OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-07 12:21:27 +01:00
Code Issues Actions 37 Packages Projects Releases Wiki Activity
64,995 Commits 8,304 Branches 150 Tags 1.5 GiB
3058700f7f
Commit Graph

28 Commits

Author SHA1 Message Date
Guilherme Leobas
0a580da582 Add batch decomposition for torch.linalg.eigh (#110640) 
Closes https://github.com/pytorch/pytorch/issues/108481

Pull Request resolved: https://github.com/pytorch/pytorch/pull/110640
Approved by: https://github.com/kshitij12345, https://github.com/zou3519
 2023-10-09 21:36:49 +00:00
kshitij12345
b8a3998c23 add batch rule for missing inplace ops (#110692) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/110692
Approved by: https://github.com/ezyang
 2023-10-06 20:53:28 +00:00
kshitij12345
371d8ba599 vmap: decompose real and imag instead of registering batch rule (#110508) 
Clean-up

Pull Request resolved: https://github.com/pytorch/pytorch/pull/110508
Approved by: https://github.com/zou3519
 2023-10-06 06:01:12 +00:00
vfdev-5
d9fe1713c3 Enabled batch rule decompositions for upsample*.vec ops (#110333) 
Follow-up PR to https://github.com/pytorch/pytorch/pull/110172
Pull Request resolved: https://github.com/pytorch/pytorch/pull/110333
Approved by: https://github.com/zou3519
 2023-10-03 06:58:18 +00:00
vfdev-5
c62be12061 Added batch rules for _upsample_bi*2d_aa and _upsample_bi*2d_aa_backward (#110172) 
Description:
- Added batch rules for `_upsample_bi*2d_aa` and `_upsample_bi*2d_aa_backward`
- Added few more test cases into `sample_inputs_upsample_aten`

Pull Request resolved: https://github.com/pytorch/pytorch/pull/110172
Approved by: https://github.com/kshitij12345, https://github.com/zou3519
 2023-09-28 17:42:48 +00:00
SherlockNoMad
d997969b8b [Reland] Add sym_size/stride/numel/storage_offset to native_function.yaml (#103107) 
Differential Revision: D46459100

Pull Request resolved: https://github.com/pytorch/pytorch/pull/103107
Approved by: https://github.com/angelayi, https://github.com/soulitzer
 2023-06-12 19:18:49 +00:00
Nikita Shulga
20cf42de2c Revert "[Reland] Add sym_size/stride/numel/storage_offset to native_function.… (#100749)" 
This reverts commit bb454891ed.
 2023-05-16 18:17:02 -07:00
Sherlock Huang
bb454891ed [Reland] Add sym_size/stride/numel/storage_offset to native_function.… (#100749) 
…yaml (#91… (#91919)

Summary:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/91919 Approved by: https://github.com/ezyang

Fixes #ISSUE_NUMBER

Pull Request resolved: https://github.com/pytorch/pytorch/pull/92402

Reviewed By: ezyang

Differential Revision: D42565586

Pulled By: SherlockNoMad

fbshipit-source-id: 1c2986e45307e076d239836a1b45441a9fa3c9d9
ghstack-source-id: 969f4928486e04c57aaf98e20e3c3ca946c51613

Fixes #ISSUE_NUMBER

Pull Request resolved: https://github.com/pytorch/pytorch/pull/100749
Approved by: https://github.com/zhxchen17, https://github.com/albanD
 2023-05-12 22:57:42 +00:00
Li-Huai (Allan) Lin
c0674c439c [vmap] Add max_pool3d batch rule (#99522) 
Also add a helper to integrate `max_pool2d_with_indices` and `max_pool3d_with_indices`

Pull Request resolved: https://github.com/pytorch/pytorch/pull/99522
Approved by: https://github.com/zou3519
 2023-04-20 05:08:19 +00:00
Li-Huai (Allan) Lin
d31a00e713 [vamp] Add max_pool1d batch_rule (#99517) 
Fixes #97558

Pull Request resolved: https://github.com/pytorch/pytorch/pull/99517
Approved by: https://github.com/Skylion007, https://github.com/zou3519
 2023-04-20 05:08:17 +00:00
kshitij12345
5e014bfbbd [vmap] ldl_factor: batch rule (#97518) 
Ref https://github.com/pytorch/pytorch/issues/96855

Will look into `ldl_solve` separately.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/97518
Approved by: https://github.com/zou3519
 2023-03-25 04:37:32 +00:00
Danni Li
7711d24717 vmap support for linalg.lu_factor (#94328) 
Differential Revision: D43093457

