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 38 Packages Projects Releases Wiki Activity
66,257 Commits 8,304 Branches 150 Tags 1.5 GiB
c46fc46dba
Commit Graph

4205 Commits

Author SHA1 Message Date
drisspg
c46fc46dba expose mem-eff to autograd (#110495) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/110495
Approved by: https://github.com/jbschlosser
 2023-11-13 17:47:40 +00:00
Jason Ansel
b00311ce9e [dynamo] Add run_inductor_tests entrypoint (#113278) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113278
Approved by: https://github.com/yanboliang
 2023-11-11 08:54:43 +00:00
Jason Ansel
66d09f8217 [inductor] Move things into torch/testing/_internal/inductor_utils.py (#113275) 
This PR is just moving things around, so code shared by multiple tests files is in torch/testing/_internal/inductor_utils.py.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/113275
Approved by: https://github.com/yanboliang
ghstack dependencies: #113242
 2023-11-11 03:17:35 +00:00
Isuru Fernando
3b915f9de0 [pt2] enable meta tests for foreach ops (#113484) 
Try https://github.com/pytorch/pytorch/pull/113059 again.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/113484
Approved by: https://github.com/lezcano
 2023-11-11 02:43:41 +00:00
Peter Bell
c4fe817a69 [inductor] Fix test_dist on pre-sm80 and add skipCUDAIf decorator (#113384) 
`test_dist` uses bfloat16 which isn't well supported by triton on pre-sm80
hardware, so split the test in two and add a skip. This also adds a
`skipCUDAIf` decorator which only skips on CUDA devices so the test still runs
on CPU.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/113384
Approved by: https://github.com/lezcano
 2023-11-10 19:45:33 +00:00
Chien-Chin Huang
08641a3232 Make FakeProcessGroup traceable (#113314) 
This PR mimics what we have done to trace ProcessGroup. This allows use to use FakeProcessGroup with torch.compile. FakeProcessGroup allows us to use world_size > 1 without creating multiple processes thus enabling the usage of PDB to debug bucketing DDP allreduce in the Inductor. We can theoretically use GLOO with world_size==1 to achieve the same goal. However, the `wait()` seems to be optimized away when the world_size is 1.

Differential Revision: [D51136463](https://our.internmc.facebook.com/intern/diff/D51136463/)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/113314
Approved by: https://github.com/wanchaol
 2023-11-10 16:03:38 +00:00
leslie-fang-intel
65e99357ae [Inductor] [Quant] Enable QConv2d Unary int8-mixed-bf16 Lowering (#112550) 
**Summary**
- PR 5 for enabling Int8-Mixed-BF16 PT2E PTQ Quantization with Inductor https://github.com/pytorch/pytorch/issues/111640.
- Enable the QConv2d Unary int8-mixed-bf16 weight prepack and post grad lowering inside inductor.

**TestPlan**
```
python -m pytest test_mkldnn_pattern_matcher.py -k test_qconv2d
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112550
Approved by: https://github.com/jgong5, https://github.com/jerryzh168
 2023-11-10 08:59:40 +00:00
PyTorch MergeBot
68bf0f1e7d Revert "[inductor] Move things into torch/testing/_internal/inductor_utils.py (#113275)" 
This reverts commit c967dc526a.

