frozenleaves/pytorch
1
0
Fork 0
mirror of https://github.com/pytorch/pytorch.git synced 2025-10-20 21:14:14 +08:00
Code Issues Packages Projects Releases Wiki Activity

Revert "Deprecate registering autograd kernels at not an autograd key (#104481)"

Browse Source 
This reverts commit ed13ab666419ae5dd3adbdb048c8f96f62b14b3d.

Reverted https://github.com/pytorch/pytorch/pull/104481 on behalf of https://github.com/atalman due to failed in periodic tests ([comment](https://github.com/pytorch/pytorch/pull/104481#issuecomment-1631552846))
This commit is contained in:
PyTorch MergeBot
2023-07-11 21:48:22 +00:00
parent 2f95a3d0fc
commit 24aa8b9b9a
10 changed files with 28 additions and 651 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
test/test_dispatch.py
View File

@ -391,7 +391,7 @@ CPU: impl_t_t [kernel]
CUDA: default_def_name_t_t [math kernel]
XLA: default_def_name_t_t [math kernel]
AutogradOther: default_def_name_t_t [math kernel]
AutogradCPU: registered in pytorch framework [backend fallback]
AutogradCPU: fallthrough registered in pytorch framework [backend fallback]
AutogradCUDA: default_def_name_t_t [math kernel]
AutogradXLA: default_def_name_t_t [math kernel]
''')
@ -456,7 +456,7 @@ CPU: fn_cpu [kernel]
CUDA: fn_math [math kernel]
XLA: fn_math [math kernel]
AutogradOther: fn_math [math kernel]
AutogradCPU: registered in pytorch framework [backend fallback]
AutogradCPU: fallthrough registered in pytorch framework [backend fallback]
AutogradCUDA: fn_math [math kernel]
AutogradXLA: fn_math [math kernel]
''')
@ -587,10 +587,10 @@ Undefined: fn_defaultbackend [default backend kernel]
CPU: fn_cpu [kernel]
CUDA: fn_defaultbackend [default backend kernel]
XLA: fn_defaultbackend [default backend kernel]
AutogradOther: registered in pytorch framework [backend fallback]
AutogradCPU: registered in pytorch framework [backend fallback]
AutogradCUDA: registered in pytorch framework [backend fallback]
AutogradXLA: registered in pytorch framework [backend fallback]
AutogradOther: fallthrough registered in pytorch framework [backend fallback]
AutogradCPU: fallthrough registered in pytorch framework [backend fallback]
AutogradCUDA: fallthrough registered in pytorch framework [backend fallback]
AutogradXLA: fallthrough registered in pytorch framework [backend fallback]
''')
def test_computed_table_with_cpu_autograd_defaultbackend(self):
@ -814,9 +814,9 @@ XLA fn_XLA [kernel]
Lazy fn_Lazy [kernel]
FPGA fn_CompositeImplicitAutograd [math kernel]
AutogradOther fn_CompositeImplicitAutograd [math kernel]
AutogradCPU [backend fallback]
AutogradXLA [backend fallback]
AutogradLazy [backend fallback]
AutogradCPU fallthrough [backend fallback]
AutogradXLA fallthrough [backend fallback]
AutogradLazy fallthrough [backend fallback]
'''
)
@ -836,8 +836,8 @@ Lazy fn_Lazy [kernel]
FPGA fn_CompositeImplicitAutograd [math kernel]
AutogradOther fn_CompositeImplicitAutograd [math kernel]
AutogradCPU fn_AutogradCPU [kernel]
AutogradXLA [backend fallback]
AutogradLazy [backend fallback]
AutogradXLA fallthrough [backend fallback]
AutogradLazy fallthrough [backend fallback]
'''
)
self.assertExpectedInline(
@ -869,10 +869,10 @@ CPU fn_CPU [kernel]
XLA fn_XLA [kernel]
Lazy fn_Lazy [kernel]
FPGA fn_CompositeExplicitAutograd [default backend kernel]
AutogradOther [backend fallback]
AutogradOther fallthrough [backend fallback]
AutogradCPU fn_AutogradCPU [kernel]
AutogradXLA [backend fallback]
AutogradLazy [backend fallback]
AutogradXLA fallthrough [backend fallback]
AutogradLazy fallthrough [backend fallback]
'''
)
@ -906,7 +906,7 @@ XLA fn_CompositeImplicitAutograd [math kernel]
Lazy fn_CompositeImplicitAutograd [math kernel]
FPGA fn_FPGA [kernel]
AutogradOther ambiguous_autogradother [ambiguous autogradother]
AutogradCPU [backend fallback]
AutogradCPU fallthrough [backend fallback]
AutogradXLA fn_CompositeImplicitAutograd [math kernel]
AutogradLazy fn_CompositeImplicitAutograd [math kernel]
'''