OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-06 00:20:18 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity
95,310 Commits 8,304 Branches 150 Tags 1.5 GiB
b3861ac8e7
Commit Graph

671 Commits

Author SHA1 Message Date
soulitzer
b3861ac8e7 [reland] Warn if AccumulateGrad stream does not match producer node stream (#166136)
Some checks failed
docker-builds / docker-build (pytorch-linux-jammy-linter, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3-clang12-executorch, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3-clang12-onnx, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3-clang18-asan, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3-gcc11-inductor-benchmarks, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3.10-clang12, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3.10-gcc11, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3.12-halide, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3.12-triton-cpu, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3.13-clang12, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-py3.14-clang12, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-rocm-n-py3, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-rocm-n-py3-benchmarks, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-xpu-n-1-py3, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-xpu-n-py3, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-jammy-xpu-n-py3-inductor-benchmarks, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-noble-riscv64-py3.12-gcc14, linux.12xlarge) (push) Has been cancelled
Details
docker-builds / docker-build (pytorch-linux-noble-rocm-n-py3, linux.12xlarge) (push) Has been cancelled
Details
ossf-scorecard / Scorecards analysis (push) Has been cancelled
Details
Close nonexistent disable issues / close-nonexistent-disable-issues (push) Has been cancelled
Details
Index PyTorch Tests for Target Determination / get-label-type (push) Has been cancelled
Details
nightly / get-label-type (push) Has been cancelled
Details
nightly / update-commit-hashes (main, .ci/docker/ci_commit_pins, triton, triton-lang) (push) Has been cancelled
Details
nightly / update-commit-hashes (main, .github/ci_commit_pins, audio, pytorch) (push) Has been cancelled
Details
nightly / update-commit-hashes (main, .github/ci_commit_pins, vision, pytorch) (push) Has been cancelled
Details
nightly / update-commit-hashes (main, .github/ci_commit_pins, vllm, vllm-project) (push) Has been cancelled
Details
Index PyTorch Tests for Target Determination / index (push) Has been cancelled
Details
nightly / Link checks (push) Has been cancelled
Details
nightly / docs build (push) Has been cancelled
Details
nightly / docs push (push) Has been cancelled
Details 
ghstack-source-id: 59641aa32dc6fd027abf3276017432b693aa71f8
Pull-Request-resolved: https://github.com/pytorch/pytorch/pull/165065

Fixes #ISSUE_NUMBER

Opening a new PR for codev

Pull Request resolved: https://github.com/pytorch/pytorch/pull/166136
Approved by: https://github.com/ngimel
 2025-11-01 12:33:48 +00:00
Scott Wolchok
7d16fcf2df Re-re-re-re-apply "C++-accessible Placements via pybind11 (#163030)" (#166132) 
Was reverted (again!) due to a merge conflict that crept in sometime during the "export to github -> land internally -> merge on github" process.

D85096233

Pull Request resolved: https://github.com/pytorch/pytorch/pull/166132
Approved by: https://github.com/Skylion007, https://github.com/ezyang, https://github.com/malfet
 2025-10-27 21:19:32 +00:00
PyTorch MergeBot
75b8295868 Revert "Warn if AccumulateGrad stream does not match producer node stream (#165065)" 
This reverts commit 12f742941d.

Reverted https://github.com/pytorch/pytorch/pull/165065 on behalf of https://github.com/clee2000 due to broke internal builds D85273204 usages of TORCH_API void add need to be updated? ([comment](https://github.com/pytorch/pytorch/pull/165065#issuecomment-3438061854))
 2025-10-23 17:02:49 +00:00
Eddie Yan
e64a814ae7 [CUDA] Add experimental green context support for SM carveout (#159104) 
Low-level PyTorch APIs should be usable/stable enough at this point but we might move the underlying driver API usage a bit from here...

