Skip to content
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

pow_Tensor_Scalar: fixing error result when scalar exp is a bool. #638

Merged
merged 4 commits into from
Jul 24, 2024
Merged

pow_Tensor_Scalar: fixing error result when scalar exp is a bool. #638

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
c118c74
fix bool pow bug
yuchengliu1 Jul 23, 2024
88a6e91
update skip
yuchengliu1 Jul 23, 2024
9759514
update
yuchengliu1 Jul 23, 2024
c5736a6
Merge branch 'main' into yucheng_pow
fengyuan14 Jul 24, 2024
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
8 changes: 7 additions & 1 deletion src/ATen/native/xpu/Pow.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,7 +58,13 @@ Tensor& XPUNativeFunctions::pow_out(
Tensor XPUNativeFunctions::pow(const Tensor& self, const Scalar& exponent) {
Tensor out;
auto iter = pow_tensor_scalar_meta(self, exponent, out);
native::xpu::pow_tensor_scalar_kernel(iter, exponent);
if (exponent.equal(0.0) || exponent.equal(false)) {
iter.output().fill_(1);
} else if (exponent.equal(1.0) || exponent.equal(true)) {
iter.output().copy_(self);
} else {
native::xpu::pow_tensor_scalar_kernel(iter, exponent);
}
return iter.output();
}

Expand Down
7 changes: 4 additions & 3 deletions test/xpu/run_test_with_skip.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -797,12 +797,13 @@ def launch_test(test_case, skip_list=None, exe_list=None):
skip_list = (
"test_fmod_remainder_by_zero_integral_xpu_int64", # zero division is an undefined behavior: different handles on different backends
"test_div_rounding_numpy_xpu_float16", # Calculation error. XPU implementation uses opmath type.
# RuntimeError: false INTERNAL ASSERT FAILED at "torch-xpu-ops/src/ATen/native/xpu/sycl/PowKernels.cpp":233, please report a bug to PyTorch. invalid combination of type in Pow function, common dtype: Short, exp is integral? 0
# fail in complex_exponents=[-1.0 - 1.5j, 3.3j]
# Mismatched elements: 33 / 100 (33.0%)
# Greatest absolute difference: 0.00038337233127094805 at index (4,) (up to 1e-05 allowed)
# Greatest relative difference: 1.9085073290625587e-06 at index (6,) (up to 1.3e-06 allowed)
"test_pow_xpu_int16",
"test_pow_xpu_int32",
"test_pow_xpu_int64",
"test_pow_xpu_int8",
"test_pow_xpu_uint8",
# AssertionError: Jiterator is only supported on CUDA and ROCm GPUs, none are available.
"_jiterator_",
# Unexpected success
Expand Down