pytorch

mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00

Files

Jason Ansel 24b501903c Minor sympy usage fix in fbcode (#93171 )

Summary: To supports older versions of sympy.

Test Plan:
```
buck2 run @//mode/opt @//mode/inplace -c python.package_style=inplace -c fbcode.enable_gpu_sections=true //caffe2/benchmarks/dynamo:torchbench -- -dcuda --performance --inductor --only hf_T5
```

Differential Revision: D42812188

Pull Request resolved: https://github.com/pytorch/pytorch/pull/93171
Approved by: https://github.com/eellison

2023-01-30 23:34:22 +00:00

_awaits

[jit] Support Awaitable type (#90863 )

2023-01-30 17:38:59 +00:00

[jit] Support Awaitable type (#90863 )

2023-01-30 17:38:59 +00:00

_C_flatbuffer

…

_decomp

Fix to use upsample_bicubic2d.vec decomp for dynamic shape support (#92854 )

2023-01-25 05:08:02 +00:00

_dispatch

Add crossref debug mode for functionalization, catches stride errors (#89498 )

2022-11-23 04:18:25 +00:00

_dynamo

[Reland] Refactor stack_trace preservation for node meta preservation (#90803 ) (#92400 )

2023-01-30 23:30:43 +00:00

_export

[Export] Introduce as_none in ex.Argument union type (#93210 )

2023-01-30 21:32:49 +00:00

_functorch

[Reland] Refactor stack_trace preservation for node meta preservation (#90803 ) (#92400 )

2023-01-30 23:30:43 +00:00

_inductor

Minor sympy usage fix in fbcode (#93171 )

2023-01-30 23:34:22 +00:00

_lazy

Add step closures (#84300 )

2022-09-06 20:55:34 +00:00

_prims

Properly compute device for elementwise operations with CPU scalar tensor (#93073 )

2023-01-26 21:27:57 +00:00

_prims_common

[NVFUSER] refactor nvfuser build (#89621 )

2023-01-26 02:50:44 +00:00

_refs

[inductor] Lower torch.exp2 and use it for torch.pow(2, x) (#92632 )

2023-01-20 22:06:23 +00:00

_subclasses

Add @count util to torch, use it to track benchmark stats (#93013 )

2023-01-26 03:09:12 +00:00

amp

Add AutocastHPU support (#84927 )

2022-10-12 19:37:16 +00:00

[quant][fx][pt2e][refactor] Refactor prepare.py for upcoming quantize_pt2e changes (#92641 )

2023-01-30 22:57:20 +00:00

autograd

[Py3.11] Remove skip logic from vmap and forward_ad (#91825 )

2023-01-25 22:40:56 +00:00

backends

Dynamo benchmark: add CPU specific changes (#88477 )

2023-01-07 09:26:06 +00:00

contrib

Change docstring type callable to Callable for consistency (#82487 )

2022-08-01 17:26:09 +00:00

cpu

Add correct __all__ for torch.distributed and torch.cuda submodules (#85702 )

2022-10-10 19:15:24 +00:00

csrc

[jit] Support Awaitable type (#90863 )

2023-01-30 17:38:59 +00:00

cuda

Revert "Improve bsr @ strided performance in baddmm for bfloat16/half with Triton kernels. (#88078 )"

2023-01-19 23:37:59 +00:00

distributed

[pytorch] fix backend_type for backend/PG plugin (#93129 )

2023-01-30 23:16:08 +00:00

distributions

[Dynamo] Fix calling UserDefinedObject.func should pass self object (#92050 )

2023-01-21 05:47:01 +00:00

fft

Fix typos used in documents under torch directory (#88300 )

2022-11-02 09:38:13 +00:00

func

[functorch] move batch_norm_replacement to torch.func (#91412 )

2023-01-12 19:15:41 +00:00

futures

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

[Reland] Refactor stack_trace preservation for node meta preservation (#90803 ) (#92400 )

2023-01-30 23:30:43 +00:00

jit

[jit] Support Awaitable type (#90863 )

2023-01-30 17:38:59 +00:00

legacy

…

lib

Apply Clang-Tidy readability-container-size-empty (#93236 )

2023-01-29 23:28:19 +00:00

linalg

Fix terminology within linalg.slogdet docs (#91129 )

2022-12-20 01:55:27 +00:00

masked

unify reduction types from different operators: scatter, scatter_reduce, segment_reduce (#91499 )

2023-01-13 04:32:34 +00:00

monitor

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

multiprocessing

Set min supported Python version to 3.8 (#93155 )

2023-01-29 18:28:46 +00:00

nested

Fix inaccuracy in nt constructor documentation + broken rendering (#89152 )

2022-11-16 22:32:46 +00:00

Use the old aten underscored function for Predictor (#93096 )

2023-01-28 03:14:18 +00:00

onnx

[ONNX] Export 'aten::index_put(self, mask, v)' when rank(mask) < rank(self) (#92862 )

2023-01-27 02:00:56 +00:00

optim

[optim][adam] use fastest impl whenever possible, add util (#93184 )

2023-01-30 19:58:55 +00:00

package

Set min supported Python version to 3.8 (#93155 )

2023-01-29 18:28:46 +00:00

profiler

Call profiler step via optimizer post hook (#90101 )

