Fixes#156587
This sets lower bounds for fsspec and networkx in both setup.py and requirements,txt.
- fsspec>= 0.8.5 (released December 15, 2020)
- netowrkx>= 2.5.1 (released April 3, 2021)
These are the first stable versions released after Python 3.9 came out on October 5, 2020. Since Python 3.8 is no longer maintained, setting these minimums helps ensure PyTorch won't be installed alongside unexpectedly old versions of these packages.
Tested with these versions locally to make sure they don't break anything. Adding CI for lower-bound testing could be a follow up later if need.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/158565
Approved by: https://github.com/janeyx99
Change `build-system.build-backend`: `setuptools.build_meta:__legacy__` -> `setuptools.build_meta`. Also, move static package info from `setup.py` to `pyproject.toml`.
Now the repo can be installed from source via `pip` command instead of `python setup.py develop`:
```bash
python -m pip install --verbose --editable .
python -m pip install --verbose --no-build-isolation --editable .
```
In addition, the SDist is also buildable:
```bash
python -m build --sdist
python -m install dist/torch-*.tar.gz # build from source using SDist
```
Note that we should build the SDist with a fresh git clone if we will upload the output to PyPI. Because all files under `third_party` will be included in the SDist. The SDist file will be huge if the git submodules are initialized.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/155998
Approved by: https://github.com/ezyang, https://github.com/cyyever, https://github.com/atalman
ghstack dependencies: #157557
Change `build-system.build-backend`: `setuptools.build_meta:__legacy__` -> `setuptools.build_meta`. Also, move static package info from `setup.py` to `pyproject.toml`.
Now the repo can be installed from source via `pip` command instead of `python setup.py develop`:
```bash
python -m pip install --verbose --editable .
python -m pip install --verbose --no-build-isolation --editable .
```
In addition, the SDist is also buildable:
```bash
python -m build --sdist
python -m install dist/torch-*.tar.gz # build from source using SDist
```
Note that we should build the SDist with a fresh git clone if we will upload the output to PyPI. Because all files under `third_party` will be included in the SDist. The SDist file will be huge if the git submodules are initialized.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/155998
Approved by: https://github.com/ezyang, https://github.com/cyyever, https://github.com/atalman
I'm not sure if this is the right think to do, but cmake 4.0.0 got released on pypi and our builds are failing with it
Example:
aa70d62041 (39555975425-box)
I guess we have to go change all the cmake_minimum_required to >=3.5?
backwards compat still failing because its building with the base commit which this pr can't really change until it gets merged, but at least manywheel binary builds got past where they were originally failing
Also pin the conda installation, but the most recent version on conda is 3.31.2
Pull Request resolved: https://github.com/pytorch/pytorch/pull/150158
Approved by: https://github.com/cyyever, https://github.com/malfet
Fixes#148877
---
On 9 March 2025, [setuptools](https://pypi.org/project/setuptools/#history) published a new version and it is causing an issue on `pytorch` with the following error:
```
AttributeError: module 'distutils' has no attribute '_msvccompiler'. Did you mean: 'ccompiler'?
```
Last known working version is [75.8.2](https://pypi.org/project/setuptools/75.8.2/)
Currently it is affecting Windows ARM64 nightly build, however soon it might affect also Windows x64 builds. (conda version is not updated yet [setuptools conda](https://anaconda.org/anaconda/setuptools)
Locally both `Windows ARM64` and `Windows x64` are having same problem with the latest `setuptools` (>75.8.2)
---
This PR is pinning `setuptools` version.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/148879
Approved by: https://github.com/seemethere
Newer versions of distutils no longer import `_msvccompiler` upon init(on Windows platform, that was not the case on other platforms even before 74), but it's still accessible if one chooses to import it directly.
Test plan:
```
% python -c 'from setuptools import distutils; print(distutils.__version__, hasattr(distutils, "_msvccompiler")); from distutils import _msvccompiler; import setuptools; print(setuptools.__version__, _msvccompiler.__file__)'
3.10.9 False
65.5.0 /usr/local/fbcode/platform010/Python3.10.framework/Versions/3.10/lib/python3.10/site-packages/setuptools/_distutils/_msvccompiler.py
```
and
```
% python -c 'from setuptools import distutils; print(distutils.__version__, hasattr(distutils, "_msvccompiler")); from distutils import _msvccompiler; import setuptools; print(setuptools.__version__, _msvccompiler.__file__)'
3.13.0 False
75.6.0 /Users/malfet/py312-venv/lib/python3.13/site-packages/setuptools/_distutils/_msvccompiler.py
```
Gave up trying to appease the linker, so rewrote it as following function:
```python
def _get_vc_env(vc_arch: str) -> dict[str, str]:
try:
from setuptools import distutils # type: ignore[import]
return distutils._msvccompiler._get_vc_env(vc_arch) # type: ignore[no-any-return]
except AttributeError:
from setuptools._distutils import _msvccompiler #type: ignore[import]
return _msvccompiler._get_vc_env(vc_arch) # type: ignore[no-any-return]
```
This PR also undoes setuptools version restriction introduced by https://github.com/pytorch/pytorch/pull/136489 as premise for restriction is incorrect
Fixes https://github.com/pytorch/pytorch/issues/141319
Pull Request resolved: https://github.com/pytorch/pytorch/pull/141363
Approved by: https://github.com/huydhn, https://github.com/atalman
Summary:
The existing tests didn't cover a case where we had multiple autotunes in a single graph. Add a test to demonstrate that case.