Built on top of @drisspg 's branch

Pull Request resolved: https://github.com/pytorch/pytorch/pull/159104
Approved by: https://github.com/ngimel, https://github.com/malfet, https://github.com/kwen2501

Co-authored-by: drisspg <drisspguessous@gmail.com>
Co-authored-by: Nikita Shulga <2453524+malfet@users.noreply.github.com>
 2025-10-22 21:38:52 +00:00
soulitzer
12f742941d Warn if AccumulateGrad stream does not match producer node stream (#165065) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165065
Approved by: https://github.com/ngimel
 2025-10-22 17:33:27 +00:00
Yuanyuan Chen
99c8640b5d [1/N] Change C-style casts to static_cast or reinterpret_cast (#165750) 
This series of changes try to cover C style casts into C++ alternatives.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/165750
Approved by: https://github.com/Skylion007
 2025-10-20 23:27:13 +00:00
PyTorch MergeBot
ab82456c16 Revert "[1/N] Change C-style casts to static_cast or reinterpret_cast (#165750)" 
This reverts commit e1e8491b31.

Reverted https://github.com/pytorch/pytorch/pull/165750 on behalf of https://github.com/pytorch-auto-revert due to Reverted automatically by pytorch's autorevert, to avoid this behaviour add the tag autorevert: disable ([comment](https://github.com/pytorch/pytorch/pull/165750#issuecomment-3422413890))
 2025-10-20 14:51:58 +00:00
Yuanyuan Chen
e1e8491b31 [1/N] Change C-style casts to static_cast or reinterpret_cast (#165750) 
This series of changes try to cover C style casts into C++ alternatives.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/165750
Approved by: https://github.com/Skylion007
 2025-10-20 04:36:19 +00:00
Yuanyuan Chen
032bed95cd Various C++ code fixes in LSAN integration (#165818) 
This PR extracts the C++ code fixes from #154584, which are fixes in enabling LSAN.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165818
Approved by: https://github.com/ezyang
 2025-10-18 17:59:23 +00:00
Nikita Shulga
ce109b3f79 Add torch.backends.mkldnn.is_acl_available() method (#165678) 
That tells whether or not PyTorch was compiled with Arm Compute Library
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165678
Approved by: https://github.com/Skylion007, https://github.com/atalman, https://github.com/albanD
ghstack dependencies: #165583, #165584, #165676
 2025-10-16 22:34:21 +00:00
PyTorch MergeBot
f975bd58af Revert "Warn if AccumulateGrad stream does not match producer node stream (#165065)" 
This reverts commit a70ef954b9.

Reverted https://github.com/pytorch/pytorch/pull/165065 on behalf of https://github.com/izaitsevfb due to breaks lint ([comment](https://github.com/pytorch/pytorch/pull/165065#issuecomment-3391387386))
 2025-10-10 17:29:29 +00:00
soulitzer
a70ef954b9 Warn if AccumulateGrad stream does not match producer node stream (#165065) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165065
Approved by: https://github.com/ngimel
ghstack dependencies: #162815
 2025-10-10 16:46:01 +00:00
FFFrog
5390324984 [CodeClean] Replace std::runtime_error with TORCH_CHECK (#164129) 
As the title stated.

**Changes**:
- torch/csrc/Module.cpp
- torch/csrc/utils.cpp
- torch/csrc/stable
- torch/lib/libshm
Pull Request resolved: https://github.com/pytorch/pytorch/pull/164129
Approved by: https://github.com/albanD
 2025-10-09 19:01:07 +00:00
Simon Layton
6a7f5c0d21 Add scaled_mm python API, test (#164142) 
Summary:

* Add `torch.nn.functional.scaled_mm` as an abstraction around the C++
  methods
* Wraps `torch._scaled_mm_v2` API by default, but user can force use of
  the older `torch._scaled_mm` interface.
* Scaled MM tests now run on the new API

Test Plan:

`pytest test/test_scaled_matmul_cuda.py`

Reviewers:

Subscribers:

Tasks:

Tags:
Signed-off-by: Simon Layton <simonlaytonmeta.com>

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164142
Approved by: https://github.com/drisspg
ghstack dependencies: #164141
 2025-10-09 12:43:18 +00:00
Natalia Gimelshein
37c6087334 Add split-K control to cuBLAS reduced-precision settings (#164766) 
## Summary
- add a CuBLASReductionOption enum so the CUDA context can track reduced-precision and split-K options
- extend the Python bindings, backend helpers, and docs to accept an optional allow_splitk argument for fp16/bf16 matmul controls
- update cuBLAS/cuBLASLt call sites plus dynamo guards and tests to respect the new combinations