Reverted https://github.com/pytorch/pytorch/pull/113275 on behalf of https://github.com/PaliC due to the diff this is stacked on top of appears to be causing inductor failures internally ([comment](https://github.com/pytorch/pytorch/pull/113275#issuecomment-1805131017))
 2023-11-10 05:40:55 +00:00
Jason Ansel
c967dc526a [inductor] Move things into torch/testing/_internal/inductor_utils.py (#113275) 
This PR is just moving things around, so code shared by multiple tests files is in torch/testing/_internal/inductor_utils.py.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/113275
Approved by: https://github.com/yanboliang
 2023-11-10 00:11:09 +00:00
Jerry Zhang
501d118255 [quant][pt2e] Add transform_for_annotation method in Quantizer (#113115) 
Summary:
Adding the method so that people can do some transformations before annotation to make the graph easier to annotate

Test Plan:
python test/test_quantization.py TestQuantizePT2E.test_transform_for_annotation

Reviewers:

Subscribers:

Tasks:

Tags:

Differential Revision: [D51141080](https://our.internmc.facebook.com/intern/diff/D51141080)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113115
Approved by: https://github.com/kimishpatel
 2023-11-09 20:23:29 +00:00
Jerry Zhang
12c257cc00 [qunat][pt2e] Support allow_implicit_sharing flag (#112929) 
Summary:
For a Node: node1 and edge: (node1, node2), since they are observing the same
Tensor, we may want to implicitly share observers, this flag allows people to
turn off this behavior for the output of the node

See the test_allow_implicit_sharing test for use case

Test Plan:
python test/test_quantization.py TestQuantizePT2E.test_allow_implicit_sharing

Reviewers:

Subscribers:

Tasks:

Tags:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112929
Approved by: https://github.com/kimishpatel
 2023-11-08 23:47:17 +00:00
lcskrishna
78b8465565 [Distributed] Limit world_size to 8 for FSDP Unit tests (#103412) 
There are few unit tests in FSDP that can support upto 8 GPUs.
In this case, for example test_fsdp_uneven has an input size of [8,3]. For each process/rank we pass the data as input[self.rank] as below. So when we use 16 GPUs for our tests, these tests throw an index/key error. So basically to avoid such corner cases, I would like to add this change to use 8GPUs if there are more than 8 GPUs. This is applicable to both ROCm and CUDA builds as well.

https://github.com/pytorch/pytorch/blob/main/test/distributed/fsdp/test_fsdp_uneven.py#L44
https://github.com/pytorch/pytorch/blob/main/test/distributed/fsdp/test_fsdp_uneven.py#L55

Pull Request resolved: https://github.com/pytorch/pytorch/pull/103412
Approved by: https://github.com/jithunnair-amd, https://github.com/pruthvistony, https://github.com/malfet
 2023-11-08 17:21:38 +00:00
wz337
f2963642c2 [DDP] Add device_mesh to DDP ctor (#112761) 
As title.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112761
Approved by: https://github.com/fegin
 2023-11-08 03:08:08 +00:00
Iris Zhang
2fc940e0c4 [DTensorTestbase] Add run_subtests to DTensorTestbase and fix test_ddp checkpoint test error (#113051) 
This PR:

1. Adds  `run_subtests` to `DTensorTestbase`, which runs test functions given by `test_fn` as subtests. This amortizes the costly dist setup.
2. Update `test/distributed/checkpoint/test_state_dict.py` to use `DTensorTestbase`. This fixes the "Duplicate GPU detected: rank 0 and rank 1 both on CUDA device 11000" when running on 1 GPU, as the skip_if_lt_x_gpu is currently happening after dist setup.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113051
Approved by: https://github.com/fegin
 2023-11-07 18:48:01 +00:00
Li-Huai (Allan) Lin
c4bb77323d [MPS] Add searchsorted op (#112829) 
The metal kernels implemented are closely following `Bucketization.cu`.

