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.

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

Align kernel_sum/ effective_bias useage #146

Merged
merged 1 commit into from
Sep 16, 2024

Conversation

AdrianLundell
Copy link
Collaborator

Use
kernel_sum[i] = SUM_j(kernel[i][j]input_offset + input_offsetfilter_offset) + bias[i]

for all occurrences of kernel_sums in int8 operators, mirroring the behavior of effective_bias in the LSTM operator. This avoids one extra read/write per output.

Use
   kernel_sum[i] = SUM_j(kernel[i][j]*input_offset + input_offset*filter_offset) + bias[i]

for all occurences of kernel_sums in int8 operators, mirroring the behavour of effective_bias
in the LSTM operator. This avoids one extra read/write per output.

Change-Id: If31c3122b7b6bfaa8a1632436d31d20937d6c13d
@AdrianLundell AdrianLundell merged commit 21a5bba into ARM-software:main Sep 16, 2024
2 of 3 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant