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

[ca] skip on some PYTORCH_TEST_WITH_DYNAMO=1 autograd tests (#156374)

Browse Source 
These aren't supported. Not sure how they passed CI

Pull Request resolved: https://github.com/pytorch/pytorch/pull/156374
Approved by: https://github.com/jansel
This commit is contained in:
Simon Fan
2025-06-21 11:25:11 -07:00
committed by PyTorch MergeBot
parent fab85fc5f9
commit f1968a5e76
21 changed files with 5 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

0
test/compiled_autograd_skips/TestAutograd.test_callback_propagates_errors_from_device_thread Normal file
View File

0
test/compiled_autograd_skips/TestAutograd.test_graph_save_on_cpu_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_inplace_on_view_python_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_pin_memory_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_reentrant_parent_error_on_cpu_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_sparse_backward_cuda_complex128 Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_sparse_backward_cuda_float64 Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_sparse_ctor_getter_backward_cuda_complex128 Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_sparse_ctor_getter_backward_cuda_float64 Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_sparse_mask_autograd_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradDeviceTypeCUDA.test_where_functional_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradFunctional.test_hessian_vectorize_raises_no_warnings_logging_tensor Normal file
View File

0
test/compiled_autograd_skips/TestAutogradFunctional.test_jacobian_vectorize_correctness_different_devices_base_tensor Normal file
View File

0
test/compiled_autograd_skips/TestAutogradFunctional.test_jacobian_vectorize_correctness_different_devices_logging_tensor Normal file
View File

0
test/compiled_autograd_skips/TestAutogradFunctional.test_jacobian_vectorize_raises_no_warnings_logging_tensor Normal file
View File

0
test/compiled_autograd_skips/TestAutogradMultipleDispatchCUDA.test_autograd_composite_implicit_and_dispatch_registration_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradMultipleDispatchCUDA.test_autograd_multiple_dispatch_registrations_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradMultipleDispatchCUDA.test_backward_single_threaded_cuda Normal file
View File

0
test/compiled_autograd_skips/TestAutogradStreamSynchronization.test_side_stream_backward_overlap Normal file
View File

0
test/compiled_autograd_skips/TestSelectiveActivationCheckpoint.test_flops_and_mem Normal file
View File

6
torch/testing/_internal/common_utils.py
View File

@ -1580,6 +1580,10 @@ TEST_WITH_TORCHDYNAMO: bool = TestEnvironment.def_flag(
env_var="PYTORCH_TEST_WITH_DYNAMO",
implied_by_fn=lambda: TEST_WITH_TORCHINDUCTOR or TEST_WITH_AOT_EAGER,
)
TEST_WITHOUT_COMPILED_AUTOGRAD: bool = TestEnvironment.def_flag(
"TEST_WITHOUT_COMPILED_AUTOGRAD",
env_var="PYTORCH_TEST_WITHOUT_COMPILED_AUTOGRAD",
)
if TEST_WITH_TORCHDYNAMO:
import torch._dynamo
@ -1594,7 +1598,7 @@ if TEST_WITH_TORCHDYNAMO:
torch._inductor.config.fallback_random = True
else:
# only dynamo for now
torch._dynamo.config.compiled_autograd = True
torch._dynamo.config.compiled_autograd = not TEST_WITHOUT_COMPILED_AUTOGRAD
# seems like this is only used in test/torch_np