## Testing
- python test/test_cuda.py TestCuda.test_cublas_allow_fp16_reduced_precision_reduction_get_set -v *(fails: ModuleNotFoundError: No module named 'psutil')*

------
https://chatgpt.com/codex/tasks/task_e_68e404623178832f8a3e1d34e1e175da

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164766
Approved by: https://github.com/malfet, https://github.com/albanD
 2025-10-08 18:48:45 +00:00
PyTorch MergeBot
df640df68a Revert "Reapply "C++-accessible Placements via pybind11 (#163030)" (#164519)" 
This reverts commit 8c0bc879b9.

Reverted https://github.com/pytorch/pytorch/pull/164519 on behalf of https://github.com/malfet due to Still breaks internal workflows ([comment](https://github.com/pytorch/pytorch/pull/164519#issuecomment-3378469432))
 2025-10-07 19:46:17 +00:00
Scott Wolchok
8c0bc879b9 Reapply "C++-accessible Placements via pybind11 (#163030)" (#164519) 
This makes Placement data representation available in C++ via pybind11. Reapply with fix for internal errors.

D83788896

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164519
Approved by: https://github.com/Skylion007, https://github.com/ezyang
 2025-10-06 23:19:14 +00:00
PyTorch MergeBot
331191ce4b Revert "[BE] Make PyObjectSlot use a global PyInterpreter (#162659)" 
This reverts commit 29cbcbac42.

Reverted https://github.com/pytorch/pytorch/pull/162659 on behalf of https://github.com/izaitsevfb due to reverted internally, see [D83214133](https://www.internalfb.com/diff/D83214133) ([comment](https://github.com/pytorch/pytorch/pull/162659#issuecomment-3369348172))
 2025-10-05 21:39:57 +00:00
Yuanyuan Chen
5103ecc5d8 [1/N] Fix clang-tidy readability checks (#164561) 
Check all `.cpp` files except `jit` files for readability thoroughly.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164561
Approved by: https://github.com/Skylion007
 2025-10-04 09:40:38 +00:00
Lakshay Garg
f006aee601 Speed up FP precision lookup (#164044) 
This commit simplifies the precision lookup and setting logic
by reducing the number of branches and using a custom hash
function. Fixes #161822. The issue described in #163709 still
persists. This is meant as a short term fix.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164044
Approved by: https://github.com/ngimel, https://github.com/eqy
 2025-10-03 21:35:20 +00:00
PyTorch MergeBot
2a7c486750 Revert "Speed up FP precision lookup (#164044)" 
This reverts commit 723ba21393.

Reverted https://github.com/pytorch/pytorch/pull/164044 on behalf of https://github.com/yangw-dev due to broke internal build In file included from xplat/caffe2/aten/src/ATen/DeviceAccelerator.cpp:1: xplat/caffe2/aten/src/ATen/Context.h:502:38: error: shift count >= width of type [-Werror,-Wshift-count-overflow] 502 | return std::hash<size_t>{}((k1 << 32) | k2); ([comment](https://github.com/pytorch/pytorch/pull/164044#issuecomment-3363016702))
 2025-10-02 21:00:44 +00:00
PyTorch MergeBot
f6f7676756 Revert "C++-accessible Placements via pybind11 (#163030)" 
This reverts commit 3e03deab6f.

Reverted https://github.com/pytorch/pytorch/pull/163030 on behalf of https://github.com/swolchok due to doesn't pass pyre ([comment](https://github.com/pytorch/pytorch/pull/163030#issuecomment-3362450379))
 2025-10-02 18:25:24 +00:00
PyTorch MergeBot
c6329524d8 Revert "Add magic TORCH_MAKE_PYBIND_ENUM_FASTER macro (#163527)" 
This reverts commit 50c0550f5a.

