC @atalman
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103091
Approved by: https://github.com/malfet, https://github.com/atalman
 2023-06-12 17:29:20 +00:00
pbialecki
0a5c930499 Re-enable CUDA 12.1 builds for Windows (#100268) 
Related: https://github.com/pytorch/pytorch/pull/98492
This PR enables Windows builds after the needed AMIs are ready.

CC @atalman

Pull Request resolved: https://github.com/pytorch/pytorch/pull/100268
Approved by: https://github.com/atalman, https://github.com/malfet
 2023-04-28 21:11:27 +00:00
Nikita Shulga
3fef633333 Add CUDA-12.1 manywheel build to trunk (#99458) 
As CUDA-11.7 is getting deprecated anyway.

Also, fix the problem when script actually generated the same workflow twice, overriding 11.8 ones with 11.7+11.7-with-pypi

<!--
copilot:poem
-->
### <samp>🤖 Generated by Copilot at 0c6c182</samp>

> _Oh we are the PyTorch crew and we have a job to do_
> _We build and test the manywheel package with CUDA 11.8_
> _So heave away, me hearties, heave away with all your might_
> _We'll smoke the Linux binary and make sure it runs all right_

Pull Request resolved: https://github.com/pytorch/pytorch/pull/99458
Approved by: https://github.com/dagitses, https://github.com/atalman
 2023-04-19 00:13:32 +00:00
pbialecki
ef3ea30eed Add CUDA 12.1 workflows (#98492) 
CC @atalman @malfet
Pull Request resolved: https://github.com/pytorch/pytorch/pull/98492
Approved by: https://github.com/malfet
 2023-04-11 00:33:53 +00:00
atalman
91166ef7e7 Remove rocm python 3.11 restriction (#97818) 
Removes restrictions for rocm 5.4.2 and python 3.11

Pull Request resolved: https://github.com/pytorch/pytorch/pull/97818
Approved by: https://github.com/malfet
 2023-03-29 02:51:13 +00:00
Huy Do
4c0dce50fd [BE] Apply ufmt to run_test and GitHub Python util scripts (#97588) 
This has been bugging me for a while as I'm working on these Python scripts and they are not tracked by ufmt linter.  So I add these script into that linter.

```
[[linter]]
code = 'UFMT'
include_patterns = [
    '.github/**/*.py',
    'test/run_test.py',
```

This change should just work and not break anything as ufmt (black + usort) linter is very safe to use for standalone util scripts.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/97588
Approved by: https://github.com/kit1980
 2023-03-26 04:52:55 +00:00
zhuhong61
73ee4964d3 Add new checks in CI system to verify the built linux pip wheel with cpu-cxx11-abi (#79409) 
We added the linux pip wheel with cpu-cxx11-abi in pytorch/builder, see: https://github.com/pytorch/builder/pull/990 and https://github.com/pytorch/builder/pull/1023

The purpose of this PR is to add new checks in pytorch CI system to verify the linux pip wheel with cpu-cxx11-abi.

Co-authored-by: Zhu Hong <hong.zhu@intel.com>
Co-authored-by: Guo Yejun <yejun.guo@intel.com>

Pull Request resolved: https://github.com/pytorch/pytorch/pull/79409
Approved by: https://github.com/malfet
 2023-02-14 12:59:03 +00:00
Nikita Shulga
89df0e4253 Enable Python-3.11 binary builds across the board (#94430) 
Most of the work is outside of repositories and consists of cloning projects https://github.com/AnacondaRecipes/ and building:
- [typing_extensions](https://github.com/AnacondaRecipes/typing_extensions-feedstock)
- [pyyaml](https://github.com/AnacondaRecipes/pyyaml-feedstock)
- [setuptools](https://github.com/AnacondaRecipes/setuptools-feedstock) v 59.8.0, needed to build `numpy`. Trick here is to add `add_pip_as_python_dependency: off` to ones `.condarc`
- [cython](https://github.com/AnacondaRecipes/cython-feedstock)
- [mkl-service](https://github.com/AnacondaRecipes/mkl-service-feedstock)
- [numpy-base](https://github.com/AnacondaRecipes/numpy-feedstock) (against mkl-2021.4), i.e. add `blas_impl: "mkl"` and `mkl: ">=2021.4.0,<2022.0a0"` to ones `conda_build_config.yaml`
- [mkl_random](https://github.com/AnacondaRecipes/mkl_random-feedstock)
- [mkl_fft](https://github.com/AnacondaRecipes/mkl_fft-feedstock)
- [numpy](https://github.com/AnacondaRecipes/numpy-feedstock)
- [mpmath](https://github.com/AnacondaRecipes/mpmath-feedstock)
- [sympy](https://github.com/AnacondaRecipes/sympy-feedstock)

