More context in [#132471](https://github.com/pytorch/pytorch/issues/132471) and https://github.com/pytorch/pytorch/issues/132366.
TLDR:
When cuda is available and users move tensors to cuda, we cannot really reuse the default pg if default pg is gloo, as lots of collectives are not supported on gloo for cuda tensors. For example, `dtensor.full_tensor()` would result in a mysterious SIGTERM when all_gather a cuda tensor using gloo. Without the change in this PR, users would have to know the context and explicitly move the cuda tensor to cpu before invoking most collectives, which I think is not so ideal UX.
Therefore, given most collectives are not supported on gloo for cuda tensors, we should init a new pg if the default pg is gloo when torch.cuda.is_available() and device_type is cuda.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/132709
Approved by: https://github.com/awgu, https://github.com/wanchaol
**Overview**
This PR supports constructing an ND mesh with `from_group()` by passing in `group: List[ProcessGroup]` and `mesh: Union[torch.Tensor, "ArrayLike"]` together. The `ndim` of the device mesh returned from `from_group()` is equal to the number of `ProcessGroup`s passed. If the `ndim` is greater than 1, then the `mesh` argument is required (since there is no simple way to recover the `mesh` tensor from the process groups otherwise).
This PR also adds `mesh_dim_names` as an argument to forward to the device mesh for convenience.
<details>
<summary> Old Approach </summary>
**Overview**
- This PR mainly adds `mesh_shape` to `from_group()` so that the user can construct an ND (N > 1) device mesh from a process group. This is to unblock HSDP, where we can pass the overall data parallel process group to `from_group()` with `mesh_shape = (replicate_dim_size, shard_dim_size)` and `from_group()` will construct subgroups for the user. (The user can then get the subgroups from the submeshes.)
- Constructing the 2D `DeviceMesh` from an existing shard process group and replicate process group is hard because we cannot easily recover the array of ranks in their parent group on each rank in general.
- This PR also adds `mesh_dim_names` to `from_group()` so that the user can name the mesh dimensions of the constructed device mesh.
</details>
Pull Request resolved: https://github.com/pytorch/pytorch/pull/126258
Approved by: https://github.com/wanchaol
This PR adds a `DeviceMesh.from_group()` static method to convert an existing process group to a device mesh.
Motivation: We need `DeviceMesh.from_group()` to allow FSDP2 to interoperate with distributed libraries that do not use `DeviceMesh` for all parallelisms.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/124787
Approved by: https://github.com/wanchaol
ghstack dependencies: #124651, #124741, #124767, #124768, #124780
This PR adds a private init backend option, to tackle the issues sub
mesh creation:
in device mesh slicing we don't want to create process groups again,
so explicitly turn the group creation off it's useful
Also I think there might be more submesh creation functionality so
having this flag would ensure that there's no new group created
Differential Revision: [D56497780](https://our.internmc.facebook.com/intern/diff/D56497780)
Pull Request resolved: https://github.com/pytorch/pytorch/pull/124780
Approved by: https://github.com/awgu
## Summary
After this PR, the functional collective Python APIs will stop honoring `TORCH_DISABLE_NATIVE_FUNCOL` and only use native funcol ops. Specifically, this PR:
- Removed `use_native_funcol()`.
- Removed the code path in the Python APIs when `use_native_funcol()` is `False`.
- Changed the CI tests that runs on both native funcol and legacy funcol through the Python API to only run with native funcol.
## Test Changes
`test_functional_api.py`
- Removed the tests where only one of output_split_sizes or input_split_sizes is specified. This behavior is unreliable has been removed from the native funcol.
- Removed `TestWaitiness` which tests an implementation detail of the legacy funcol. We have equivalent tests for native funcol in `test/distributed/test_c10d_functional_native.py` b7fac76fc2/test/distributed/test_c10d_functional_native.py (L114-L116)
`test/distributed/_tensor/test_dtensor.py`
`test/distributed/_tensor/test_dtensor_compile.py`
`test/distributed/test_device_mesh.py`
`test/distributed/_tensor/experimental/test_tp_transform.py`
`test/distributed/_tensor/test_matrix_ops.py`
`test/distributed/test_inductor_collectives.py`
- All these tests were double running with both native funcol and legacy funcol. Changed to only run with native funcol.
`test/distributed/test_c10d_functional_native.py`
- Removed the `run_with_native_funcol` decorators.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/123777
Approved by: https://github.com/wanchaol
ghstack dependencies: #123776
Fixes#118849
Add a map for parent_to_child_mappings in _mesh_resources so we can cache and reuse submesh slicing result so that we can avoid recreating submesh and the underlying sub pg repeatedly, which could lead to funky behaviors.
We will follow up with reusing pg from the parent_mesh during submesh creation.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/122975
Approved by: https://github.com/wanchaol
Summary:
Original commit changeset: e52b8809c8d8
Original Phabricator Diff: D54778906
We have to backout this diff.
