fe80f03726
Add B200 files to labeler and update codeowners ( #165767 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/165767
Approved by: https://github.com/slayton58
2025-10-17 23:24:17 +00:00
2984bfe3da
[ez][CI] Run vllm workflow on vllm pin updates ( #163353 )
...
As in title
The auto pin update was merged without running vllm workflow
Pull Request resolved: https://github.com/pytorch/pytorch/pull/163353
Approved by: https://github.com/malfet , https://github.com/wdvr
2025-09-19 17:32:49 +00:00
f46e4bcf43
Revert "Add ciflow/vllm to vLLM commit hash update PR(s) ( #161678 )"
...
This reverts commit 0e358050304c6a350dae2bce497bd1867ecc3c9f.
Reverted https://github.com/pytorch/pytorch/pull/161678 on behalf of https://github.com/yangw-dev due to we want to keep the vllm pinn updated now, right now we have some failure ([comment](https://github.com/pytorch/pytorch/pull/161678#issuecomment-3234876332 ))
2025-08-28 20:42:19 +00:00
0e35805030
Add ciflow/vllm to vLLM commit hash update PR(s) ( #161678 )
...
As it should be, otherwise, PR(s) like https://github.com/pytorch/pytorch/pull/161121 were merged without the signals it needed.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/161678
Approved by: https://github.com/atalman
2025-08-28 01:35:04 +00:00
7982b8c703
[BE][AOTI] Remove duplicate schema for ExternKernelNode ( #155867 )
...
Summary: The definition of `ExternKernelNode` and `ExternKernelNodes` schema in `torch/_export/serde/aoti_schema.py` is a complete duplicate of the ones in `torch/_export/serde/schema.py`.
Test Plan:
CI
Rollback Plan:
Differential Revision: D76558294
Pull Request resolved: https://github.com/pytorch/pytorch/pull/155867
Approved by: https://github.com/jingsh
2025-06-14 02:03:27 +00:00
7d282da449
Add automatic categorization for release notes: inductor (aoti) ( #151569 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/151569
Approved by: https://github.com/desertfire
ghstack dependencies: #151453
2025-04-18 03:39:06 +00:00
a41413829c
Use release notes label for module: distributed_checkpoint ( #148352 )
...
module: distributed_checkpoint is redundant with oncall: distributed checkpointing.
@fduwjj let us know that module: distributed_checkpoint is just used for release notes, so let's use the release notes label for the release notes, which the bot will pick up better.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/148352
Approved by: https://github.com/fegin
2025-03-03 21:33:28 +00:00
40ccb7a86d
cpp_wrapper: Move #includes to per-device header files ( #145932 )
...
Summary:
This prepares us for the next PR in the stack, where we introduce pre-compiled per-device header files to save compilation time.
Reland https://github.com/pytorch/pytorch/pull/143909 after merge conflicts.
Co-authored-by: Benjamin Glass <[bglass@quansight.com ](mailto:bglass@quansight.com )>
Differential Revision: D68656960
Pulled By: benjaminglass1
Pull Request resolved: https://github.com/pytorch/pytorch/pull/145932
Approved by: https://github.com/yushangdi , https://github.com/benjaminglass1
Co-authored-by: bglass@quansight.com <bglass@quansight.com >
2025-01-29 21:08:45 +00:00
96f4abba17
[dtensor] move all tests to distribute/tensor folder ( #144166 )
...
as titled, mainly moving files
Pull Request resolved: https://github.com/pytorch/pytorch/pull/144166
Approved by: https://github.com/Skylion007
2025-01-08 00:32:33 +00:00
6c54963f75
Revert "[dtensor] move all tests to distribute/tensor folder ( #144166 )"
...
This reverts commit 2e1ea8598f477322965c28fb52e6e5f53876d8dd.