Anaconda build system is really modern, so in order to be able to build:
- x86 MacOS packages, one need to install Macos 10.10 SDK from 2014, still available at https://github.com/phracker/MacOSX-SDKs/releases and reference it as conda build sysroot, as follows: `CONDA_BUILD_SYSROOT: /Library/Developer/CommandLineTools/SDKs/MacOSX10.10.sdk`
- Windows packages "MSVC v141 - VS 2017 C++ x64/86 build tools (v14.16)" is needed, which likely is still available as Visual Studio component

As well as make a pretty trivial tweak to build rules in cf4fa8900b
Pull Request resolved: https://github.com/pytorch/pytorch/pull/94430
Approved by: https://github.com/seemethere, https://github.com/weiwangmeta, https://github.com/albanD, https://github.com/atalman
 2023-02-10 06:10:27 +00:00
blorange-amd
11f51e798f Upgrade nightly wheels to ROCm5.4.2 (#93090) 
Test PR1225: https://github.com/pytorch/builder/pull/1225

Pull Request resolved: https://github.com/pytorch/pytorch/pull/93090
Approved by: https://github.com/atalman
 2023-02-09 04:53:11 +00:00
atalman
2910695942 Remove cuda 11.6 from nightly (#93979) 
Remove cuda 11.6 from CI replace with 11.7
Following the Release readme here: https://github.com/pytorch/pytorch/blob/master/RELEASE.md#release-compatibility-matrix
Pull Request resolved: https://github.com/pytorch/pytorch/pull/93979
Approved by: https://github.com/Skylion007, https://github.com/clee2000, https://github.com/malfet
 2023-02-02 20:27:19 +00:00
PyTorch MergeBot
f5e9c8ce54 Revert "Remove CUDA 11.6 from nightly builds (#93404)" 
This reverts commit c76ac8eef2.

Reverted https://github.com/pytorch/pytorch/pull/93404 on behalf of https://github.com/clee2000 due to breaking lint
 2023-02-02 17:10:01 +00:00
atalman
c76ac8eef2 Remove CUDA 11.6 from nightly builds (#93404) 
Remove CUDA 11.6 from nightly builds.
Following the Release readme here: https://github.com/pytorch/pytorch/blob/master/RELEASE.md#release-compatibility-matrix
Pull Request resolved: https://github.com/pytorch/pytorch/pull/93404
Approved by: https://github.com/malfet
 2023-02-02 14:26:52 +00:00
Nikita Shulga
87b9ab4870 [CI] Add Py-3.11 wheels for all platforms (#93400) 
As python-3.11 is now available on Conda for both MacOS and Windows

Disable dimtorch for Python-3.11 on Windows as its current implementation relies on internal symbols which are not exposed on Windows runtime (and to be frank, not sure why they are exposed on Linux/Mac), see https://github.com/pytorch/pytorch/issues/93854

As with the previous PR, most of the changes are not in PyTorch repo, but in builder, namely:
b71049dcbc
ece340ef7e
b0071ac366

Pull Request resolved: https://github.com/pytorch/pytorch/pull/93400
Approved by: https://github.com/weiwangmeta, https://github.com/atalman
 2023-02-01 19:51:19 +00:00
🌌
b179a097ea Add platform markers for linux x86_64 only extra_install_requires (#93066) 
Like #89924 #91083

#85097 added new extra dependencies on nvidia-*. They are linux x86_64 (GPU) only packages, but were not marked as such, causing issues installing pytorch 1.13 via Poetry (and possibly other tools that follow PyPI's metadata API) on Linux aarch64 systems. This "fixes" the issue by adding the `and platform_machine == 'x86_64'` marker on these dependencies.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/93066
Approved by: https://github.com/malfet
 2023-01-31 22:23:51 +00:00
Nikita Shulga
295fd20eb5 [CI] Add Python-3.11 Linux conda builds (#93186) 
This PR almost a no-op, as most of the logic resides in the builder repo, namely:
6342242c50
8f361d91e1

