Skip to content

[AUTOGENERATED] [release/2.8] skip 3D NCHW FP16 batchnorm test due to Native accuracy issue #2652

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Sep 18, 2025
Merged

[AUTOGENERATED] [release/2.8] skip 3D NCHW FP16 batchnorm test due to Native accuracy issue #2652

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 2 additions & 4 deletions test/test_nn.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5180,6 +5180,8 @@ def test_batchnorm_nhwc_cuda(self):
name_fn=lambda f, b, m, t: f"{f}_vs_{b}{'_mixed' if m else ''}_{dtype_name(t)}"
)
def test_batchnorm(self, dims, mode, memory_format, ref_backend, mixed, dtype):
if self._testMethodName == "test_batchnorm_3D_train_NCHW_vs_native_mixed_float16":
self.skipTest("3D float16 NCHW train failed on CUDA and ROCm due to Native batchnorm accuracy issue SWDEV-541024")
if torch.version.hip:
if self._testMethodName in ("test_batchnorm_2D_train_NHWC_vs_NCHW_mixed_bfloat16",
"test_batchnorm_2D_train_NCHW_vs_cpu_mixed_bfloat16",
Expand All @@ -5195,10 +5197,6 @@ def test_batchnorm(self, dims, mode, memory_format, ref_backend, mixed, dtype):
) and _get_torch_rocm_version() >= (6, 4):
self.skipTest("bfloat16 NCHW train failed due to native tolerance issue SWDEV-507600")

if self._testMethodName == "test_batchnorm_3D_train_NCHW_vs_native_mixed_float16" \
and _get_torch_rocm_version() < (6, 4):
self.skipTest("3D float16 NCHW train failed on ROCm<=6.3 ")

if dims == 3 and memory_format in ("NHWC", "NCHW"):
memory_format = memory_format + "3D"

Expand Down