OSSForks/pytorch
1
0
Fork 0
mirror of https://github.com/zebrajr/pytorch.git synced 2025-12-07 12:21:27 +01:00
Code Issues Actions 38 Packages Projects Releases Wiki Activity

Fixed error string assertion in test_invalid_devices (#137772)

Browse Source 
ROCm distribution returns different error string for this operation so the test fails this assertion.

Pull Request resolved: https://github.com/pytorch/pytorch/pull/137772
Approved by: https://github.com/Skylion007
This commit is contained in:
iupaikov-amd 2024-10-13 18:10:07 +00:00 committed by PyTorch MergeBot
parent 65d665bae5
commit c09b567a91
1 changed files with 9 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
torch/testing/_internal/distributed/nn/api/remote_module_test.py
View File

@ -12,7 +12,7 @@ from torch.distributed.nn import RemoteModule
from torch.distributed.nn.api.remote_module import _REMOTE_MODULE_PICKLED_ATTRIBUTES
from torch.distributed.nn.api.remote_module import _RemoteModule
from torch.testing._internal.common_distributed import skip_if_lt_x_gpu
from torch.testing._internal.common_utils import TemporaryFileName
from torch.testing._internal.common_utils import TemporaryFileName, TEST_WITH_ROCM
from torch.testing._internal.distributed.rpc.rpc_agent_test_fixture import (
RpcAgentTestFixture,
)
@ -613,8 +613,15 @@ class CudaRemoteModuleTest(CommonRemoteModuleTest):
)
]
if TEST_WITH_ROCM:
errorString = (r"HIP error: invalid device ordinal\n"
r"HIP kernel errors might be asynchronously reported at some other API call, "
r"so the stacktrace below might be incorrect.\n"
r"For debugging consider passing AMD_SERIALIZE_KERNEL=3")
else:
errorString = r"CUDA error: invalid device ordinal"
with self.assertRaisesRegex(
RuntimeError, r"CUDA error: invalid device ordinal"
RuntimeError, errorString
):
[
m.forward()