Remove `conda-forge` channel dependency for test job, but add `malfet` channel for 3.11 testing (as numpy is not in default channel yet)
Build and upload following dependencies to `pytorch-nightly` channel:
```
anaconda copy --to-owner pytorch-nightly malfet/numpy/1.23.5
anaconda copy --to-owner pytorch-nightly malfet/numpy-base/1.23.5
anaconda copy --to-owner pytorch-nightly malfet/mkl-service/2.4.0
anaconda copy --to-owner pytorch-nightly malfet/mkl_random/1.2.2
anaconda copy --to-owner pytorch-nightly malfet/mkl_fft/1.3.1

anaconda copy --to-owner pytorch-nightly malfet/sympy/1.11.1
anaconda copy --to-owner pytorch-nightly malfet/mpmath/1.2.1
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/93186
Approved by: https://github.com/atalman, https://github.com/ZainRizvi
 2023-01-31 20:24:03 +00:00
Wei Wang
345695e8f7 Remove PY37 from binary build matrix (#92919) 
Similar to https://github.com/pytorch/test-infra/pull/1416 but for binary build
Pull Request resolved: https://github.com/pytorch/pytorch/pull/92919
Approved by: https://github.com/atalman
 2023-01-26 01:25:47 +00:00
Syed Tousif Ahmed
66b3325304 Adds more nvidia pypi dependencies (#89944) 
This PR adds more nvidia pypi dependencies for cuda 11.7 wheel. Additionally, it pins cufft version to 10.9.0.58 to resolve https://github.com/pytorch/pytorch/issues/88038

Depends on: https://github.com/pytorch/builder/pull/1196

Pull Request resolved: https://github.com/pytorch/pytorch/pull/89944
Approved by: https://github.com/atalman
 2023-01-04 15:08:08 +00:00
pbialecki
8e55d5831a add cu118 workflows for Windows (#91216) 
CC @atalman

Pull Request resolved: https://github.com/pytorch/pytorch/pull/91216
Approved by: https://github.com/atalman
 2022-12-22 01:11:24 +00:00
pbialecki
2d68cc4bc2 Add cu118 workflows (#90826) 
CC @atalman @malfet

Pull Request resolved: https://github.com/pytorch/pytorch/pull/90826
Approved by: https://github.com/atalman
 2022-12-20 14:34:18 +00:00
Jithun Nair
a8629a1c18 Upgrade nightly wheels to ROCm5.3 (#89101) 
Dependent on PR https://github.com/pytorch/builder/pull/1193

Pull Request resolved: https://github.com/pytorch/pytorch/pull/89101
Approved by: https://github.com/kit1980
 2022-11-24 10:53:22 +00:00
Jacob Hayes
2e358cc98f Add platform markers for linux only extra_install_requires (#88826) 
Fixes #88049

https://github.com/pytorch/pytorch/pull/85097 added new extra dependencies on `nvidia-*`. They are linux (GPU) only packages, but were not marked as such, causing issues installing pytorch 1.13 via Poetry (and possibly other tools that follow PyPI's metadata API) on non-Linux systems. This "fixes" the issue by adding the `; platform_system = 'Linux'` marker on these dependencies, but the main problem of different metadata for different wheels is a [somewhat larger issue](https://github.com/pytorch/pytorch/issues/88049#issuecomment-1302555269).

https://github.com/pytorch/pytorch/pull/85097 used `;` as a delimiter for splitting the different deps, but that is the delimiter used in markers, so I changed to split on `|`.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/88826
Approved by: https://github.com/neersighted, https://github.com/lalmei, https://github.com/malfet
 2022-11-18 14:09:21 +00:00
atalman
a807f1987a Stop cuda-10.2 binary builds (#85873) 
Deprecate cuda 10.2 nightly

Pull Request resolved: https://github.com/pytorch/pytorch/pull/85873
Approved by: https://github.com/malfet
 2022-09-29 15:04:24 +00:00
atalman
3007093007 Add new cudnn buid for linux only (#85549) 
Add new cudnn buid for linux only

New pypi packaged are available only for linux:
https://pypi.org/project/nvidia-cudnn-cu11/
Pull Request resolved: https://github.com/pytorch/pytorch/pull/85549
Approved by: https://github.com/malfet
 2022-09-23 20:21:36 +00:00
atalman
eb94df28c7 Use pip install cu117 (#85097) 
Creates new wheel workflow specific to CUDA 11.7 that does not bundle the cudnn and cublas.

Workflow:
https://github.com/pytorch/pytorch/actions/runs/3094622781

New Package:
manywheel-py3_10-cuda11_7-with-pypi-cudnn | 843 MB

Old Package:
manywheel-py3_10-cuda11_7 | 1.65 GB

Testing workflow:

[manywheel-py3_7-cuda11_7-with-pypi-cudnn-build / build](https://github.com/pytorch/pytorch/actions/runs/3091145546/jobs/5000867662#logs):
```
Bundling without cudnn and cublas.
+ DEPS_LIST=("/usr/local/cuda/lib64/libcudart.so.11.0" "/usr/local/cuda/lib64/libnvToolsExt.so.1" "/usr/local/cuda/lib64/libnvrtc.so.11.2" "/usr/local/cuda/lib64/libnvrtc-builtins.so.11.7" "$LIBGOMP_PATH")
+ DEPS_SONAME=("libcudart.so.11.0" "libnvToolsExt.so.1" "libnvrtc.so.11.2" "libnvrtc-builtins.so.11.7" "libgomp.so.1")
.....
pytorch_extra_install_requirements: nvidia-cuda-runtime-cu11, nvidia-cudnn-cu11, nvidia-cublas-cu11
```

[manywheel-py3_7-cuda11_7-build / build](https://github.com/pytorch/pytorch/actions/runs/3091145546/jobs/5000863250#logs)

