diff options
author | Eric Kunze <eric.kunze@arm.com> | 2021-09-21 16:56:20 -0700 |
---|---|---|
committer | Eric Kunze <eric.kunze@arm.com> | 2021-09-24 16:19:50 -0700 |
commit | b37e7fb865744536b708c8baa49b6415eab07b04 (patch) | |
tree | 179ca5c9322f342993c82a67a16208c810d4ba10 /chapters/introduction.adoc | |
parent | f1b531ac7a0a045301d39ba14d5c1dc5dc691ddc (diff) | |
download | specification-b37e7fb865744536b708c8baa49b6415eab07b04.tar.gz |
Don't allow pooling ops with excessive padding
Padding larger than the kernel size leads to the
odd effect where no values from the input tensor are used.
This leads to a divide by 0 in AVG_POOL2D, as it only divides
by the number of valid input values used, not padding.
Change MAX_POOL2D to have the same limitation for consistency.
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
Change-Id: I9cd0c0619f88f570d30c83053280460ae3d1cbfb
Diffstat (limited to 'chapters/introduction.adoc')
0 files changed, 0 insertions, 0 deletions