Benchmark:
```
[----------------------------- searchsorted ----------------------------]
                                                         |  cpu   |  mps
1 threads: --------------------------------------------------------------
      Batch size: 8; In features: 64; Sorter: True       |    44  |   530
      Batch size: 8; In features: 64; Sorter: False      |    31  |    12
      Batch size: 8; In features: 256; Sorter: True      |   131  |   520
      Batch size: 8; In features: 256; Sorter: False     |   107  |    12
      Batch size: 8; In features: 1024; Sorter: True     |   499  |   590
      Batch size: 8; In features: 1024; Sorter: False    |   398  |    12
      Batch size: 16; In features: 64; Sorter: True      |    71  |   540
      Batch size: 16; In features: 64; Sorter: False     |    57  |    12
      Batch size: 16; In features: 256; Sorter: True     |   242  |   610
      Batch size: 16; In features: 256; Sorter: False    |   200  |    12
      Batch size: 16; In features: 1024; Sorter: True    |   999  |   720
      Batch size: 16; In features: 1024; Sorter: False   |   842  |    12
      Batch size: 32; In features: 64; Sorter: True      |   124  |   509
      Batch size: 32; In features: 64; Sorter: False     |   103  |    12
      Batch size: 32; In features: 256; Sorter: True     |   477  |   650
      Batch size: 32; In features: 256; Sorter: False    |   407  |    12
      Batch size: 32; In features: 1024; Sorter: True    |  1940  |   833
      Batch size: 32; In features: 1024; Sorter: False   |  1710  |    12
      Batch size: 64; In features: 64; Sorter: True      |   231  |   590
      Batch size: 64; In features: 64; Sorter: False     |   194  |    12
      Batch size: 64; In features: 256; Sorter: True     |   937  |   710
      Batch size: 64; In features: 256; Sorter: False    |   800  |    13
      Batch size: 64; In features: 1024; Sorter: True    |  3980  |  1290
      Batch size: 64; In features: 1024; Sorter: False   |  3330  |    12
      Batch size: 128; In features: 64; Sorter: True     |   448  |   650
      Batch size: 128; In features: 64; Sorter: False    |   390  |    13
      Batch size: 128; In features: 256; Sorter: True    |  1830  |   850
      Batch size: 128; In features: 256; Sorter: False   |  1590  |    12
      Batch size: 128; In features: 1024; Sorter: True   |  7790  |  2850
      Batch size: 128; In features: 1024; Sorter: False  |  6670  |    13
```

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112829
Approved by: https://github.com/malfet
 2023-11-07 17:22:08 +00:00
PyTorch MergeBot
9a28a7b498 Revert "Add support for torch.Generator type in TorchScript (#110413)" 
This reverts commit 27e31ab6e8.

Reverted https://github.com/pytorch/pytorch/pull/110413 on behalf of https://github.com/PaliC due to breaking internal builds ([comment](https://github.com/pytorch/pytorch/pull/110413#issuecomment-1799003164))
 2023-11-07 15:53:32 +00:00
Fei Kou
220c3bae6d Pass in parallel strategy to tp_transform API (#112286) 
Summary: Support passing in parallel strategy map and apply TP transform based on that. This is to make it easier manually select layers from real model to parallelize and benchmark.

Test Plan: buck test mode/opt  -c fbcode.enable_gpu_sections=true //caffe2/test/distributed/_tensor/experimental:tp_transform

Differential Revision: D50591039

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112286
Approved by: https://github.com/wanchaol
 2023-11-07 09:28:58 +00:00
Oguz Ulgen
f6008be266 Move all triton related testing utils into shared file (#113008) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/113008
Approved by: https://github.com/zou3519, https://github.com/jansel
ghstack dependencies: #112752
 2023-11-07 05:29:29 +00:00
Peter Bell
65ecb36621 Move ShapeEnv config out of dynamo (#112933) 
Previously there was a circular dependency between fx and dynamo that happened
to work out since ShapeEnv didn't access the config at module init time.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112933
Approved by: https://github.com/ezyang
 2023-11-07 01:10:25 +00:00
Aaron Gokaslan
8219bf051b [BE]: Apply RUF015 to torch folder (#113025) 
Removes unnecessary allocations of iterators. There is a small chance this may have side effects as the entire iterator is no longer consumed, but this is a way more efficient method for retrieving the first element.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/113025
Approved by: https://github.com/ezyang, https://github.com/malfet
 2023-11-07 00:48:15 +00:00
Antonio Kim
27e31ab6e8 Add support for torch.Generator type in TorchScript (#110413) 
- Add support for `torch.Generator` type in TorchScript
- Add `generator` args to all `torch.nn.init` functions that call `uniform_` or `normal_`
- Add support for `torch.Generator` in LTC's TorchScript backend (CC: @wconstab)

CC: @eellison @davidberard98 @GlebKazantaev @behzad-a
Pull Request resolved: https://github.com/pytorch/pytorch/pull/110413
Approved by: https://github.com/wconstab, https://github.com/albanD, https://github.com/glebk-cerebras, https://github.com/davidberard98
 2023-11-06 21:27:02 +00:00
Richard Zou
185515368b Add generated opcheck test for if the pt2_compliant_tag is incorrectly applied (#112759) 
Summary:
If there are xfails in the failures_dict and the operator has the
pt2_compliant_tag, then we raise an error. These generated tests are separate
from those in the failures dict because we don't actually need any sample
inputs to check this.

Test Plan: - New tests

Differential Revision: D50936201

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112759
Approved by: https://github.com/ezyang
 2023-11-06 13:45:35 +00:00
CaoE
455241bbd3 Add Half for aten2, logaddexp, logaddexp2, hypot, and nextafter on CPU (#112138) 
Add Half for aten2, logaddexp, logaddexp2, hypot, and nextafter on CPU.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112138
Approved by: https://github.com/cpuhrsch
 2023-11-06 06:01:29 +00:00
CaoE
26b5e27ace Add Half support for cummax, cummin, cumprod, logcumsumexp, and prod on CPU (#112132) 
Add Half support for cummax, cummin, cumprod, logcumsumexp, and prod on CPU.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112132
Approved by: https://github.com/cpuhrsch
 2023-11-05 12:31:38 +00:00
David Berard
d084a024ae [easy] skipIfTorchInductor - use condition variable (#112774) 
Fixes #112465
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112774
Approved by: https://github.com/jon-chuang, https://github.com/aaronenyeshi
 2023-11-03 17:55:32 +00:00
drisspg
458e7d09fd Add meta func for scaled mm (#112609) 
# Summary
Adds a meta implementation for _scaled_mm which is required for dynamic shapes

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112609
Approved by: https://github.com/eellison, https://github.com/malfet
 2023-11-03 03:44:22 +00:00
rzou
32039883d1 Set default for IS_FBCODE flag (#112766) 
Summary:
If IS_FBCODE is False, then we print an OSS repro if a test fails. We do
set IS_FBCODE manually on most internal tests, but we don't do it for
all of them. This PR changes it so that the IS_FBCODE gets set to the
correct default value (and then tests are able to override them if
they'd like).

Test Plan:
- Tested locally
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112766
Approved by: https://github.com/williamwen42
 2023-11-03 00:01:07 +00:00
PyTorch MergeBot
2e29172942 Revert "Add meta func for scaled mm (#112609)" 
This reverts commit 75174c3797.

Reverted https://github.com/pytorch/pytorch/pull/112609 on behalf of https://github.com/huydhn due to Sorry for reverting this change, but it is failing ROCm jobs 75174c3797 ([comment](https://github.com/pytorch/pytorch/pull/112609#issuecomment-1791704037))
 2023-11-02 23:37:16 +00:00
drisspg
75174c3797 Add meta func for scaled mm (#112609) 
# Summary
Adds a meta implementation for _scaled_mm which is required for dynamic shapes

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112609
Approved by: https://github.com/eellison, https://github.com/malfet
 2023-11-02 18:42:41 +00:00
Ting Lu
e06288f8f1 skip test in test_eager_transforms.py while Triton lacks ARM support (#112092) 
fix the failure with test_compile_vmap_hessian in test_eager_transforms.py. Skipping the test while we wait for ARM support from Triton. cc @ptrblck @eqy

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112092
Approved by: https://github.com/eqy, https://github.com/huydhn
 2023-11-01 21:33:18 +00:00
Kurt Mohler
fd209543d5 Add torch.utils.deterministic.fill_uninitialized_memory flag (#111377) 
Part of #109802

Pull Request resolved: https://github.com/pytorch/pytorch/pull/111377
Approved by: https://github.com/albanD, https://github.com/aaronenyeshi
 2023-11-01 16:10:09 +00:00
Peter Bell
04024926f4 Use pytree.tree_map_ everywhere (#112417) 
Wherever we discard the output of `tree_map` it's better to call `tree_map_`
which doesn't unflatten the mapped results and so is a lot cheaper.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112417
Approved by: https://github.com/lezcano
ghstack dependencies: #112391, #112392, #112393, #112394
 2023-10-31 15:57:06 +00:00
Peter Bell
66c32d099a Use pytree.arg_tree_leaves everywhere (#112394) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112394
Approved by: https://github.com/lezcano
ghstack dependencies: #112391, #112392, #112393
 2023-10-31 15:57:06 +00:00
CaoE
a310cc8968 Add Half support for kthvalue, cross, hist, and logit on CPU (#112135) 
Add Half support for kthvalue, cross, hist, and logit on CPU.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112135
Approved by: https://github.com/cpuhrsch
 2023-10-31 09:12:47 +00:00
Guilherme Leobas
b8a10a8a2d Add batch decomposition for torch.unsafe_chunk (#110862) 
This updates the docs as well to show `torch.unsafe_chunk`. Should the `unsafe_*` functions should not appear in the docs?