Reverted https://github.com/pytorch/pytorch/pull/163527 on behalf of https://github.com/swolchok due to breaking import torch in debug builds, see #164297 ([comment](https://github.com/pytorch/pytorch/pull/163527#issuecomment-3361919142))
 2025-10-02 15:42:42 +00:00
Scott Wolchok
3e03deab6f C++-accessible Placements via pybind11 (#163030) 
This makes Placement data representation available in C++ via pybind11.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163030
Approved by: https://github.com/ezyang
 2025-10-02 02:38:23 +00:00
Lakshay Garg
723ba21393 Speed up FP precision lookup (#164044) 
This commit simplifies the precision lookup and setting logic
by reducing the number of branches and using a custom hash
function. Fixes #161822. The issue described in #163709 still
persists. This is meant as a short term fix.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/164044
Approved by: https://github.com/ngimel, https://github.com/eqy
 2025-10-02 00:59:19 +00:00
Han Qi
b5c4f46bb9 Add functions to setup PrivateUse1 as a python backend device. (#157859) 
Fixes #156052 and #156444.

This PR setup the privateuseone key in Python to be used as a python backend for pytorch.
Meaning that, after calling `setup_privateuseone_for_python_backend('npy')`, one can use a subclass to with that device to hold arbitrary python data as "device data" and use `torch.library` to register ops that takes that Tensor.

Changes done in this PR:

1. Register an vanilla Device Guard: I extended NoOpDeviceGuard to have allow device index of 0 and to not raise errors when event related functions are accessed. If I don't do those, when calling backward I would get errors. (CPU backend uses NoOpDeviceGuard just fine, although there seems to be special treatment of CPU in the autograd engine.
2. Tensor subclass allows not having `__torch_dispatch__` if the device is not CUDA or CPU. The comment of the check suggests it was to avoid segfault when calling into ops that expects a storage. Here we have a different device so will not call into those ops.
3. python function that invokes the other incantations to setup the privateusekey backend.

This took inspiration of https://github.com/bdhirsh/pytorch_open_registration_example and https://github.com/tinygrad/tinygrad/blob/master/extra/torch_backend/wrapped_tensor.cpp; great thanks to @bdhirsh and @geohot.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/157859
Approved by: https://github.com/albanD
 2025-10-01 21:32:59 +00:00
PyTorch MergeBot
410ed3006b Revert "Add functions to setup PrivateUse1 as a python backend device. (#157859)" 
This reverts commit 1310d6a1f9.

Reverted https://github.com/pytorch/pytorch/pull/157859 on behalf of https://github.com/jeanschmidt due to introduce linting errors ([comment](https://github.com/pytorch/pytorch/pull/157859#issuecomment-3352140098))
 2025-09-30 13:24:37 +00:00
Han Qi
1310d6a1f9 Add functions to setup PrivateUse1 as a python backend device. (#157859) 
Fixes #156052 and #156444.

This PR setup the privateuseone key in Python to be used as a python backend for pytorch.
Meaning that, after calling `setup_privateuseone_for_python_backend('npy')`, one can use a subclass to with that device to hold arbitrary python data as "device data" and use `torch.library` to register ops that takes that Tensor.

Changes done in this PR:

1. Register an vanilla Device Guard: I extended NoOpDeviceGuard to have allow device index of 0 and to not raise errors when event related functions are accessed. If I don't do those, when calling backward I would get errors. (CPU backend uses NoOpDeviceGuard just fine, although there seems to be special treatment of CPU in the autograd engine.
2. Tensor subclass allows not having `__torch_dispatch__` if the device is not CUDA or CPU. The comment of the check suggests it was to avoid segfault when calling into ops that expects a storage. Here we have a different device so will not call into those ops.
3. python function that invokes the other incantations to setup the privateusekey backend.

This took inspiration of https://github.com/bdhirsh/pytorch_open_registration_example and https://github.com/tinygrad/tinygrad/blob/master/extra/torch_backend/wrapped_tensor.cpp; great thanks to @bdhirsh and @geohot.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/157859
Approved by: https://github.com/albanD
 2025-09-30 08:39:36 +00:00
Scott Wolchok
50c0550f5a Add magic TORCH_MAKE_PYBIND_ENUM_FASTER macro (#163527) 
See comment on the macro definition. In short, pybind11 3.x
added `py::native_enum`, and also had to add overhead for that new way
to bind enums on the critical path for calling functions that take
regular old `py::enum_`s as arguments (for example, `__eq__`).