D54778906 seems to be causing test failures for APF blocking trunk health and hence release. Just starting to look at the issue. T182209248
Test Plan: Sandcastle
Reviewed By: satgera
Differential Revision: D54825114
Pull Request resolved: https://github.com/pytorch/pytorch/pull/121763
Approved by: https://github.com/osalpekar
```
Between the time we switch to the native funcol by default and the time when
we are confident that we can remove the legacy implementation, we want to
ensure that the legacy funcol remains covered by unit tests. This is to
prepare for any potential (but unlikely) reverts. The following utilities
help achieve this goal.
run_with_{native,legacy}_funcol - mark a test to run with only
{native,legacy} funcol. These decorators are for impl specific tests (e.g.
verifying generated code with FileCheck).
run_with_both_funcol_impls - parametrize a test to run with both legacy and
native funcol.
run_with_both_funcol_impls_with_arg - same as run_with_both_funcol_impls, but
passes `enable_native_funcol` to the test so impl specific checks can be
carried out.
```
This PR also marks some tests we want to cover in this fashion. More tests will be marked in subsequent PRs.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/119950
Approved by: https://github.com/wanchaol
ghstack dependencies: #119881
Summary:
The reuse subgroup logic is causing GLOO to timeout on two internal modelstore tests (relevant tests in test plan).
We temporarily disabling re-use subgroup during root-causing to allow the internal tests to be able to run again, as they are now omitted shown in T176426987.
Test Plan:
CI
Pull Request resolved: https://github.com/pytorch/pytorch/pull/118940
Approved by: https://github.com/wanchaol
### Summary
Run the relevant tests in `test/distributed/_tensor/test_dtensor_compile.py` and `test/distributed/test_device_mesh.py` with native funcol enabled, in addition to with them being disabled.
All tests excepts `test_tp_compile_comm_reordering` pass. This is expected because the native funcols have slightly different IRs, so the reordering pass needs to be adjusted. This test is disabled for now.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/118437
Approved by: https://github.com/LucasLLC
ghstack dependencies: #118910, #118911
### Summary
- Added `group_name` as the third field in `dim_group_infos`.
- `DeviceMeshTest` now runs both w/ and w/0 `_USE_NATIVE_C10D_FUNCTIONAL=1` in CI.
### Other fixes
- Convert `reduceOp` to lower case before passing it into c10d_functional ops.
- Added a finalizer to handle unwaited collectives (this mirrors the treatment for Python functional collective ops).
Pull Request resolved: https://github.com/pytorch/pytorch/pull/118423
Approved by: https://github.com/wanchaol, https://github.com/LucasLLC, https://github.com/wconstab
Currently, we create new_group for sub_group pg during mesh initialization. The PR changes this so we will:
1) re-use sub_group pg if it exsits,
2) create new sub_group pg if it does not exist.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115716
Approved by: https://github.com/wanchaol
Summary:
This change is to make the input tensor contiguous for DTensor reduce scatter in the case no padding is needed.
There's no exception thrown during training, but we ran into numerical value correctness issue without the change.
Test Plan:
**CI**
CI test
**WHEN model test**:
- Verified loss for each iteration within the expected range.
- Verified NE on-par with this change with 4B training data.
Differential Revision: D52170822
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115847
Approved by: https://github.com/wanchaol
Summary:
Rename _device_mesh.py to device_mesh.py, update all callsites, add documentation.
We created stubs for public class and methods in torch.distributed.device_mesh so that torch.distributed.device_mesh can be imported with or without distributed is available().
Original diff reverted: D51629761
Original PR reverted: https://github.com/pytorch/pytorch/pull/115099
Prior to landing, CI signals are all passed. Shipit added the "ci/trunk" label to the PR and DID NOT wait for it and went ahead committing. More context can be found in the reverted PR above.
Test Plan: CI.
Differential Revision: D51861018
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115193
Approved by: https://github.com/fegin
Summary:
Rename _device_mesh.py to device_mesh.py, update all callsites, adds documentation.
Original diff reverted: D51629761
Original PR reverted: https://github.com/pytorch/pytorch/pull/114991
It was failing because failing a public module binding tests in MacOS, and this is due to the change in import order for torch/distributed/fsdp/_common_utils.py. Since this original import would still work, we remove the changes in this file.
Test Plan: CI.
Differential Revision: D51825114
Pull Request resolved: https://github.com/pytorch/pytorch/pull/115099
Approved by: https://github.com/wanchaol, https://github.com/fegin
Plan B for https://github.com/pytorch/pytorch/pull/112839
Motivation for the change:
1. We need to remove `funcol` as a dependency for device_mesh.py to resolve circular dependency issues when introducing device_mesh as an arg for DDP. In the meantime, we should not go from funcol to non-funcol as @voznesenskym suggested. Therefore, we want to remove this all_gather check completely.
2. For large scale, it would not make sense to validate the mesh at global scale anyway.
Pull Request resolved: https://github.com/pytorch/pytorch/pull/112928
Approved by: https://github.com/wanchaol
