Will Constable e665f03ad8 Fix dynamo func defaults handling for torch.device, size, dtype (#92880) ... Previously, these torch types were not handled in the wrap_bound_arg handler. Add a unit test and verify it is fixed. Fixes #91084 Pull Request resolved: https://github.com/pytorch/pytorch/pull/92880 Approved by: https://github.com/ezyang		2023-01-24 21:50:43 +00:00
..
__init__.py	Properly guard all numpy usage within dynamo and remove UnspecializedNumpyVariable (#90795)	2023-01-06 22:36:38 +00:00
base.py	Cache guards once per variable tracker, rather than re-propagating them repeatedly (#89827)	2022-12-02 01:45:05 +00:00
builder.py	Remove is_aot_autograd_safe_to_run (#91927)	2023-01-21 23:54:48 +00:00
builtin.py	Propagate sources in VariableBuilder and add SuperSource (#91729)	2023-01-12 05:04:18 +00:00
constant.py	Properly guard all numpy usage within dynamo and remove UnspecializedNumpyVariable (#90795)	2023-01-06 22:36:38 +00:00
dicts.py	Cache guards once per variable tracker, rather than re-propagating them repeatedly (#89827)	2022-12-02 01:45:05 +00:00
functions.py	Fix dynamo func defaults handling for torch.device, size, dtype (#92880)	2023-01-24 21:50:43 +00:00
lists.py	[dynamo] Support dynamic slicing (#91341)	2023-01-10 21:23:55 +00:00
misc.py	[Dynamo] Make torch.autograd.Function.forward support graph break and no re-compilation (#91295)	2023-01-20 06:25:09 +00:00
nn_module.py	[quant][pt2e] Support setting qconfig by module_type (#92355)	2023-01-20 03:18:21 +00:00
tensor.py	Delete get_pyobj() entirely (#92638)	2023-01-20 19:06:56 +00:00
torch.py	[dynamo] Support control flow map() operator. (#91939)	2023-01-19 22:03:01 +00:00
user_defined.py	[Dynamo] Fix calling UserDefinedObject.func should pass self object (#92050)	2023-01-21 05:47:01 +00:00