Fix #91415

### Expected behaviour

No use warning.

```python
from functorch import vmap
x = torch.randn(4, 3, 2)
z = vmap(torch.linalg.lu_factor)(x)
```
Same behaviour as for-loop:

```python
x = torch.randn(4, 3, 2)
results = []
for xi in x:
  y = torch.linalg.lu_factor(xi)
  results.append(y)
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/94328
Approved by: https://github.com/zou3519, https://github.com/Skylion007, https://github.com/Chillee
 2023-03-23 14:18:57 +00:00
Kshiteej K
24c49dbf14 [functorch] batch rule : few decomposition ops (#96744) 
Fixes https://github.com/pytorch/pytorch/issues/96741

Pull Request resolved: https://github.com/pytorch/pytorch/pull/96744
Approved by: https://github.com/zou3519
 2023-03-15 18:55:05 +00:00
Richard Zou
13011afb87 Fix vmap registration for t, t_ (#96539) 
- t, t_ are not CompositeImplicitAutograd
- They were previously registered in BatchRulesDecompositions.cpp.
- The only thing that should get registered in BatchRulesDecompositions.cpp
are CompositeImplicitAutograd
- This PR moves their registrations out of there and into
BatchRulesViews.cpp.

Test Plan:
- existing tests
Pull Request resolved: https://github.com/pytorch/pytorch/pull/96539
Approved by: https://github.com/srossross, https://github.com/kshitij12345, https://github.com/Chillee
 2023-03-13 16:08:32 +00:00
Sean Ross-Ross
6650aac8ce move more operators to BatchRulesDecompositions (#93164) 
Moving operators over to `BatchRulesDecompositions.cpp` to remove xfails. I noticed that composite-compliant does not mean inductor or vmap compliant, so I added more `isTensorSubclassLike` checks

Pull Request resolved: https://github.com/pytorch/pytorch/pull/93164
Approved by: https://github.com/lezcano, https://github.com/kshitij12345
 2023-02-03 16:36:05 +00:00
PyTorch MergeBot
f7bd5d0ccb Revert "[Reland] Add sym_size/stride/numel/storage_offset to native_function.yaml (#91… (#92402)" 
This reverts commit 965f4ea3ba.

Reverted https://github.com/pytorch/pytorch/pull/92402 on behalf of https://github.com/zhxchen17 due to Caused a regression for an export model.
 2023-02-03 03:12:43 +00:00
Sherlock Huang
965f4ea3ba [Reland] Add sym_size/stride/numel/storage_offset to native_function.yaml (#91… (#92402) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/91919
Approved by: https://github.com/ezyang

Fixes #ISSUE_NUMBER

Pull Request resolved: https://github.com/pytorch/pytorch/pull/92402
Approved by: https://github.com/ezyang
 2023-02-01 04:47:49 +00:00
Khushi Agrawal
4c074ddfd2 [functorch][reland] vmap: bitwise operators (#92836) 
Previous PR: #91971

Fixes: https://github.com/pytorch/functorch/issues/1069

Pull Request resolved: https://github.com/pytorch/pytorch/pull/92836
Approved by: https://github.com/Chillee
 2023-01-26 06:12:47 +00:00
Sean Ross-Ross
d354499faf adding some more missing ops to vmap (#92110) 
removes some xfails that were a part of https://github.com/pytorch/functorch/issues/1009 and https://github.com/pytorch/functorch/issues/1087

Pull Request resolved: https://github.com/pytorch/pytorch/pull/92110
Approved by: https://github.com/zou3519
 2023-01-25 19:43:12 +00:00
PyTorch MergeBot
7ddcf4e0c3 Revert "[functorch] vmap: bitwise operators (#91971)" 
This reverts commit e54f7b3edd.

Reverted https://github.com/pytorch/pytorch/pull/91971 on behalf of https://github.com/malfet due to Broke functorch bitwise, see e54f7b3edd
 2023-01-23 14:52:16 +00:00
Khushi Agrawal
e54f7b3edd [functorch] vmap: bitwise operators (#91971) 
Fixes https://github.com/pytorch/functorch/issues/1069

Pull Request resolved: https://github.com/pytorch/pytorch/pull/91971
Approved by: https://github.com/kshitij12345, https://github.com/Chillee
 2023-01-23 09:03:13 +00:00
Henry Cheng
b6cfd62285 vmap support for torch.linalg.vander (#91749) 
Adds vmap support for torch.linalg.vander in a similar manner to how view_as_complex is implemented.

#91700

Pull Request resolved: https://github.com/pytorch/pytorch/pull/91749
Approved by: https://github.com/lezcano
 2023-01-19 14:49:54 +00:00
PyTorch MergeBot
befe815466 Revert "Add sym_size/stride/numel/storage_offset to native_function.yaml (#91919)" 
This reverts commit 0388400f3f.

Reverted https://github.com/pytorch/pytorch/pull/91919 on behalf of https://github.com/atalman due to Break internal build
 2023-01-17 21:03:18 +00:00
Sherlock Huang
0388400f3f Add sym_size/stride/numel/storage_offset to native_function.yaml (#91919) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/91919
Approved by: https://github.com/ezyang
 2023-01-17 03:39:57 +00:00
Sean Ross-Ross
0100293a7b feat: adding greater_equal Scalar variant (#91324) 
Fixes https://github.com/pytorch/functorch/issues/1080

