This is an automated email from the ASF dual-hosted git repository.
github-bot pushed a change to branch nightly
in repository https://gitbox.apache.org/repos/asf/tvm.git
from 67df20faee [Target][TOPI] Use LLVM for x86 CPU feature lookup (#15685)
add 64ab31ec3e [UnitTest][Metal] Parametrize allreduce GPU tests (#15749)
add 08a6ee520f [Hexagon] F2qi avgpool bug fix (#15599)
add 61d5be0470 [Relay][Bugfix] fix the wrong calculate logic of operator
flip in PyTorch frontend (#15752)
No new revisions were added by this update.
Summary of changes:
python/tvm/relay/frontend/pytorch.py | 5 +-
.../transform/fake_quantization_to_integer.py | 31 +++
python/tvm/topi/hexagon/qnn/avg_pool2d.py | 22 +-
.../python/contrib/test_hexagon/infrastructure.py | 22 ++
.../test_hexagon/test_pass_fq2i_avg_pool2d.py | 290 +++++++++++++++++++++
.../test_hexagon/test_relay_simplify_conv_pat.py | 23 +-
tests/python/frontend/pytorch/test_forward.py | 11 +-
.../{test_allreduce_cuda.py => test_allreduce.py} | 109 ++++----
8 files changed, 421 insertions(+), 92 deletions(-)
create mode 100644
tests/python/contrib/test_hexagon/test_pass_fq2i_avg_pool2d.py
rename tests/python/unittest/{test_allreduce_cuda.py => test_allreduce.py}
(50%)
