5d410342e9
!13566 fix transformer readme
...
From: @yuchaojie
Reviewed-by: @oacjiewen,@c_34
Signed-off-by: @c_34
4 years ago
483ab836c9
fix transformer readme
4 years ago
15d37e5db9
imporve convergence of loss in bert
4 years ago
fa1fbc088c
!13163 modify api detect_overflow name in TrainOneStepWithLossScaleCell
...
From: @wangnan39
Reviewed-by: @kingxian,@hwhewei
Signed-off-by: @kingxian
4 years ago
7188a14215
modify api detect_overflow name in TrainOneStepWithLossScaleCell
4 years ago
4d82df0b09
!12652 modify tinybert readme
...
From: @wang_hua_2019
Reviewed-by: @c_34,@linqingke
Signed-off-by: @c_34
4 years ago
8133238f2f
modify tinybert readme
4 years ago
ec9793861f
fix grad accu
4 years ago
95adf66d30
!12512 optimize class TrainOneStepWithLossScaleCell
...
From: @wangnan39
Reviewed-by:
Signed-off-by:
4 years ago
4d69e21933
optimizer class TrainOneStepWithLossScaleCell
4 years ago
923b53d2e6
replace tab with spaces
4 years ago
f4193137e5
LSTM Ascend parameter type fp16 change to fp32
4 years ago
7d9a783993
[auto-monad] Support side-effects by auto-monad
...
The basic idea is: exploits data dependency to control the execution order
of side-effect operations, and keep the semantics of ANF unchanged.
The ControlDepend primitive is removed and there are two primitives added:
1. UpdateState:
```
a = Assign(para, value)
```
became:
```
a = Assign(para, value, u)
u = UpdateState(u, a)
```
2. Load:
```
x = Add(para, value)
```
became:
```
p = Load(para, u)
x = Add(p, value)
u = UpdateState(u, p)
```
4 years ago
74652eb942
!12044 modify pack to stack
...
From: @jinyaohui
Reviewed-by:
Signed-off-by:
4 years ago
3bf5681ebc
!12093 adapt api for optimizer import
...
From: @sl_wang
Reviewed-by: @wang_zi_dong
Signed-off-by:
4 years ago
8022f9a6ed
modify pack to stack
4 years ago
8f8eee4b5e
bert thor supports lr configuration in config.py
4 years ago
e36d08aacc
!12096 fix link error in README of GPT
...
From: @alouhahahahaha
Reviewed-by: @kisnwang
Signed-off-by:
4 years ago
b2fc857fe0
fix link error in README
4 years ago
51e7b6481a
!12034 modify instructions of preprocessing msra dataset
...
From: @shibeiji
Reviewed-by: @c_34,@guoqi1024
Signed-off-by: @c_34
4 years ago
1c12b84757
!11992 [Bert][Gpu]Sync modify of bert script from r1.1 to master
...
From: @hanhuifeng2020
Reviewed-by: @gaoxiong1,@anyrenwei
Signed-off-by: @anyrenwei
4 years ago
95ccebb811
fix bugs of MSRA dataset preprocessing
4 years ago
ed4207f698
!12022 Add a new optimizer THOR option to BERT pretrain script.
...
From: @i_yangmh
Reviewed-by: @wang_zi_dong,@kisnwang
Signed-off-by: @kisnwang
4 years ago
0b312c2385
thor generalization code submit
4 years ago
67a4c62b4b
Add new optimizer THOR option to BERT pretrain script.
4 years ago
53d4510ea6
[Bert][Gpu]Sync modify of bert script from r1.1 to master
4 years ago
65f1d90509
!11946 gnmt modified
...
From: @gaojing22
Reviewed-by: @c_34,@wuxuejian
Signed-off-by: @c_34
4 years ago
e897eb4c41
!11915 Change TensorAdd to Add, merge from r1.1 to master
...
From: @liangzhibo
Reviewed-by: @ginfung,@zh_qh
Signed-off-by: @zh_qh
4 years ago
df265b6d6b
!11912 update submoudle akg, close graph kernel ascend ci testcases
...
From: @looop5
Reviewed-by:
Signed-off-by:
4 years ago
930fdc4e31
modify uncorrect name
4 years ago
9ec100d069
Change TensorAdd to Add, from r1.1 to master
4 years ago
0161209e40
update submoudle akg, close graph kernel ascend ci testcases
4 years ago
cc18b206c9
bert ner for adaption of MSRA dataset
4 years ago
9fa0499fa0
Change GatherV2 to Gather r1.1 to master
4 years ago
8b1a8a6bc1
bert ner for msra dataset
4 years ago
3ec3f038ad
extract bert embedding tables in construct
4 years ago
01d84b1ce5
!11680 fix GRU dataset
...
From: @qujianwei
Reviewed-by: @c_34,@liangchenghui
Signed-off-by: @c_34
4 years ago
4fef53a4a2
!11592 add chn/eng switch for transformer readme
...
From: @yuchaojie
Reviewed-by: @c_34,@liangchenghui
Signed-off-by: @c_34
4 years ago
19fc6306d3
fix for dataset input
4 years ago
74c2b957f7
!11632 fix gru net usablity bug
...
From: @qujianwei
Reviewed-by: @linqingke,@liangchenghui
Signed-off-by: @liangchenghui
4 years ago
6adb82c2c6
gru_fix_bug
4 years ago
dc70469d4a
!11597 Fix code spell in Bert and CRNN
...
From: @c_34
Reviewed-by: @liangchenghui
Signed-off-by:
4 years ago
4899a0b4c1
fix code spell
4 years ago
d2f818968c
add chn/eng switch for transformer readme
4 years ago
14a6713d08
!11541 modify transformer export
...
From: @changzherui
Reviewed-by: @guoqi1024,@kingxian
Signed-off-by: @guoqi1024
4 years ago
0beb38fdb0
!11491 Add GRU network
...
From: @qujianwei
Reviewed-by:
Signed-off-by:
4 years ago
a37ad24136
init for gru
...
fix for copyright
fix for create_dataset
fix for some bug
fix for parser_output
fix for parse output
add readme and fix for some bug
fix test2016 to test
delete space
fix some bug for create dataset
fix for comments and overflow
fix a wrong bug
fix for weight init
4 years ago
a0ff6613c3
modify transformer export
4 years ago
303048a186
Add lstm ascend distribute train
4 years ago
8eae3262f3
fix some description about bert and yolov3
4 years ago
mindspore-ci-bot
yuchaojie
chenhaozhe
wangnan39@huawei.com
wang_hua_2019
Ziyan
xsmq
ttudu
He Wei
jinyaohui
mwang
alouhahaha
shibeiji
MingHan-Y
hanhuifeng2020
gaojing
l00591931
looop5
qujianwei
changzherui