```py
import torch
from functorch import vmap

def f(x):
    return torch.greater_equal(torch.cumsum(x, dim=0), .5 * 10)

x = torch.randn([10,10])
vmap(f)(x)
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/91324
Approved by: https://github.com/zou3519
 2023-01-05 20:25:38 +00:00
Joel Schlosser
1effabe257 Support per-parameter test decoration (#91658) 
Continuation of #79979.

Fixes #79161

This PR does the following:
* Expands the `parametrize_fn()` signature from returning a 3-tuple of `(test, test_name, param_kwargs)` to returning a 4-tuple of `(test, test_name, param_kwargs, decorator_fn)`. Expected signature for the addition is `decorator_fn(param_kwargs) -> List[decorator]` i.e. given the full set of test params, return a list of decorators to apply.
    * `modules`, `ops`, and `parametrize` now fit the new signature, returning `decorator_fn`s instead of applying decorators themselves.
    * `instantiate_parametrized_tests()` and `instantiate_device_type_tests()` now call the returned `decorator_fn`, passing in the full set of `param_kwargs` (after composition + `device` / `dtype` additions) and applying the returned decorators.
    * Composing multiple `parametrize_fn`s also composes the corresponding `decorator_fn`s; the composed `decorator_fn` simply concatenates the decorator lists returned by the constituents.
* Expands `DecorateInfo.is_active` to support callables:
```python
DecorateInfo(
    unittest.expectedFailure, "TestOps", "test_python_ref_executor",
    device_type='cuda', active_if=lambda params: params['executor'] == 'nvfuser'
),
```
* Adds several tests to `test/test_testing.py` ensuring proper decoration using `@parametrize`, `@modules`, and `@ops`.
* (minor) Fixes a couple `ModuleInfo` naming oddities uncovered during testing.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/91658
Approved by: https://github.com/malfet
 2023-01-04 21:08:32 +00:00
Sean Ross-Ross
cb3204823e adding test to audit CompositeImplicitAutograd ops that do not have a batching rule (#91367) 
Fixes https://github.com/pytorch/functorch/issues/1087

It looks like there are `306` rules that should be looked into
```
test/functorch/test_vmap_registrations.py .x.....xxxxxxx.x.x.x.x.x.x.x.x........xx.x.x..x.x.xxx...xxxx.x.x.x........x.........xxxxx..x..x.....xx...xx.....xxx.xxxxxxxxxxxxxxxxx.. [ 24%]
.........x.x......x.xxxxxx..x..xx.x.xxx.x.......x.xxx.xx..xxx.xxx...xxxxx.x....xxxxxxxxxxxxxxx....xx.xxx.xx.x...xx...xx...xxxxxx...xxxxx..x...xxxxxxxxxxxx..xx..xx.xx.x..xxxx..xx [ 56%]
.xx..x.x....xxxxxx.x.xx...xxxxx.xx...x..x.x.xx...xx.xxxxxx.xxxxxx..x........xxxxxxxx..xxxxxxxx..xx.xxxxxxxxxxxxxxxxxxxxxxx..........xxxx.xxxx.........xxxxxxxx..xxx..xxx.x.x.x.xx [ 88%]
xx.xxx.x......xxx.x.xxxxxxxx....x......xxxxxxxxx.xx.x.x.x.......xx                                                                                                                [100%]

=================================================================== 249 passed, 1185 deselected, 306 xfailed in 3.17s ===================================================================

```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/91367
Approved by: https://github.com/zou3519
 2023-01-03 04:21:39 +00:00
Sean Ross-Ross
dcce5677fd Adding test when registering a batching rule for a CompositeImplicitAutograd operation (#89465) 
This is a Follow on from https://github.com/pytorch/pytorch/pull/88771 which should close out https://github.com/pytorch/functorch/issues/1009 I've got another PR where I'm moving some operators over https://github.com/pytorch/pytorch/pull/89762

you can see that the new test file is being picked [run here](https://github.com/pytorch/pytorch/actions/runs/3617298059/jobs/6096218583#step:10:472)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/89465
Approved by: https://github.com/zou3519
 2022-12-12 16:21:07 +00:00