Allow dynamic shapes of tuple type for inputs of dataclass type
#117917
Conversation
BoyuanFeng
requested review from
avikchaudhuri,
gmagogsfm,
tugsbayasgalan and
zhxchen17
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/117917
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (2 Unrelated Failures)As of commit 3c872a2 with merge base b2a3d6b ( FLAKY - The following job failed but was likely due to flakiness present on trunk:
UNSTABLE - The following job failed but was likely due to flakiness present on trunk and has been marked as unstable:
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
|
This pull request was exported from Phabricator. Differential Revision: D52932856 |
…ytorch#117917) Summary: In `torch.export.export(f, args, kwargs, ..., dynamic_shpapes=None, ...)`, `dataclass` is an acceptable type of inputs (for args and kwargs). The `dynamic_shapes` of the `dataclass` inputs needs to be the same `dataclass` type which replaces each tensor attributes with `dynamic_shapes` of the corresponding tensors. (https://github.com/pytorch/pytorch/blob/main/torch/export/dynamic_shapes.py#L375) However, some `dataclass` may have limitations on the types of attributes (e.g., having to be tensors) such that the same `dataclass` cannot be constructed for dynamic shapes. For an input of `dataclass` type, this task enables a `dynamic_shapes` of a tuple type that specifies dynamic shape specifications for each tensor of the input in the same order as the input dataclass type's flatten_fn (https://github.com/pytorch/pytorch/blob/main/torch/utils/_pytree.py#L103) Test Plan: buck test //caffe2/test:test_export Differential Revision: D52932856
BoyuanFeng
force-pushed
the
export-D52932856
branch
from
ebe68f6
to
3c872a2
Compare
|
This pull request was exported from Phabricator. Differential Revision: D52932856 |
|
@pytorchbot merge -f 'Landed internally' (Initiating merge automatically since Phabricator Diff has merged, using force because this PR might not pass merge_rules.json but landed internally) |
Merge startedYour change will be merged immediately since you used the force (-f) flag, bypassing any CI checks (ETA: 1-5 minutes). Please use Learn more about merging in the wiki. Questions? Feedback? Please reach out to the PyTorch DevX Team |
Summary:
In
torch.export.export(f, args, kwargs, ..., dynamic_shpapes=None, ...),dataclassis an acceptable type of inputs (for args and kwargs). Thedynamic_shapesof thedataclassinputs needs to be the samedataclasstype which replaces each tensor attributes withdynamic_shapesof the corresponding tensors. (https://github.com/pytorch/pytorch/blob/main/torch/export/dynamic_shapes.py#L375)However, some
dataclassmay have limitations on the types of attributes (e.g., having to be tensors) such that the samedataclasscannot be constructed for dynamic shapes.For an input of
dataclasstype, this task enables adynamic_shapesof a tuple type that specifies dynamic shape specifications for each tensor of the input in the same order as the input dataclass type's flatten_fn (https://github.com/pytorch/pytorch/blob/main/torch/utils/_pytree.py#L103)Test Plan: buck test //caffe2/test:test_export
Differential Revision: D52932856