98dc6eedc2
!3877 Add new hms ops of floor, round and ceil with type of int8
...
Merge pull request !3877 from liuwenhao/master
5 years ago
27a0a2e333
!3880 fix auto mix precision issue in pynative
...
Merge pull request !3880 from wangqiuliang/fix-gradients-issue-in-pynative
5 years ago
952f96d860
!3875 fix reduce for loop size bug
...
Merge pull request !3875 from zhaozhenlong/lite/issue/reduce_for_loop_size_unsign_compare
5 years ago
a182a87a6d
!3820 support tensor attr shape and dtype in graph mode
...
Merge pull request !3820 from zhangbuxue/support_dtype_and_shape_as_attr_in_graph_mode
5 years ago
374b7b8583
Add TBE op SquaredDifference for VM.
5 years ago
a367b3c286
Add new hms ops of floor, round and ceil with type of int8
5 years ago
58e6227a23
fix space to depth bug
5 years ago
577535b387
fix gradients issue in pynative
5 years ago
87323cf0f1
fix thread pool bug
5 years ago
131cad16e8
batch_to_space,depth_to_space,argmin,argmax support int8
5 years ago
201bcdd9af
!3866 add register func for mean op
...
Merge pull request !3866 from fuzhiye/mindspore
5 years ago
b075674cf2
support tensor attr shape and dtype in graph mode
5 years ago
0d77de77e0
add register func for mean op
5 years ago
02c921550e
!3865 fix anf_export quant function
...
Merge pull request !3865 from yankai10/merge_branch
5 years ago
e9c4a697d5
!3868 set tensor allocator
...
Merge pull request !3868 from 张学同/to_merge
5 years ago
15d5cc396d
change type of Shape from int32 to int64
5 years ago
e01036f6ca
!3816 If not set pretrained model,we should not be allowed to run training tasks
...
Merge pull request !3816 from zhouyuanshen/master
5 years ago
88a44a0ef0
optimize arm cpu op: conv_depthwise, deconv_depthwise
5 years ago
a1bdecb9e6
!3856 fix matrixDiag msg , FloorDiv grad, SoftmaxCrossEntropyLogit mag
...
Merge pull request !3856 from fangzehua/fix_matrix_msg
5 years ago
36b8941412
!3873 fix quantdtypecast
...
Merge pull request !3873 from zhongligeng/master
5 years ago
aa6be43437
fix reduce for loop size bug
5 years ago
be507b96e0
!3473 [refine]change base class of parameter
...
Merge pull request !3473 from vlne-v1/change_base_class_of_parameter
5 years ago
8beb1b0fb5
!3799 conv1x1 & deconv change
...
Merge pull request !3799 from ling/conv1x1
5 years ago
f926650c64
if AbstractFunction comparison succeed in NewContext, then the evaluator should use the same one, otherwise one of the evaluator will not be evaluated.
...
if funcgraph or metafuncgraph call it recursively, then anf_node should be used as tracking_id to discriminate the first occurcance and the
recursive occurance.
add anf_node to PrimitiveAbstractClosure hash() to reduce cost of GetEvaluatorFor().
ignore the tracking_id to make cse work.
5 years ago
8ff7c0b640
!3859 embeddinglookup support cpu in auto parallel
...
Merge pull request !3859 from yao_yf/embeddinglookup_parallel_ops_host_device
5 years ago
3ce0f33b27
add internal output tensor
5 years ago
fa86096f13
conv1x1 & deconv matmul 8x8
5 years ago
45ae76e86a
!3863 Modify logic
...
Merge pull request !3863 from lijiaqi/bug
5 years ago
f6dc928701
fix quantdtypecast
5 years ago
6d4c07c886
Update python pattern expression
5 years ago
3454153523
set tensor alloctor
5 years ago
ebda55506e
!3716 Adapt the eof of Euler OS
...
Merge pull request !3716 from jiangzhiwen/dataset/adapt_eof
5 years ago
15d0365748
fix anf exporter
5 years ago
fa96dfd161
!3774 Sync device data type to return tensor
...
Merge pull request !3774 from JoyLvliang/sync-device-data-type-to-return-tensor
5 years ago
a80432f08e
add 4 grad ops
5 years ago
1091cdeedf
!3822 Decouple infer trace from LogWriter
...
Merge pull request !3822 from hewei/decouple_debug_trace
5 years ago
d01ca09e41
Fix profiling data Flush failed in GLOG_v > 1
5 years ago
fc259aebcf
!3736 quant aware training add without bn fold and bn fold
...
Merge pull request !3736 from chenzhongming/master
5 years ago
1354cf098b
!3642 use akg to compile gpu ops rather than _akg
...
Merge pull request !3642 from lingyunli63/use_full_akg_to_support_gpu
5 years ago
8cef3ad328
!3861 [MS][LITE]fix benchmark inf error problem
...
Merge pull request !3861 from cjh9368/fix_benchmark_inf_bug
5 years ago
4edcce367d
!3864 fix fp16 init func bug && enable arm32 assembly code
...
Merge pull request !3864 from fuzhiye/mindspore
5 years ago
471ecfcaae
adapt euler eof
5 years ago
25fdb67e3d
!3744 [Auto parallel] Add a new graph operation in the DP algorithm
...
Merge pull request !3744 from Xiaoda/14-add-new-graph-operations-in-auto-parallel
5 years ago
315c6a5742
quant aware training add without bn fold and bn fold
5 years ago
878f03ad6c
Modify logic
5 years ago
5eed256f1a
fix benchmark inf error
5 years ago
b38d8ee837
use akg.ms instead of _akg
5 years ago
488d1904b6
!3802 fix nopnode internal output error
...
Merge pull request !3802 from kisnwang/fix-ps-task-error
5 years ago
7773eafd11
!3848 add anf fusion pass
...
Merge pull request !3848 from zhengjun10/master
5 years ago
ccb80694fd
!3845 add image2d for opencl depthwise op and testcase
...
Merge pull request !3845 from wandongdong/master
5 years ago
mindspore-ci-bot
liuxiao93
liuwenhao4
zhongligeng
kingfo
z00505269
chenjianping
buxue
fuzhiye
Yi Huaijie
yangruoqi713
zhaozhenlong
zhousiyi
kswang
ling
BowenK
yankai
fangzehua
caifubi
jiangzhiwen
chenzomi
李嘉琪
cjh9368
lingyunli63