Reverted https://github.com/pytorch/pytorch/pull/144166 on behalf of https://github.com/huydhn due to Sorry for reverting your change, but inductor/test_compiled_autograd needs to be updated ([comment](https://github.com/pytorch/pytorch/pull/144166#issuecomment-2575969871 ))
2025-01-07 18:31:36 +00:00
2e1ea8598f
[dtensor] move all tests to distribute/tensor folder ( #144166 )
...
as titled, mainly moving files
Pull Request resolved: https://github.com/pytorch/pytorch/pull/144166
Approved by: https://github.com/Skylion007
2025-01-07 06:45:14 +00:00
47f44303ff
Add ciflow/inductor automatically in more cases ( #140824 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/140824
Approved by: https://github.com/malfet
2024-11-15 19:54:20 +00:00
64ccebd2e0
update labeler for module: compiled autograd ( #137954 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/137954
Approved by: https://github.com/yf225
2024-10-16 01:56:21 +00:00
361db32d47
Consolidate SymDispatchMode into ProxyTensorMode ( #132674 )
...
Instead of having a separate context variable for SymDispatchMode, we
now simply delegate to the current active proxy tensor mode when we
need to trace a SymInt. We maintain a separate `__sym_dispatch__` magic
method as the calling convention is different than `__torch_dispatch__`.
Consolidating the modes in this ways means that we can consistently
disable both of these modes in tandem simply by removing the mode
from the proxy mode infra slot.
Signed-off-by: Edward Z. Yang <ezyang@meta.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/132674
Approved by: https://github.com/zou3519 , https://github.com/bdhirsh
2024-08-08 12:02:54 +00:00
a9ff190867
Revert "Consolidate SymDispatchMode into ProxyTensorMode ( #132674 )"
...
This reverts commit ffdf48e63b94930c81f05b06444721109d0b243d.
Reverted https://github.com/pytorch/pytorch/pull/132674 on behalf of https://github.com/PaliC due to We need to now revert https://github.com/pytorch/pytorch/pull/132216 in OSS and there is a dependency on this pr ([comment](https://github.com/pytorch/pytorch/pull/132674#issuecomment-2274062785 ))
2024-08-07 18:25:33 +00:00
ffdf48e63b
Consolidate SymDispatchMode into ProxyTensorMode ( #132674 )
...
Instead of having a separate context variable for SymDispatchMode, we
now simply delegate to the current active proxy tensor mode when we
need to trace a SymInt. We maintain a separate `__sym_dispatch__` magic
method as the calling convention is different than `__torch_dispatch__`.
Consolidating the modes in this ways means that we can consistently
disable both of these modes in tandem simply by removing the mode
from the proxy mode infra slot.
Signed-off-by: Edward Z. Yang <ezyang@meta.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/132674
Approved by: https://github.com/zou3519 , https://github.com/bdhirsh
2024-08-06 17:03:17 +00:00
6ebec38453
Add ciflow/linux-aarch64 to auto labeler on mkldnn PR's ( #125599 )
...
Trigger Aarch64 CI on oneDNN changes, to detect issues like this: https://github.com/pytorch/pytorch/issues/125548
Pull Request resolved: https://github.com/pytorch/pytorch/pull/125599
Approved by: https://github.com/malfet , https://github.com/snadampal
2024-05-06 17:26:17 +00:00
4946638f06
[AOTI] Add ABI-compatiblity tests ( #123848 )
...
Summary: In AOTInductor generated CPU model code, there can be direct references to some aten/c10 utility functions and data structures, e.g. at::vec and c10::Half. These are performance critical and thus it doesn't make sense to create C shim for them. Instead, we make sure they are implemented in a header-only way, and use this set of tests to guard future changes.
There are more header files to be updated, but we will do it in other followup PRs.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/123848
Approved by: https://github.com/jansel
ghstack dependencies: #123847
2024-04-19 00:51:24 +00:00
8bad7b63c8
[ez] Add more files to trigger inductor ( #122669 )
...
