Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
OPTHREE
Paddle
提交
ddf57836
P
Paddle
项目概览
OPTHREE
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
Paddle
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
1
Issue
1
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
未验证
提交
ddf57836
编写于
4月 12, 2018
作者:
J
jiaozhenyu
提交者:
GitHub
4月 12, 2018
浏览文件
操作
浏览文件
下载
差异文件
Merge pull request #9812 from jshower/develop
Changing network configuration, avoid nan
上级
4c55a602
ad6ddf53
变更
1
显示空白变更内容
内联
并排
Showing
1 changed file
with
9 addition
and
8 deletion
+9
-8
python/paddle/fluid/tests/book/test_label_semantic_roles.py
python/paddle/fluid/tests/book/test_label_semantic_roles.py
+9
-8
未找到文件。
python/paddle/fluid/tests/book/test_label_semantic_roles.py
浏览文件 @
ddf57836
...
...
@@ -37,7 +37,7 @@ depth = 8
mix_hidden_lr
=
1e-3
IS_SPARSE
=
True
PASS_NUM
=
10
PASS_NUM
=
10
0
BATCH_SIZE
=
10
embedding_name
=
'emb'
...
...
@@ -77,7 +77,8 @@ def db_lstm(word, predicate, ctx_n2, ctx_n1, ctx_0, ctx_p1, ctx_p2, mark,
emb_layers
.
append
(
mark_embedding
)
hidden_0_layers
=
[
fluid
.
layers
.
fc
(
input
=
emb
,
size
=
hidden_dim
)
for
emb
in
emb_layers
fluid
.
layers
.
fc
(
input
=
emb
,
size
=
hidden_dim
,
act
=
'tanh'
)
for
emb
in
emb_layers
]
hidden_0
=
fluid
.
layers
.
sums
(
input
=
hidden_0_layers
)
...
...
@@ -94,8 +95,8 @@ def db_lstm(word, predicate, ctx_n2, ctx_n1, ctx_0, ctx_p1, ctx_p2, mark,
for
i
in
range
(
1
,
depth
):
mix_hidden
=
fluid
.
layers
.
sums
(
input
=
[
fluid
.
layers
.
fc
(
input
=
input_tmp
[
0
],
size
=
hidden_dim
),
fluid
.
layers
.
fc
(
input
=
input_tmp
[
1
],
size
=
hidden_dim
)
fluid
.
layers
.
fc
(
input
=
input_tmp
[
0
],
size
=
hidden_dim
,
act
=
'tanh'
),
fluid
.
layers
.
fc
(
input
=
input_tmp
[
1
],
size
=
hidden_dim
,
act
=
'tanh'
)
])
lstm
=
fluid
.
layers
.
dynamic_lstm
(
...
...
@@ -109,8 +110,8 @@ def db_lstm(word, predicate, ctx_n2, ctx_n1, ctx_0, ctx_p1, ctx_p2, mark,
input_tmp
=
[
mix_hidden
,
lstm
]
feature_out
=
fluid
.
layers
.
sums
(
input
=
[
fluid
.
layers
.
fc
(
input
=
input_tmp
[
0
],
size
=
label_dict_len
),
fluid
.
layers
.
fc
(
input
=
input_tmp
[
1
],
size
=
label_dict_len
)
fluid
.
layers
.
fc
(
input
=
input_tmp
[
0
],
size
=
label_dict_len
,
act
=
'tanh'
),
fluid
.
layers
.
fc
(
input
=
input_tmp
[
1
],
size
=
label_dict_len
,
act
=
'tanh'
)
])
return
feature_out
...
...
@@ -171,7 +172,7 @@ def train(use_cuda, save_dirname=None, is_local=True):
# check other optimizers and check why out will be NAN
sgd_optimizer
=
fluid
.
optimizer
.
SGD
(
learning_rate
=
fluid
.
layers
.
exponential_decay
(
learning_rate
=
0.0
00
1
,
learning_rate
=
0.01
,
decay_steps
=
100000
,
decay_rate
=
0.5
,
staircase
=
True
))
...
...
@@ -233,7 +234,7 @@ def train(use_cuda, save_dirname=None, is_local=True):
print
(
"second per batch: "
+
str
((
time
.
time
(
)
-
start_time
)
/
batch_id
))
# Set the threshold low to speed up the CI test
if
float
(
pass_precision
)
>
0.0
5
:
if
float
(
pass_precision
)
>
0.0
1
:
if
save_dirname
is
not
None
:
# TODO(liuyiqun): Change the target to crf_decode
fluid
.
io
.
save_inference_model
(
save_dirname
,
[
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录