i-robot
9eb414b768
!19389 ascend compile reconstruct
Merge pull request !19389 from liubuyu/compile_re
4 years ago
zengzitao
b8095efcab
gk fix warnings in master
4 years ago
linqingke
efab3997f0
enable send and recv op fusion.
4 years ago
chujinjin
7e3fbed627
fix lstm split error
4 years ago
i-robot
8947a11e49
!21544 [GraphKernel] fix bert and add graph kernel ops.
Merge pull request !21544 from chenlei_autodiff/add_sin
4 years ago
lby
a5029f061c
ascend kernel build refactory
4 years ago
i-robot
57bbc7fb94
!21486 Add Op Infer for Graph Kernel C++
Merge pull request !21486 from jiaoy1224/eliminate_empty
4 years ago
chenlei_autodiff
0271535429
[GraphKernel] fix bert and add graph kernel ops.
4 years ago
i-robot
322bb37070
!21045 pangu_train_and_eval
Merge pull request !21045 from yao_yf/pangu_train_and_eval
4 years ago
Yang Jiao
74c96bf4ee
add more infer
4 years ago
i-robot
caee1ec751
!21530 code review of gpu backend
Merge pull request !21530 from limingqi107/bug_fix
4 years ago
i-robot
ef2d3ad850
!21507 fix insert cast in cpu pynative mode
Merge pull request !21507 from baihuawei/insert_cast_fix
4 years ago
guyuehuo
55a2f23d1c
add prelu Optimized pass
4 years ago
limingqi107
5766234426
code review of gpu backend
4 years ago
dayschan
78dfab9c52
Save and restore the output abstract of original graph
For convenience, we may change some operators' shape in equivalent way,
such as changing the scalar value (shape is empty) to a tensor with shape [1].
It's ok for the intermediate tensors, but not for the outputs.
So we save the output shapes in pre-process stage, and restore them in post-process stage.
4 years ago
baihuawei
4a3717043b
fix insert cast in cpu pynative mode
4 years ago
i-robot
2f4fc75fcc
!21396 modify the limit of axis of reduce ops
Merge pull request !21396 from Simson/opinfer
4 years ago
i-robot
09dba31600
!21444 Fix Reshape Attrs
Merge pull request !21444 from jiaoy1224/master
4 years ago
i-robot
3693625d6f
!21395 fix same node is used by two comm op
Merge pull request !21395 from zhoufeng/xiu-ba-ge
4 years ago
simson
5b34859541
modify the limit of axis of reduce ops
4 years ago
Yang Jiao
2b4a784b86
fix reshape
4 years ago
hwjiaorui
e97df3a58f
clean code
4 years ago
yao_yf
1203413d32
pangu train and eval
4 years ago
i-robot
9629b4ebd9
!21071 [Graph Kernel] Add Low Precision Optimition
Merge pull request !21071 from cyun/master_729
4 years ago
zhoufeng
b09d411dc4
fix same node is used by two comm op
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
4 years ago
cy
4105a247b7
add low precison
4 years ago
i-robot
8610b7e38d
!21301 fix kernel size bug in conv2d
Merge pull request !21301 from Simson/opinfer
4 years ago
i-robot
85999a87a7
!21279 code clean master
Merge pull request !21279 from zhaosida/code_clean_master
4 years ago
i-robot
b917ceca36
!21248 clean code
Merge pull request !21248 from zhoufeng/code-check
4 years ago
simson
d458699df2
fix kernel size bug in conv2d
4 years ago
i-robot
28857c7f3a
!18154 reduce infer with dynamic shape
Merge pull request !18154 from Simson/opinfer
4 years ago
zhaosida
00118a1467
code clean master
4 years ago
i-robot
7ca2d0758d
!21147 cpu inset cast for graph output node
Merge pull request !21147 from baihuawei/insert_cast
4 years ago
simson
1bfbb6c4f4
reduce infer with dynamic shape
4 years ago
zhoufeng
67a3558c0e
clean code
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
4 years ago
i-robot
645e32dde7
!21135 pclint code clean
Merge pull request !21135 from yuchaojie/code-clean
4 years ago
zhoufeng
2143241092
change neighbor exchange to all to all
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
4 years ago
baihuawei
20b3942ca6
inset cast
4 years ago
i-robot
ef182b31d6
!20928 Fix network's AR and optimizer cannot be in parallel.
Merge pull request !20928 from linqingke/highest-performance
4 years ago
yuchaojie
210bfeb6e0
pclint code clean
4 years ago
i-robot
7b55c9858b
!21103 Reorganize Empty Graph
Merge pull request !21103 from jiaoy1224/eliminate_empty
4 years ago
i-robot
0816f95653
!21091 unified runtime codedex fixed
Merge pull request !21091 from limingqi107/bug_fix4
4 years ago
i-robot
92362f8c94
!21053 code check clean
Merge pull request !21053 from yuchaojie/code-clean
4 years ago
linqingke
7a463a885e
Fix network's AR and optimizer cannot be in parallel.
4 years ago
Yang Jiao
fd7ab25fc2
reorganiz empty graph
4 years ago
i-robot
7b20a5adf7
!20966 Update graph kernel support for argmax/argmin
Merge pull request !20966 from zichun_ye/argmax_min_kernelgraph
4 years ago
limingqi107
8c690e3d18
unified runtime codedex fixed
4 years ago
i-robot
65ec61a480
!21033 Modify Double Comparison for Arithmetic Simplify
Merge pull request !21033 from jiaoy1224/master
4 years ago
Zichun Ye
22172f18bc
update graph kernel support for argmax/argmin
fix pylint problem
fix conflict
fix op list
fix check warning
fix code based on review comments
update akg commit
fix check warning
4 years ago
yuchaojie
905e798423
code check clean
4 years ago