To catch https://github.com/pytorch/pytorch/pull/122562/files
Pull Request resolved: https://github.com/pytorch/pytorch/pull/122669
Approved by: https://github.com/desertfire
2024-03-26 15:19:30 +00:00
664dd61b29
Add some more symbolic shapes related files to ciflow/inductor ( #120887 )
...
Signed-off-by: Edward Z. Yang <ezyang@meta.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/120887
Approved by: https://github.com/janeyx99 , https://github.com/malfet
2024-02-29 16:59:32 +00:00
1e272fb6d6
[export] Undo "module: export" labeling ( #116042 )
...
Delete the auto-labeling of "module: export" as this is not really used, and we want to delete the "module: export" label.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/116042
Approved by: https://github.com/clee2000
2023-12-18 21:23:17 +00:00
72eab5aa43
Configures distributed_checkpoint label ( #115833 )
...
Configures the existing `module: distributed_checkpoint` label
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115833
Approved by: https://github.com/wconstab , https://github.com/wz337
2023-12-15 18:17:25 +00:00
9f7b3a4e18
Move autolabeler to "oncall: distributed" not "module:.." ( #115447 )
...
Reasoning for the change is spelled out in this issue
https://github.com/pytorch/pytorch/issues/115168
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115447
Approved by: https://github.com/huydhn , https://github.com/malfet
2023-12-08 22:53:20 +00:00
71bf4f3b87
[CI] Add torch/_functorch/_aot_autograd to auto-label rule ( #115283 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115283
Approved by: https://github.com/bdhirsh
2023-12-06 20:07:53 +00:00
3fccc0446c
Add dtensor and fsdp/2d tests to inductor_distributed CI ( #114642 )
...
Smuggle important and not too slow tests to run on this trunk job,
instead of just on the periodic job where they currently reside.
- test_dtensor_compile took 70sec, test_fsdp_2d_parallel took 198sec
locally
As a follow up, organize the distributed-mgpu tests better and maybe
rename this job to reflect its more 'general dist mgpu'
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114642
Approved by: https://github.com/wanchaol , https://github.com/malfet
2023-11-28 23:06:18 +00:00
e0ec71deab
Fix module: distributed labeler ( #114324 )
...
Removes preceding `/` which was preventing labeler from working. (looks like a typo in the original PR)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/114324
Approved by: https://github.com/XilunWu , https://github.com/fegin
2023-11-22 03:43:14 +00:00
63fc48257a
Configure labeler for 'module: distributed' ( #112812 )
...
To opt-in to getting notified based on this label, simply add yourself
to the summary field at the top of this issue:
https://github.com/pytorch/pytorch/issues/24422
Uses same file paths as current CODEOWNERS.
Note: can easily add sub-labels for components within distributed if we
want.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112812
Approved by: https://github.com/wanchaol , https://github.com/fduwjj
2023-11-03 21:51:49 +00:00
058cb70ad9
[CI] Add auto label rule for torch/_export ( #111181 )
...
Summary: Auto label all torch/_export changes with ciflow/inductor to trigger AOTInductor tests.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/111181
Approved by: https://github.com/angelayi
2023-10-13 20:07:47 +00:00
e1502c0cdb
Add some more files to ciflow/inductor ( #105112 )
...
Signed-off-by: Edward Z. Yang <ezyang@meta.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/105112
Approved by: https://github.com/albanD
2023-07-13 14:44:42 +00:00
73e179a5ca
Follow file move for functorch bits for ciflow/inductor ( #105019 )
...
Signed-off-by: Edward Z. Yang <ezyang@meta.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/105019
Approved by: https://github.com/Skylion007
2023-07-11 23:29:08 +00:00
f6b4106554
[export] Automatically add label for export ( #103458 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103458
Approved by: https://github.com/clee2000
2023-06-13 08:24:01 +00:00
1b31665e78
Make all CI commit pin changes trigger ciflow/inductor ( #103443 )
...
