diff options
author | Rickard Bolin <rickard.bolin@arm.com> | 2024-06-26 13:17:17 +0000 |
---|---|---|
committer | Rickard Bolin <rickard.bolin@arm.com> | 2024-06-26 14:46:27 +0000 |
commit | 86ba3fd1eef6b6ad9047ef4aaed3461e0b250a1a (patch) | |
tree | cf1f2b5e47828b5f13842754ec007de10c6c16ab /ethosu/vela/tensor_allocation.py | |
parent | c8450913d7f557a3cf6d547caebe691f41d2a00d (diff) | |
download | ethos-u-vela-86ba3fd1eef6b6ad9047ef4aaed3461e0b250a1a.tar.gz |
MLBEDSW-9222: Fix rescale precision errors
After updating to NumPy 2.0, some variables that were previously
implicitly promoted to doubles instead remained floats, causing
precision loss. They are now explicitly promoted to doubles instead.
Change-Id: Ibc1838e2a3a05116e291a6320525b38972f0478e
Signed-off-by: Rickard Bolin <rickard.bolin@arm.com>
Diffstat (limited to 'ethosu/vela/tensor_allocation.py')
0 files changed, 0 insertions, 0 deletions