```
Bundling with cudnn and cublas.
+ DEPS_LIST=("/usr/local/cuda/lib64/libcudart.so.11.0" "/usr/local/cuda/lib64/libnvToolsExt.so.1" "/usr/local/cuda/lib64/libnvrtc.so.11.2" "/usr/local/cuda/lib64/libnvrtc-builtins.so.11.7" "/usr/local/cuda/lib64/libcudnn_adv_infer.so.8" "/usr/local/cuda/lib64/libcudnn_adv_train.so.8" "/usr/local/cuda/lib64/libcudnn_cnn_infer.so.8" "/usr/local/cuda/lib64/libcudnn_cnn_train.so.8" "/usr/local/cuda/lib64/libcudnn_ops_infer.so.8" "/usr/local/cuda/lib64/libcudnn_ops_train.so.8" "/usr/local/cuda/lib64/libcudnn.so.8" "/usr/local/cuda/lib64/libcublas.so.11" "/usr/local/cuda/lib64/libcublasLt.so.11" "$LIBGOMP_PATH")
+ DEPS_SONAME=("libcudart.so.11.0" "libnvToolsExt.so.1" "libnvrtc.so.11.2" "libnvrtc-builtins.so.11.7" "libcudnn_adv_infer.so.8" "libcudnn_adv_train.so.8" "libcudnn_cnn_infer.so.8" "libcudnn_cnn_train.so.8" "libcudnn_ops_infer.so.8" "libcudnn_ops_train.so.8" "libcudnn.so.8" "libcublas.so.11" "libcublasLt.so.11" "libgomp.so.1")
```

cc: @malfet @ptrblck
Pull Request resolved: https://github.com/pytorch/pytorch/pull/85097
Approved by: https://github.com/malfet
 2022-09-21 16:30:25 +00:00
atalman
5014282792 Removing cuda 11.3 nightly builds (#84866) 
Removing cuda 11.3 nightly builds

Pull Request resolved: https://github.com/pytorch/pytorch/pull/84866
Approved by: https://github.com/weiwangmeta, https://github.com/malfet
 2022-09-15 15:47:36 +00:00
Jeff Daily
2dae93b212 [ROCm] update nightly builds to rocm5.2 (#82353) 
### Description
Updates ROCm arches for nightly binaries to 5.1.1 and 5.2, dropping 5.0.

Corresponding builder PR:
https://github.com/pytorch/builder/pull/1033

### Testing
PR CI with extra label to enable binary builds.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/82353
Approved by: https://github.com/janeyx99, https://github.com/malfet
 2022-08-11 00:33:59 +00:00
albanD
36120ce5c0 Adding 3.11 nightlies for linux PyPi (#82302) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/82302
Approved by: https://github.com/malfet
 2022-08-01 15:01:51 +00:00
pbialecki
99c464ae26 Add CUDA 11.7 workflows (#81095) 
Ref https://github.com/pytorch/builder/issues/1042

CC @atalman

Pull Request resolved: https://github.com/pytorch/pytorch/pull/81095
Approved by: https://github.com/atalman
 2022-07-18 09:03:03 +00:00
Pruthvi Madugundu
e549e97484 Upgrade CI to ROCm 5.1 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/76365
Approved by: https://github.com/seemethere, https://github.com/jeffdaily
 2022-04-28 00:02:07 +00:00
atalman
31a6e6cabc Remove cuda 11.5 builds since we have 11.6 
CUDA 11.5 was experimental, but now that we are moving to 11.6, we don't need to keep 11.5 around.
Please note following PR's where merged for domain libraries
[vision](https://github.com/pytorch/vision/pull/5862)
[audio](https://github.com/pytorch/audio/pull/2346)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/76257