Also added a test dependency on redis and removed the "fake redis" from the previous PR (#133579)
Test Plan: unit tests
Reviewed By: oulgen
Differential Revision: D61178861
Pull Request resolved: https://github.com/pytorch/pytorch/pull/133868
Approved by: https://github.com/oulgen
Summary: Previously we were mocking out FbRemoteFxGraphCacheBackend which meant that we were missing testing a whole bunch of the cache code. Cache at a lower level (CacheClient, LocalAutotuneCacheBackend, ManifoldClient, Redis) so we cover a larger amount of the caching code.
Test Plan: unit tests
Reviewed By: oulgen
Differential Revision: D60937966
Pull Request resolved: https://github.com/pytorch/pytorch/pull/133579
Approved by: https://github.com/oulgen
0.12.0 Major Updates:
- Add context manager to temporarily set the dictionary sorting mode
- Add accessor APIs
- Use `stable` tag for `pybind11` for Python 3.13 support
- Fix potential segmentation fault for pickling support
0.12.1 Updates:
- Fix warning regression during import when launch with strict warning filters
Closes#130155
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130139
Approved by: https://github.com/zou3519
ghstack dependencies: #130895
------
The opposite of #130836. Pin `sympy >= 1.13.0` for Python >= 3.9 and `sympy == 1.12.1` for Python 3.8.
- #130836
See the PR description of #130836 for more details.
`sympy` 1.13.0 introduces some breaking changes which break our tests. More specifically:
- Ref [Backwards compatibility breaks and deprecations](https://github.com/sympy/sympy/wiki/release-notes-for-1.13.0#backwards-compatibility-breaks-and-deprecations)
> BREAKING CHANGE: Float and Integer/Rational no longer compare equal with a == b. From now on Float(2.0) != Integer(2). Previously expressions involving Float would compare unequal e.g. x*2.0 != x*2 but an individual Float would compare equal to an Integer. In SymPy 1.7 a Float will always compare unequal to an Integer even if they have the same "value". Use sympy.numbers.int_valued(number) to test if a number is a concrete number with no decimal part. ([#25614](https://github.com/sympy/sympy/pull/25614) by [@smichr](https://github.com/smichr))
`sympy >= 1.13.0` is required to enable Python 3.13 support. This should be part of #130689.
- #130689
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130895
Approved by: https://github.com/ezyang
0.12.0 Major Updates:
- Add context manager to temporarily set the dictionary sorting mode
- Add accessor APIs
- Use `stable` tag for `pybind11` for Python 3.13 support
- Fix potential segmentation fault for pickling support
0.12.1 Updates:
- Fix warning regression during import when launch with strict warning filters
Closes#130155
Pull Request resolved: https://github.com/pytorch/pytorch/pull/130139
Approved by: https://github.com/zou3519
`ninja` is required to build C++ extensions in tests.
```pytb
ERROR: test_autograd_cpp_node (__main__.TestCompiledAutograd)
----------------------------------------------------------------------
Traceback (most recent call last):
File "/home/PanXuehai/Projects/pytorch/torch/testing/_internal/common_utils.py", line 2741, in wrapper
method(*args, **kwargs)
File "test/inductor/test_compiled_autograd.py", line 1061, in test_autograd_cpp_node
module = torch.utils.cpp_extension.load_inline(
File "/home/PanXuehai/Projects/pytorch/torch/utils/cpp_extension.py", line 1643, in load_inline
return _jit_compile(
File "/home/PanXuehai/Projects/pytorch/torch/utils/cpp_extension.py", line 1718, in _jit_compile
_write_ninja_file_and_build_library(
File "/home/PanXuehai/Projects/pytorch/torch/utils/cpp_extension.py", line 1800, in _write_ninja_file_and_build_library
verify_ninja_availability()
File "/home/PanXuehai/Projects/pytorch/torch/utils/cpp_extension.py", line 1849, in verify_ninja_availability
raise RuntimeError("Ninja is required to load C++ extensions")
RuntimeError: Ninja is required to load C++ extensions
To execute this test, run the following from the base repo dir:
python test/inductor/test_compiled_autograd.py -k TestCompiledAutograd.test_autograd_cpp_node
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/127380
Approved by: https://github.com/ezyang