Differential Revision: [D82873169](https://our.internmc.facebook.com/intern/diff/D82873169/)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163527
Approved by: https://github.com/ezyang
 2025-09-26 17:59:22 +00:00
PyTorch MergeBot
00059db034 Revert "[RELAND] Always build USE_DISTRIBUTED (#160449) and Make distributed modules importable even when backend not built (#159889) (#162594)" 
This reverts commit 09cb34c1dc.

Reverted https://github.com/pytorch/pytorch/pull/162594 on behalf of https://github.com/malfet due to reverted internally and now can be safely reverted in OSS ([comment](https://github.com/pytorch/pytorch/pull/162594#issuecomment-3334176367))
 2025-09-25 13:47:46 +00:00
Brian Hirsh
7d710403b0 Reapply "Make functionalization ViewMeta serializable with pickle. (#143712)" (#163769) 
### Summary:
NOTE: This is a re-export of https://github.com/pytorch/pytorch/pull/161994 ; the changes between these two PRs is exclusively to the buck/build files

(Summary from #161994 )
Attempted rebase of https://github.com/pytorch/pytorch/pull/143712.

This reverts commit 6c713ccb5e.

cc voznesenskym penguinwu EikanWang jgong5 Guobing-Chen XiaobingSuper zhuhaozhe blzheng wenzhe-nrv jiayisunx chenyang78 kadeng chauhang amjames Lucaskabela

imported-using-ghimport

Test Plan: Imported from OSS

Differential Revision: D81524507

Pulled By: Lucaskabela

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163769
Approved by: https://github.com/dolpm

Co-authored-by: Brian Hirsh <hirsheybar@fb.com>
 2025-09-25 10:27:37 +00:00
PaliC
29cbcbac42 [BE] Make PyObjectSlot use a global PyInterpreter (#162659) 
This pr gets rid of the pyobj_interpreter_ variable from PyObjectSlot and saves a word in the process

Gonna ask for review from @huydhn as there are some changes to CI.

Testing: imported internally and the failed android build seems to work now!

Pull Request resolved: https://github.com/pytorch/pytorch/pull/162659
Approved by: https://github.com/albanD, https://github.com/huydhn
 2025-09-25 08:53:19 +00:00
Edward Yang
09cb34c1dc [RELAND] Always build USE_DISTRIBUTED (#160449) and Make distributed modules importable even when backend not built (#159889) (#162594) 
Summary:
Original: D81957844 and D81957923

Also, https://github.com/pytorch/pytorch/pull/162142 is patched in as well

#buildall

Test Plan:
sandcastle and oss ci

Rollback Plan:

Reviewed By: H-Huang

Pull Request resolved: https://github.com/pytorch/pytorch/pull/162594
Approved by: https://github.com/H-Huang, https://github.com/dcci
 2025-09-22 21:12:18 +00:00
PyTorch MergeBot
edafc902d7 Revert "[BE] Make PyObjectSlot use a global PyInterpreter (#162659)" 
This reverts commit d1993c27ae.

Reverted https://github.com/pytorch/pytorch/pull/162659 on behalf of https://github.com/wdvr due to reverted internally, please see D82771705 @PaliC ([comment](https://github.com/pytorch/pytorch/pull/162659#issuecomment-3317110247))
 2025-09-22 06:22:37 +00:00
PyTorch MergeBot
f0078941cf Revert "[RELAND] Always build USE_DISTRIBUTED (#160449) and Make distributed modules importable even when backend not built (#159889) (#162594)" 
This reverts commit 6c334885d4.