Approved by: https://github.com/seemethere, https://github.com/malfet
 2022-04-23 00:24:58 +00:00
PyTorch MergeBot
1118b157bc Revert "Remove 11.5 experimental builds now that we have 11.6" 
This reverts commit 5d059d20ad.

Reverted https://github.com/pytorch/pytorch/pull/75749 on behalf of https://github.com/janeyx99
 2022-04-14 00:54:22 +00:00
Jane Xu
5d059d20ad Remove 11.5 experimental builds now that we have 11.6 
CUDA 11.5 was experimental, but now that we are hoping to jump to 11.6, we don't need to keep 11.5 around.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/75749
Approved by: https://github.com/seemethere
 2022-04-13 20:28:52 +00:00
Jeff Daily
d38d950f22 [ROCm] libtorch nightly now correctly uses rocm runners 
Fixes nightly rocm libtorch builds.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/74992
Approved by: https://github.com/seemethere, https://github.com/malfet
 2022-04-12 16:41:34 +00:00
Andrey Talman
e7f4f5dd9b CUDA 11.6 workflows (#75518) 
Summary:
Adding CUDA 11.6 workflows .
Please note we still depend on conda-forge for cuda 11.6.
Issue created to remove conda-forge dependency: [75532](https://github.com/pytorch/pytorch/issues/75532)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/75518

Reviewed By: janeyx99

Differential Revision: D35516057

Pulled By: atalman

fbshipit-source-id: 44a3a0f8954d98adca2280b2e9f203267ebe98cd
(cherry picked from commit 97a4e52ecee8540453e2871714275796dc1c4abb)
 2022-04-11 14:05:41 +00:00
Jeff Daily
92f01e19f3 [ROCm] libtorch nightly builds 
Enables libtorch builds for ROCm, starting with ROCm version 4.5.2.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/57506
Approved by: https://github.com/malfet
 2022-03-29 20:36:55 +00:00
Nikita Shulga
9929a9fc8f [GHA] Migrate win/linux binary-smoke workflows from CircleCI 
Partially fixes https://github.com/pytorch/pytorch/issues/72956
Pull Request resolved: https://github.com/pytorch/pytorch/pull/73788
Approved by: https://github.com/seemethere
 2022-03-07 15:22:23 +00:00
PyTorch MergeBot
ea698c148a Revert "[GHA] Migrate win/linux binary-smoke workflows from CircleCI" 
This reverts commit 486bd9f306.

Reverted https://github.com/pytorch/pytorch/pull/73788 on behalf of https://github.com/malfet
 2022-03-06 19:07:04 +00:00
Nikita Shulga
486bd9f306 [GHA] Migrate win/linux binary-smoke workflows from CircleCI 
Partially fixes https://github.com/pytorch/pytorch/issues/72956
Pull Request resolved: https://github.com/pytorch/pytorch/pull/73788
Approved by: https://github.com/seemethere
 2022-03-06 17:08:08 +00:00
Nikita Shulga
bebfdca093 Re-enable Windows debug libtorch 
Get rid of `BUILD_FOR_SYSTEM` environment variable
Pass `libtorch_config` environment variable for Windows builds

Fixes https://github.com/pytorch/pytorch/issues/73068

Pull Request resolved: https://github.com/pytorch/pytorch/pull/73805
Approved by: https://github.com/seemethere
 2022-03-04 23:49:07 +00:00