2023-01-13 18:07:40 +00:00

quantization

[ao] making _is_activation_post_process private with BC (#90554 )

2022-12-16 08:09:33 +00:00

signal

Reland "Add torch.utils.device_mode" (#91796 )

2023-01-09 20:57:12 +00:00

sparse

Revert "Improve bsr @ strided performance in baddmm for bfloat16/half with Triton kernels. (#88078 )"

2023-01-26 16:22:29 +00:00

special

Adding multigammaln ref and fix arange (#85153 )

2022-09-20 17:52:56 +00:00

testing

[Pytorch][Executorch] Handwritten view copy out ops should resize out (#91194 )

2023-01-30 23:07:14 +00:00

utils

Set min supported Python version to 3.8 (#93155 )

2023-01-29 18:28:46 +00:00

__config__.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

__future__.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

__init__.py

[jit] Support Awaitable type (#90863 )

2023-01-30 17:38:59 +00:00

_appdirs.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_classes.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_deploy.py

Deprecate TypedStorage, its derived classes, and all of their public methods (#85303 )

2022-11-08 18:11:01 +00:00

_guards.py

Properly resolve source_ref when constructing shape guards (#91058 )

2022-12-30 05:56:56 +00:00

_jit_internal.py

[jit] Support Awaitable type (#90863 )

2023-01-30 17:38:59 +00:00

_linalg_utils.py

Revert "Remove deprecated torch.symeig (#70988 )"

2023-01-24 19:03:40 +00:00

_lobpcg.py

Fix typo in _lobpcg.py (#91641 )

2023-01-04 15:19:05 +00:00

_lowrank.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_meta_registrations.py

[SDPA] Update SDPA API and make function Public (#92189 )

2023-01-23 20:50:46 +00:00

_namedtensor_internals.py

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

_ops.py

Fix checking of current mode in PyOperator dispatch (#92357 )

2023-01-18 23:08:36 +00:00

_python_dispatcher.py

[PolishComment] Polish code comment, revelant->relevant (#85238 )

2022-09-19 19:43:14 +00:00

_six.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_sources.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_storage_docs.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_tensor_docs.py

Revert "Remove deprecated torch.symeig (#70988 )"

2023-01-24 19:03:40 +00:00

_tensor_str.py

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

_tensor.py

Revert "Remove deprecated torch.symeig (#70988 )"

2023-01-24 19:03:40 +00:00

_torch_docs.py

Add missing scalar example in docs of torch.where (#93145 )

2023-01-28 03:46:44 +00:00

_utils_internal.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_utils.py

[follow-up] Python Attr Serialization (#88913 )

2023-01-13 17:38:51 +00:00

_VF.py

Apply ufmt to torch internal (#81643 )

2022-07-22 02:19:50 +00:00

_vmap_internals.py

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

_weights_only_unpickler.py

Sparse CSC/BSR/BSC serialization and pickle support (#89553 )

2022-11-23 20:56:48 +00:00

abi-check.cpp

…

CMakeLists.txt

Revert "[cuDNN][cuDNN V8 API] Always build assuming cuDNN >= 8.0 (#91527 )"

2023-01-16 13:28:09 +00:00

custom_class_detail.h

More tidy fixes (#93069 )

2023-01-27 06:40:50 +00:00

custom_class.h

Revert "Fix crash on unload torch cpu dll (#67632 )"

2022-08-02 00:56:18 +00:00

extension.h

…

functional.py

Update version numbers in torch.{stft,istft} deprecations (#91761 )

2023-01-05 22:17:37 +00:00

hub.py

Preventing crashing incase of no network by loading from cache (#91569 )

2023-01-11 11:56:46 +00:00

library.h

Make Python op registration work with torchdeploy/multipy (#87162 )

2022-11-03 12:56:44 +00:00

library.py

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

overrides.py

Revert "Remove deprecated torch.symeig (#70988 )"

2023-01-24 19:03:40 +00:00

py.typed

…

quasirandom.py

Integrate xdoctest - Rebased (#82797 )

2022-08-12 02:08:01 +00:00

random.py

…

README.txt

…

return_types.py

Add __all__ to torch.utils submodules (#85331 )

2022-09-27 14:45:26 +00:00

script.h

…

serialization.py

Enable xdoctest runner in CI for real this time (#83816 )

2022-12-29 05:32:42 +00:00

storage.py

Rename Tensor._storage to Tensor.untyped_storage and update docs (#91414 )

2022-12-28 19:21:34 +00:00

torch_version.py

Add __all__ to torch.distributed, futures, fx, nn, package, benchmark submodules (#80520 )

2022-07-08 14:31:24 +00:00

types.py

improve annotations (#86105 )

2022-10-05 10:33:26 +00:00

README.txt

Note [TH abstraction violation]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

TH/THC provide some hpp headers, which are proper C++ headers rather than
C headers.  These headers serve double duty as *internal implementation
detail* headers, whose contents should largely not be used by external
clients.

Ideally, we would not install these headers at all; instead, you should
use public functions (in headers like `THTensor.h`, NOT `THTensor.hpp`)
to manipulate these structs.  However, there are a few places
in torch/csrc where we violate this abstraction.  They are marked with
a pointer to this note.  Each of those sites will have to be refactored
when we refactor the guts of THTensor and related structures.