Reverted https://github.com/pytorch/pytorch/pull/162594 on behalf of https://github.com/wdvr due to reverted internally - @ezyang see D82281294 ([comment](https://github.com/pytorch/pytorch/pull/162594#issuecomment-3317017530))
 2025-09-22 05:39:07 +00:00
Sherlock Huang
033b7d1e1a [Reland] Return NoOpDeviceGuardImpl in replace of CudaDeviceGuard when device is not available (#163187) 
Reland of #160532

Summary:

To support exporting a cuda model on a CPU-only machine under fake tensor mode. User commonly need to move sample inputs to the cuda device with .to("cuda:0") or .to("cuda") call. This diff supports this.
I expect the following pattern to work
```
with FakeTensorMode(allow_non_fake_inputs=True):
    cuda_module = module.to("cuda:0")
    cuda_sample_inputs = tuple([x.to("cuda:0") for x in sample_inputs])
    with torch.no_grad():
        ep = torch.export.export(cuda_module, cuda_sample_inputs)
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163016
Approved by: https://github.com/huydhn

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163187
Approved by: https://github.com/angelayi
 2025-09-18 04:46:26 +00:00
Sahan Paliskara
d1993c27ae [BE] Make PyObjectSlot use a global PyInterpreter (#162659) 
This pr gets rid of the pyobj_interpreter_ variable from PyObjectSlot and saves a word in the process

Gonna ask for review from @huydhn as there are some changes to CI.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/162659
Approved by: https://github.com/albanD, https://github.com/huydhn
 2025-09-17 16:40:55 +00:00
PyTorch MergeBot
79fd497423 Revert "[Reland] Return NoOpDeviceGuardImpl in replace of CudaDeviceGuard when device is not available, or cpu-only build (#163016)" 
This reverts commit f1eb99e2e4.

Reverted https://github.com/pytorch/pytorch/pull/163016 on behalf of https://github.com/jeffdaily due to broke rocm CI, see export/test_export_opinfo.py::TestExportOnFakeCudaCUDA::test_fake_export_nonzero_cuda_float32 [GH job link](https://github.com/pytorch/pytorch/actions/runs/17787208381/job/50564369696) [HUD commit link](f1eb99e2e4) ([comment](https://github.com/pytorch/pytorch/pull/163016#issuecomment-3303707552))
 2025-09-17 16:17:53 +00:00
Sherlock Huang
f1eb99e2e4 [Reland] Return NoOpDeviceGuardImpl in replace of CudaDeviceGuard when device is not available, or cpu-only build (#163016) 
Reland of #160532

Summary:

To support exporting a cuda model on a CPU-only machine under fake tensor mode.
User commonly need to move sample inputs to the cuda device with .to("cuda:0") or .to("cuda") call.
This diff supports this.
I expect the following pattern to work
```
with FakeTensorMode(allow_non_fake_inputs=True):
    cuda_module = module.to("cuda:0")
    cuda_sample_inputs = tuple([x.to("cuda:0") for x in sample_inputs])
    with torch.no_grad():
        ep = torch.export.export(cuda_module, cuda_sample_inputs)
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/163016
Approved by: https://github.com/huydhn
 2025-09-17 05:01:33 +00:00
PyTorch MergeBot
4db203f875 Revert "[BE] Make PyObjectSlot use a global PyInterpreter (#162659)" 
This reverts commit 05ee8114f8.

Reverted https://github.com/pytorch/pytorch/pull/162659 on behalf of https://github.com/jeanschmidt due to seems to have introduced errors in linting see https://github.com/pytorch/pytorch/actions/runs/17750689989/job/50444910643 ([comment](https://github.com/pytorch/pytorch/pull/162659#issuecomment-3298626136))
 2025-09-16 12:52:57 +00:00
PaliC
05ee8114f8 [BE] Make PyObjectSlot use a global PyInterpreter (#162659) 
This pr gets rid of the pyobj_interpreter_ variable from PyObjectSlot and saves a word in the process

Gonna ask for review from @huydhn as there are some changes to CI.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/162659
Approved by: https://github.com/albanD, https://github.com/huydhn
 2025-09-16 00:37:09 +00:00
PyTorch MergeBot
9c93dc8123 Revert "Return NoOpDeviceGuardImpl in replace of CudaDeviceGuard when device is not available, or cpu-only build (#160532)" 
This reverts commit a956c4ab1c.

