[Compiler Toolkit] Add annotations to MoE #1937
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
[ghstack-poisoned]
meta-cla
bot
added
the
CLA Signed
SherlockNoMad
changed the title
anijain2305
approved these changes
Oct 27, 2025
sample output
```
[rank0]: # Annotation: {'EP': 'dispatch'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:485 in all_to_all_single, code: tensor = torch.ops._c10d_functional.all_to_all_single( # type: ignore[attr-defined]
[rank0]: tensor_3: "i64[8]" = torch.ops._c10d_functional.all_to_all_single(num_tokens_per_expert_3, [4, 4], [4, 4], '11')
[rank0]:
[rank0]: # Annotation: {'EP': 'dispatch'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:136 in wait_tensor, code: return torch.ops._c10d_functional.wait_tensor(tensor) # type: ignore[attr-defined]
[rank0]: num_tokens_per_expert_group_2: "i64[8]" = torch.ops._c10d_functional.wait_tensor(tensor_3); tensor_3 = None
```
```
**[rank0]: # Annotation: {'EP': 'combine'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:522 in all_to_all_single_autograd, code: tensor = torch.ops._c10d_functional_autograd.all_to_all_single( # type: ignore[attr-defined]
[rank0]: slice_20: "bf16[u18 + u19, 256]" = torch.ops.aten.slice.Tensor(index_put_6, 0, 0, -1); index_put_6 = None
[rank0]: all_to_all_single_14: "bf16[u16 + u17, 256]" = torch.ops._c10d_functional.all_to_all_single.default(slice_20, [_local_scalar_dense_16, _local_scalar_dense_17], [_local_scalar_dense_18, _local_scalar_dense_19], '11'); slice_20 = None
[rank0]:
[rank0]: # Annotation: {'EP': 'combine'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:528 in all_to_all_single_autograd, code: return _FromTorchTensor.apply(tensor)
[rank0]: wait_tensor_136: "bf16[u16 + u17, 256]" = torch.ops._c10d_functional.wait_tensor.default(all_to_all_single_14); all_to_all_single_14 = None
[rank0]:
```
[ghstack-poisoned]
yiming0416
approved these changes
Oct 27, 2025
SherlockNoMad
added a commit
that referenced
this pull request
Oct 27, 2025
Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #1937 * #1906 sample output ``` [rank0]: # Annotation: {'EP': 'dispatch'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:485 in all_to_all_single, code: tensor = torch.ops._c10d_functional.all_to_all_single( # type: ignore[attr-defined] [rank0]: tensor_3: "i64[8]" = torch.ops._c10d_functional.all_to_all_single(num_tokens_per_expert_3, [4, 4], [4, 4], '11') [rank0]: [rank0]: # Annotation: {'EP': 'dispatch'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:136 in wait_tensor, code: return torch.ops._c10d_functional.wait_tensor(tensor) # type: ignore[attr-defined] [rank0]: num_tokens_per_expert_group_2: "i64[8]" = torch.ops._c10d_functional.wait_tensor(tensor_3); tensor_3 = None ``` ``` **[rank0]: # Annotation: {'EP': 'combine'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:522 in all_to_all_single_autograd, code: tensor = torch.ops._c10d_functional_autograd.all_to_all_single( # type: ignore[attr-defined] [rank0]: slice_20: "bf16[u18 + u19, 256]" = torch.ops.aten.slice.Tensor(index_put_6, 0, 0, -1); index_put_6 = None [rank0]: all_to_all_single_14: "bf16[u16 + u17, 256]" = torch.ops._c10d_functional.all_to_all_single.default(slice_20, [_local_scalar_dense_16, _local_scalar_dense_17], [_local_scalar_dense_18, _local_scalar_dense_19], '11'); slice_20 = None [rank0]: [rank0]: # Annotation: {'EP': 'combine'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:528 in all_to_all_single_autograd, code: return _FromTorchTensor.apply(tensor) [rank0]: wait_tensor_136: "bf16[u16 + u17, 256]" = torch.ops._c10d_functional.wait_tensor.default(all_to_all_single_14); all_to_all_single_14 = None [rank0]: ```
SherlockNoMad
added a commit
that referenced
this pull request
Oct 27, 2025
Stack from [ghstack](https://github.com/ezyang/ghstack) (oldest at bottom): * __->__ #1937 * #1906 sample output ``` [rank0]: # Annotation: {'EP': 'dispatch'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:485 in all_to_all_single, code: tensor = torch.ops._c10d_functional.all_to_all_single( # type: ignore[attr-defined] [rank0]: tensor_3: "i64[8]" = torch.ops._c10d_functional.all_to_all_single(num_tokens_per_expert_3, [4, 4], [4, 4], '11') [rank0]: [rank0]: # Annotation: {'EP': 'dispatch'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:136 in wait_tensor, code: return torch.ops._c10d_functional.wait_tensor(tensor) # type: ignore[attr-defined] [rank0]: num_tokens_per_expert_group_2: "i64[8]" = torch.ops._c10d_functional.wait_tensor(tensor_3); tensor_3 = None ``` ``` **[rank0]: # Annotation: {'EP': 'combine'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:522 in all_to_all_single_autograd, code: tensor = torch.ops._c10d_functional_autograd.all_to_all_single( # type: ignore[attr-defined] [rank0]: slice_20: "bf16[u18 + u19, 256]" = torch.ops.aten.slice.Tensor(index_put_6, 0, 0, -1); index_put_6 = None [rank0]: all_to_all_single_14: "bf16[u16 + u17, 256]" = torch.ops._c10d_functional.all_to_all_single.default(slice_20, [_local_scalar_dense_16, _local_scalar_dense_17], [_local_scalar_dense_18, _local_scalar_dense_19], '11'); slice_20 = None [rank0]: [rank0]: # Annotation: {'EP': 'combine'} File: /data/users/bahuang/pytorch/torch/distributed/_functional_collectives.py:528 in all_to_all_single_autograd, code: return _FromTorchTensor.apply(tensor) [rank0]: wait_tensor_136: "bf16[u16 + u17, 256]" = torch.ops._c10d_functional.wait_tensor.default(all_to_all_single_14); all_to_all_single_14 = None [rank0]: ```
|
Hi @SherlockNoMad , I have a nit house-keeping request: Could you add a line here https://github.com/pytorch/torchtitan/tree/refs/heads/main/torchtitan/experiments#current-experiments |
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
sample output