Signed-off-by: Edward Z. Yang <ezyang@meta.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/103443
Approved by: https://github.com/desertfire , https://github.com/malfet
2023-06-12 18:12:45 +00:00
efed5e1c47
Fix triton auto update pin workflow ( #100211 )
...
* allow pytorchbot to approve pin changes
* add ciflow/inductor label when pin changes
Pull Request resolved: https://github.com/pytorch/pytorch/pull/100211
Approved by: https://github.com/huydhn
2023-04-28 17:06:31 +00:00
55b661137f
[inductor] Use decomposition for smooth_l1_loss_backward ( #100242 )
...
Summary: This forward fixes a CI check failure introduced by
https://github.com/pytorch/pytorch/pull/99429 . I also updated
auto-labeler rule to trigger inductor test for any decomposition
changes.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/100242
Approved by: https://github.com/bertmaher
2023-04-28 13:23:20 +00:00
19833486dc
Autorun binary builds when a commit pin is updated ( #96526 )
...
Automatically trigger binary builds when a commit pin is updated to ensure the new versions actually get tested.
This is to prevent a recurrence of the build breaks introduced by https://github.com/pytorch/pytorch/pull/95896#issuecomment-1463312996
Pull Request resolved: https://github.com/pytorch/pytorch/pull/96526
Approved by: https://github.com/huydhn , https://github.com/seemethere
2023-03-13 18:33:29 +00:00
4e02ad7538
Rename inductor collectives test ( #95889 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/95889
Approved by: https://github.com/kumpera
2023-03-03 04:58:39 +00:00
a8cbf70ffc
Inductor support for aten::all_reduce ( #93111 )
...
Pull Request resolved: https://github.com/pytorch/pytorch/pull/93111
Approved by: https://github.com/jansel , https://github.com/wanchaol
2023-02-17 04:42:04 +00:00
7cf0913909
Correct the label for quantization PRs ( #89888 )
...
Summary:
att
Test Plan:
NA
Reviewers:
Subscribers:
Tasks:
Tags:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/89888
Approved by: https://github.com/andrewor14
2022-11-30 23:06:49 +00:00
ed9cd47e31
Add AOTAutograd and partitioner to ciflow/inductor ( #89772 )
...
Signed-off-by: Edward Z. Yang <ezyang@fb.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/89772
Approved by: https://github.com/albanD
2022-11-28 18:39:42 +00:00
2661ff10a9
Include test/distributed/test_dynamo_distributed.py for ciflow/inductor ( #89755 )
...
Signed-off-by: Edward Z. Yang <ezyang@fb.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/89755
Approved by: https://github.com/anjali411
2022-11-28 15:02:09 +00:00
708c050af9
Add labeler with cpu, mkldnn, amp, NNC and quantization paths to start ( #87690 )
...
This PR is to dd labeler with `module: cpu`, `module: mkldnn`, `module: amp (automated mixed precision)`, `NNC` and `oncall: quantization' paths to start.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/87690
Approved by: https://github.com/ezyang , https://github.com/malfet
2022-11-01 02:06:30 +00:00
496acb6602
Add fake tensor files to ciflow/inductor ( #88052 )
...
Signed-off-by: Edward Z. Yang <ezyang@fb.com >
Pull Request resolved: https://github.com/pytorch/pytorch/pull/88052
Approved by: https://github.com/anijain2305
2022-10-31 05:35:54 +00:00
b1cf377cce
Enable inductor CI for huggingface ( #86792 )
...
Summary: Unit tests will be enabled after fixed in trunck. TorchBench and TIMM need
more setup and are coming later.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/86792
Approved by: https://github.com/jansel , https://github.com/huydhn
2022-10-21 01:38:46 +00:00
8a8cd092c8
Add labeler with dynamo/inductor paths to start ( #87024 )
...
The other missing ingredient is getting CC bot to work on labels on PRs
Pull Request resolved: https://github.com/pytorch/pytorch/pull/87024
Approved by: https://github.com/soumith , https://github.com/jansel
2022-10-16 06:13:18 +00:00
