Closed daisyden closed 2 days ago
with oneAPI 2025.0.1 upgrade, Compiler: l_compiler_p_2025.0.1.1242 We got fail on this case:
test_reductions_xpu.py::TestReductionsXPU::test_mode_large_xpu_float32
It can be reproduced on this case, we expect to get 1 in each row of value variable, but got 0.
tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], device='xpu:0')
import torch a=torch.FloatTensor([[ 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.], [ 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.], [ 1., 1., 23., 24., 25., 26., 27., 28., 29., 30.], [ 1., 1., 33., 34., 35., 36., 37., 38., 39., 40.], [ 1., 1., 43., 44., 45., 46., 47., 48., 49., 50.], [ 1., 1., 53., 54., 55., 56., 57., 58., 59., 60.], [ 1., 1., 63., 64., 65., 66., 67., 68., 69., 70.], [ 1., 1., 73., 74., 75., 76., 77., 78., 79., 80.], [ 1., 1., 83., 84., 85., 86., 87., 88., 89., 90.], [ 1., 1., 93., 94., 95., 96., 97., 98., 99., 100.]]).xpu() value, indice = torch.mode(a, -1, False) print(value)
2025 oneapi
🐛 Describe the bug
with oneAPI 2025.0.1 upgrade, Compiler: l_compiler_p_2025.0.1.1242 We got fail on this case:
test_reductions_xpu.py::TestReductionsXPU::test_mode_large_xpu_float32
It can be reproduced on this case, we expect to get 1 in each row of value variable, but got 0.
tensor([0., 0., 0., 0., 0., 0., 0., 0., 0., 0.], device='xpu:0')
Versions
2025 oneapi