25b2424f9b
!1042 clean pylint warning in test_framwork dir
...
Merge pull request !1042 from jinyaohui/clean_pylint
5 years ago
6f386c5782
!1082 Support 'break', 'continue' and 'pass' statements
...
Merge pull request !1082 from hewei/support_cont_break
5 years ago
53c2a281fb
!1107 rectify auto mix precision
...
Merge pull request !1107 from gengdongjie/master
5 years ago
6a8cd83d21
!1114 fix some python code format
...
Merge pull request !1114 from casgj/gaojing
5 years ago
793bfddb17
!1112 Fix weight decay bug for `SGD` optimizer
...
Merge pull request !1112 from seatea/fix-sgd-weight-decay-bug
5 years ago
33fa90efc9
Support 'break', 'continue' and 'pass'
...
To handle 'break' and 'continue' statement, a loop context is pushed
to a stack before we parse the loop body, and pop it after body parsed.
When a 'break', 'continue' statement is encountered, we retrieve current
loop contex from the stack, and let the current block jump to the end
block or header block;
For 'break' statement, we added an extra 'end_block' follow the 'after_block',
because 'after_block' is called from a ContionalJump in 'header_block', it can
not be set as jump target from other place. to support 'break', we let loop
body jump to the 'end_block' at the 'break' point. and 'after_block'
maybe a good place to handle loop 'else' clause in the future.
Handle 'pass' is simple, just bypass it when doing parse.
5 years ago
5de4323632
!1096 fix some Misspelling
...
Merge pull request !1096 from SanjayChan/03cell
5 years ago
0345995000
!1110 [AutoParallel]fix gatherv2 and dataset bug
...
Merge pull request !1110 from lichen/fix_gatherv2_and_dataset_bug
5 years ago
26fd75895d
pylint waring clean
5 years ago
5a119d107f
Fix weight decay bug for `SGD` optimizer.
...
Weight decay has been considered and caculated in `SGD` operation, so
there's no need to apply weight decay in `SGD` optimizer.
5 years ago
abb4b7ed5c
fix some python code format
5 years ago
a2d5ad5abe
!1099 fix bug of can not find op adapter
...
Merge pull request !1099 from fary86/fix_bug_for_can_not_find_opadapter
5 years ago
168dfb2555
!1108 Add input shape condition for transpose_reshape fusion pass
...
Merge pull request !1108 from YuJianfeng/master
5 years ago
2823ec806f
!1092 add rts kernel of labelset/labelgoto/labelswitch
...
Merge pull request !1092 from dengwentao/rts_kerenl
5 years ago
18019ad6a5
!1038 clean some context param
...
Merge pull request !1038 from jinyaohui/clean_context
5 years ago
debfd38b75
fix gatherv2 and dataset bug
5 years ago
ba3d48817e
!1070 Adapt aicpu op CTCLoss and TBE op L2Loss for GE.
...
Merge pull request !1070 from liuxiao/ops-for-VM
5 years ago
eab7d593a1
!1103 clean task sink codex
...
Merge pull request !1103 from caifubi/clean-runtime-codex
5 years ago
635acb6c27
!1109 remove unused variable from repeat op
...
Merge pull request !1109 from Jamie/fixoneline
5 years ago
ff5f2521fc
remove unused variable
5 years ago
85ff90c237
Add input shape condition for transpose_reshape fusion pass
5 years ago
06bac6f120
add rts label op
5 years ago
391a060f21
remove two context param
5 years ago
031a2931b0
add new auto mix precision flag
5 years ago
298a784878
!1090 fix mean and var shape in layernorm
...
Merge pull request !1090 from JichenZhao/layernorm_mean_var_shape
5 years ago
9a0d894b9d
clean task sink codex
5 years ago
77ef790f25
Fix bug of can not find opadapter
5 years ago
7ab3f5c348
!1086 fix some bug in quant debug
...
Merge pull request !1086 from SanjayChan/05showcase
5 years ago
48e54dcfda
!1056 [Auto parallel] Memory calculation in the inference phase
...
Merge pull request !1056 from Xiaoda/memory-estimation-in-inference-phase
5 years ago
fd2af88871
fix some Misspelling
5 years ago
00b7877ec4
!1087 add param_perfix to cell
...
Merge pull request !1087 from SanjayChan/03cell
5 years ago
6710e1feac
add param_perfix to cell
5 years ago
c1290e7b70
!1088 Add some op registrations for const value to attr
...
Merge pull request !1088 from YuJianfeng/master
5 years ago
10abb68498
add ops CTCLoss
5 years ago
fdad91355f
!1093 reset parallel context before running each parallel test case
...
Merge pull request !1093 from yihuaijie/master
5 years ago
da7d605e85
!683 WIP: specialize hyper map parameter
...
Merge pull request !683 from xychow/bypass-renorm-and-specialize-hypermap-parameter
5 years ago
6ab046e4dd
!1062 Adapt Softplus\SoftplusGrad for VM, and ApplyFtrlD for both GE and VM.
...
Merge pull request !1062 from liuxiao/softplus-softplusgrad
5 years ago
a05aa21cc2
calculating PEAK memory cost in the inference phase
5 years ago
552fc5c958
!1085 Adapt removing TopK and use TopKD
...
Merge pull request !1085 from zhaozhenlong/vm-adapt-topk
5 years ago
06a9eeb3bf
!1050 [bug][auto_mixed_precision]fix sens shape error of `TrainOneStepWithLossScaleCell`
...
Merge pull request !1050 from vlne-v1/I1GEPN-check-bprop-error-loss-scale-wrap
5 years ago
c622879a2f
fix mean and var shape in layernorm
5 years ago
462f30e077
reset auto parallel context before each test
5 years ago
eb6094315a
!1083 change the hswish to HSwish
...
Merge pull request !1083 from SanjayChan/02bug_fix
5 years ago
b9d3495c4d
!1081 format nn. and change quant.py
...
Merge pull request !1081 from SanjayChan/formast
5 years ago
98112d1a64
!997 support vm for log1p
...
Merge pull request !997 from jiangjinsheng/log1p
5 years ago
8701ca24dd
!950 add bert base version finetune and evaluation script in bert_clue dir
...
Merge pull request !950 from yoonlee666/master
5 years ago
3e89f7baaa
fix sens shape error in loss_scale one_step_wrap
5 years ago
2a695cfe24
fix some bug in quant debug
5 years ago
28a6daaed5
!1017 add profiling_mode and profiling_options to context
...
Merge pull request !1017 from jinyaohui/context_opt
5 years ago
b0b867a4bc
!1077 fix bug when reset taskinfo error
...
Merge pull request !1077 from jjfeing/master
5 years ago
mindspore-ci-bot
He Wei
jinyaohui
seatea
gaojing
lichenever
Jamie Nisbet
yujianfeng
dengwentao
gengdongjie
caifubi
fary86
chenzomi
liuxiao
Xiaoda Zhang
zhaojichen
Yi Huaijie
Wei Luning