diff options
author | Eric Kunze <eric.kunze@arm.com> | 2024-02-14 16:33:31 -0800 |
---|---|---|
committer | Eric Kunze <eric.kunze@arm.com> | 2024-03-13 16:31:19 -0700 |
commit | 0afe61f88ce3d2f445c5f01ae5567cb1b0b7f303 (patch) | |
tree | bfa03c381634090e9074d3e1167c3256a31f3c84 /chapters/tensor_ops.adoc | |
parent | 6dd341093507157aabbea00b90ca8902509cfd4f (diff) | |
download | specification-0afe61f88ce3d2f445c5f01ae5567cb1b0b7f303.tar.gz |
Modify convolution operators to improve bias handling
Accumulator size moves to an enumerated attribute, out_t for floating-point changes
to be the size of the input. Bias for floating-point also becomes the bit width
of the input type.
Signed-off-by: Eric Kunze <eric.kunze@arm.com>
Change-Id: I7369417adbb1106ce34a1978e7f511a30272c318
Diffstat (limited to 'chapters/tensor_ops.adoc')
0 files changed, 0 insertions, 0 deletions