Reverted https://github.com/pytorch/pytorch/pull/160532 on behalf of https://github.com/huydhn due to Reverted internally ([comment](https://github.com/pytorch/pytorch/pull/160532#issuecomment-3287745165))
 2025-09-13 07:42:12 +00:00
Sherlock Huang
a956c4ab1c Return NoOpDeviceGuardImpl in replace of CudaDeviceGuard when device is not available, or cpu-only build (#160532) 
Summary:

To support exporting a cuda model on a CPU-only machine under fake tensor mode.
User commonly need to move sample inputs to the cuda device with .to("cuda:0") or .to("cuda") call.
This diff supports this.
I expect the following pattern to work
```
with FakeTensorMode(allow_non_fake_inputs=True):
    cuda_module = module.to("cuda:0")
    cuda_sample_inputs = tuple([x.to("cuda:0") for x in sample_inputs])
    with torch.no_grad():
        ep = torch.export.export(cuda_module, cuda_sample_inputs)
```

Test Plan:
CI

Rollback Plan:

Differential Revision: D80181887

Pull Request resolved: https://github.com/pytorch/pytorch/pull/160532
Approved by: https://github.com/henryoier, https://github.com/ezyang
 2025-09-13 01:50:51 +00:00
Edward Yang
6c334885d4 [RELAND] Always build USE_DISTRIBUTED (#160449) and Make distributed modules importable even when backend not built (#159889) (#162594) 
Summary:
Original: D81957844 and D81957923

Also, https://github.com/pytorch/pytorch/pull/162142 is patched in as well

#buildall

Test Plan:
sandcastle and oss ci

Rollback Plan:

Reviewed By: H-Huang

Pull Request resolved: https://github.com/pytorch/pytorch/pull/162594
Approved by: https://github.com/H-Huang, https://github.com/dcci
 2025-09-12 10:54:42 +00:00
PyTorch MergeBot
6b59a19242 Revert "[RELAND] Always build USE_DISTRIBUTED (#160449) and Make distributed modules importable even when backend not built (#159889) (#162594)" 
This reverts commit 6e8f17c580.

Reverted https://github.com/pytorch/pytorch/pull/162594 on behalf of https://github.com/huydhn due to Reverted internally ([comment](https://github.com/pytorch/pytorch/pull/162594#issuecomment-3283985880))
 2025-09-12 06:52:03 +00:00
Edward Yang
6e8f17c580 [RELAND] Always build USE_DISTRIBUTED (#160449) and Make distributed modules importable even when backend not built (#159889) (#162594) 
Summary:
Original: D81957844 and D81957923

Also, https://github.com/pytorch/pytorch/pull/162142 is patched in as well

#buildall

Test Plan:
sandcastle and oss ci

Rollback Plan:

Reviewed By: H-Huang

Pull Request resolved: https://github.com/pytorch/pytorch/pull/162594
Approved by: https://github.com/H-Huang, https://github.com/dcci
 2025-09-12 03:56:18 +00:00
Edward Yang
dda071587f Revert "Make distributed modules importable even when backend not built (#159889)" (#162568) 
This reverts commit a0d026688c.

Revert "Always build USE_DISTRIBUTED. (#160449)"

This reverts commit d80297a684.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/162568
Approved by: https://github.com/huydhn
 2025-09-10 04:29:42 +00:00
Edward Yang
d80297a684 Always build USE_DISTRIBUTED. (#160449) 
Signed-off-by: Edward Yang <ezyang@meta.com>

Pull Request resolved: https://github.com/pytorch/pytorch/pull/160449
Approved by: https://github.com/wconstab, https://github.com/albanD, https://github.com/dcci
 2025-09-08 19:10:36 +00:00
PyTorch MergeBot
1e0656f063 Revert "Always build USE_DISTRIBUTED. (#160449)" 
This reverts commit de893e96c7.

Reverted https://github.com/pytorch/pytorch/pull/160449 on behalf of https://github.com/jeanschmidt due to internal changes breaks import checks, see [D81845053](https://www.internalfb.com/diff/D81845053) ([comment](https://github.com/pytorch/pytorch/pull/160449#issuecomment-3264887002))
 2025-09-08 07:04:36 +00:00
Edward Yang
de893e96c7 Always build USE_DISTRIBUTED. (#160449) 
Signed-off-by: Edward Yang <ezyang@meta.com>

Pull Request resolved: https://github.com/pytorch/pytorch/pull/160449
Approved by: https://github.com/wconstab, https://github.com/albanD, https://github.com/dcci
 2025-09-05 20:15:11 +00:00