feat: support float8 e8m0 scale nn transdata-ops-math-AtomGit

Star0

Star0

Open

feat: support float8 e8m0 scale nn transdata #1

jayshuAdd on Aug 11Merged fromfeat/transdata-scale-nn into master

feat: support float8 e8m0 scale nn transdata #1

Open

jayshuAdd on Aug 11

Merged fromfeat/transdata-scale-nnintomaster

All

2026-08-26

style: format npu format cast changes

jayshuSubmitted on Aug 26

3e711737

2026-08-25

docs: refine e8m0 scale nz format cast

jayshuSubmitted on Aug 25

ddb711cf

docs: describe e8m0 format cast as affinity format

jayshuSubmitted on Aug 25

8b66f438

fix npu format cast e8m0 nn aclnn path

jayshuSubmitted on Aug 25

35412c17

feat: support float8 e8m0 scale nn transdata

jayshuSubmitted on Aug 25

2b0c1a14

feat: transpose kernel代码切换旧接口到新接口

zhanw_codingwritten in Aug 25, and

cann-robotSubmitted on Aug 25

a4da7271

补充4个算子rt2 infershape实现 + 9个算子常量折叠HOSTCPU改造

cann-robotSubmitted on Aug 25

b21d1fef

[Docs] 修复 LogAddExp aclnn 文档乱码

chensi79written in Aug 25, and

cann-robotSubmitted on Aug 25

bac01298

fix divide zero bug

guijianweiwritten in Aug 25, and

cann-robotSubmitted on Aug 25

cd606ed6

[CANNBOT]: Eltwise 算子共性问题修改

ZZJYzjy-shinewritten in Aug 25, and

cann-robotSubmitted on Aug 25

f0797fd8