Pull Request resolved: https://github.com/pytorch/pytorch/pull/110862
Approved by: https://github.com/kshitij12345, https://github.com/zou3519
 2023-10-31 00:37:08 +00:00
Tristan Rice
013f622dd2 grid_sample: support bfloat16 (#112331) 
This adds bfloat16 support to `torch.nn.functional.grid_sample` this is particularly important when doing feature sampling such as for rendering techniques used in PyTorch3d or for camera projections to voxel grids such as in SimpleBEV.

Related to #57707

Test plan:

```
pytest test/test_nn.py -k grid_sample
pytest test/test_ops.py -k grid_sample
```
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112331
Approved by: https://github.com/zou3519
 2023-10-30 19:31:41 +00:00
CaoE
29844adbe0 Add Half support for logspace and range on CPU (#112131) 
Add Half support for logspace and range on CPU

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112131
Approved by: https://github.com/cpuhrsch
 2023-10-30 07:18:47 +00:00
Peter Bell
bbd5b935e4 Use pytree.tree_leaves everywhere (#112324) 
This changes all the instances I could find of `tree_flatten(...)[0]` or
`x, _ = tree_flatten` to use `tree_leaves`.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112324
Approved by: https://github.com/lezcano
ghstack dependencies: #112327, #112323
 2023-10-30 03:39:04 +00:00
PyTorch MergeBot
ace2713d1e Revert "Add torch.utils.deterministic.fill_uninitialized_memory flag (#111377)" 
This reverts commit f1785373c0.

Reverted https://github.com/pytorch/pytorch/pull/111377 on behalf of https://github.com/facebook-github-bot due to Diff reverted internally ([comment](https://github.com/pytorch/pytorch/pull/111377#issuecomment-1784179040))
 2023-10-29 17:41:55 +00:00
Xiaodong Wang
3080fd8383 [profiler] add send/recv src/dst info (#111811) 
Summary: There is an ask to add src/dst to nccl trace. This feels like the easiest way to do - adding it to metadata seems to require plumbing a few stacks so will be more work

Test Plan: {F1128545195}

Reviewed By: davidberard98

Differential Revision: D50560692

Pull Request resolved: https://github.com/pytorch/pytorch/pull/111811
Approved by: https://github.com/davidberard98, https://github.com/aaronenyeshi, https://github.com/fduwjj
 2023-10-28 02:48:23 +00:00
Nikita Shulga
328a4c5475 [BE] Enhance OpInfo.supported_dtype (#111995) 
Current implementation is prone to errors, as it accepts any object, but does not print an error or something if device_type is not recognized.

Remediate it by accepting both device-type and device identifies (either `torch.device` instance or "{device_type}:{ordinal}" string

Fixes https://github.com/pytorch/pytorch/issues/111179

Pull Request resolved: https://github.com/pytorch/pytorch/pull/111995
Approved by: https://github.com/albanD
 2023-10-27 19:42:01 +00:00
Isuru Fernando
c7dcba9276 Remove passing disable_fastpath in kwargs (#112250) 
Fixes an issue that came up in https://github.com/pytorch/pytorch/pull/112030

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112250
Approved by: https://github.com/lezcano
 2023-10-27 18:29:20 +00:00
Lengyue
f839a5627b Add bf16 support to replicate padding (#112099) 
Fixes #99433

Pull Request resolved: https://github.com/pytorch/pytorch/pull/112099
Approved by: https://github.com/mikaylagawarecki
 2023-10-26 20:30:49 +00:00
Joel Schlosser
42e4c648a2 New @decorateIf decorator for param-specific conditional decoration (#112033) 
Adds a new decorator `@decorateIf(decorator, predicate_fn)`. Examples:
```python
from torch.testing._internal.common_utils import decorateIf
...

@decorateIf(unittest.skip, lambda params: params["x"] == 2)
@parametrize("x", range(5))
def test_foo(self, x):
    ...

@parametrize("x,y", [(1, 'foo'), (2, 'bar'), (3, 'baz')])
@decorateIf(
    unittest.expectedFailure,
    lambda params: params["x"] == 3 and params["y"] == "baz"
)
def test_bar(self, x, y):
    ...

@decorateIf(
    unittest.expectedFailure,
    lambda params: params["op"].name == "add" and params["dtype"] == torch.float16
)
@ops(op_db)
def test_op_foo(self, device, dtype, op):
    ...

@decorateIf(
    unittest.skip,
    lambda params: params["module_info"].module_cls is torch.nn.Linear and \
        params["device"] == "cpu"
)
@modules(module_db)
def test_module_foo(self, device, dtype, module_info):
    ...
```

Follow-up for per-param decoration based on https://github.com/pytorch/pytorch/issues/79161#issuecomment-1152487359
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112033
Approved by: https://github.com/clee2000, https://github.com/pmeier
 2023-10-26 14:39:59 +00:00
Cao E
1c89ea7f72 Add Half support for softmax and log_softmax on CPU (#103315) 
Add Half support for softmax and log_softmax on CPU.
Note: This introduces a correctness issue with MPS https://github.com/pytorch/pytorch/issues/111416 and https://github.com/pytorch/pytorch/issues/111479.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/103315
Approved by: https://github.com/jgong5, https://github.com/mikaylagawarecki, https://github.com/malfet
 2023-10-26 08:38:54 +00:00
Kurt Mohler
f1785373c0 Add torch.utils.deterministic.fill_uninitialized_memory flag (#111377) 
Part of #109802

Pull Request resolved: https://github.com/pytorch/pytorch/pull/111377
Approved by: https://github.com/albanD
 2023-10-26 02:39:06 +00:00
Evgeni Burovski
7352c88f58 TST: add x{pass,fail}IfTorchDynamo (#112082) 
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112082
Approved by: https://github.com/lezcano
ghstack dependencies: #112081
 2023-10-26 01:03:45 +00:00
rzou
6365992f92 [opcheck] Add way to initialize blank failures dict (#111948) 
Summary:

Fixes #111926. The workflow is:
- create a blank file with the correct name
- run a test with PYTORCH_OPCHECK_ACCEPT=1

Test Plan:
- tested locally

Reviewers:

Subscribers:

Tasks:

Tags:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/111948
Approved by: https://github.com/ezyang
ghstack dependencies: #111912, #111915
 2023-10-25 21:20:59 +00:00
PyTorch MergeBot
7e654c8f88 Revert "WIP / TST: allow testing torch._numpy under Dynamo (#110401)" 
This reverts commit 5ed4a423de.

Reverted https://github.com/pytorch/pytorch/pull/110401 on behalf of https://github.com/huydhn due to Sorry for reverting your change, but it is failing dynamo job in trunk 5ed4a423de ([comment](https://github.com/pytorch/pytorch/pull/110401#issuecomment-1779811943))
 2023-10-25 18:21:16 +00:00
Evgeni Burovski
5ed4a423de WIP / TST: allow testing torch._numpy under Dynamo (#110401) 
Use conditional imports: when running under dynamo, import the original NumPy not torch._numpy. This is what we want to trace, not our implementation.

With this, the test suite passes with and without `PYTORCH_TEST_WITH_DYNAMO=1` (modulo a couple of test modules which are not meant to be compiled, e.g. `test_nep50_examples`). There are two new decorators, `x{fail,pass}ifTorchDynamo`, the `xpass` in most cases indicates a graph break and a fallback to eager for things we do not implement.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/110401
Approved by: https://github.com/lezcano
 2023-10-25 16:02:16 +00:00