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
PaddlePaddle
Paddle
提交
d1e1d858
P
Paddle
项目概览
PaddlePaddle
/
Paddle
接近 2 年 前同步成功
通知
2323
Star
20933
Fork
5424
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1423
列表
看板
标记
里程碑
合并请求
543
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
1,423
Issue
1,423
列表
看板
标记
里程碑
合并请求
543
合并请求
543
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
未验证
提交
d1e1d858
编写于
4月 30, 2020
作者:
W
wawltor
提交者:
GitHub
4月 30, 2020
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
add the graph batch reader for pslib mode (#24178)
Add the pslib graph batch reader mode, add the test case for this change
上级
80355949
变更
5
隐藏空白更改
内联
并排
Showing
5 changed file
with
86 addition
and
2 deletion
+86
-2
paddle/fluid/framework/data_feed.cc
paddle/fluid/framework/data_feed.cc
+22
-2
paddle/fluid/framework/data_feed.h
paddle/fluid/framework/data_feed.h
+3
-0
paddle/fluid/framework/data_feed.proto
paddle/fluid/framework/data_feed.proto
+1
-0
python/paddle/fluid/dataset.py
python/paddle/fluid/dataset.py
+3
-0
python/paddle/fluid/tests/unittests/test_dataset.py
python/paddle/fluid/tests/unittests/test_dataset.py
+57
-0
未找到文件。
paddle/fluid/framework/data_feed.cc
浏览文件 @
d1e1d858
...
@@ -813,6 +813,7 @@ void MultiSlotInMemoryDataFeed::Init(
...
@@ -813,6 +813,7 @@ void MultiSlotInMemoryDataFeed::Init(
visit_
.
resize
(
all_slot_num
,
false
);
visit_
.
resize
(
all_slot_num
,
false
);
pipe_command_
=
data_feed_desc
.
pipe_command
();
pipe_command_
=
data_feed_desc
.
pipe_command
();
finish_init_
=
true
;
finish_init_
=
true
;
input_type_
=
data_feed_desc
.
input_type
();
}
}
void
MultiSlotInMemoryDataFeed
::
GetMsgFromLogKey
(
const
std
::
string
&
log_key
,
void
MultiSlotInMemoryDataFeed
::
GetMsgFromLogKey
(
const
std
::
string
&
log_key
,
...
@@ -1065,8 +1066,27 @@ void MultiSlotInMemoryDataFeed::PutToFeedVec(
...
@@ -1065,8 +1066,27 @@ void MultiSlotInMemoryDataFeed::PutToFeedVec(
CopyToFeedTensor
(
tensor_ptr
,
feasign
,
total_instance
*
sizeof
(
int64_t
));
CopyToFeedTensor
(
tensor_ptr
,
feasign
,
total_instance
*
sizeof
(
int64_t
));
}
}
auto
&
slot_offset
=
offset_
[
i
];
auto
&
slot_offset
=
offset_
[
i
];
LoD
data_lod
{
slot_offset
};
if
(
this
->
input_type_
==
0
)
{
feed_vec_
[
i
]
->
set_lod
(
data_lod
);
LoD
data_lod
{
slot_offset
};
feed_vec_
[
i
]
->
set_lod
(
data_lod
);
}
else
if
(
this
->
input_type_
==
1
)
{
if
(
!
use_slots_is_dense_
[
i
])
{
std
::
vector
<
size_t
>
tmp_offset
;
PADDLE_ENFORCE_EQ
(
slot_offset
.
size
(),
2
,
platform
::
errors
::
InvalidArgument
(
"In batch reader, the sparse tensor lod size "
"must be 2, but received %d"
,
slot_offset
.
size
()));
const
auto
&
max_size
=
slot_offset
[
1
];
tmp_offset
.
reserve
(
max_size
+
1
);
for
(
unsigned
int
k
=
0
;
k
<=
max_size
;
k
++
)
{
tmp_offset
.
emplace_back
(
k
);
}
slot_offset
=
tmp_offset
;
LoD
data_lod
{
slot_offset
};
feed_vec_
[
i
]
->
set_lod
(
data_lod
);
}
}
if
(
use_slots_is_dense_
[
i
])
{
if
(
use_slots_is_dense_
[
i
])
{
if
(
inductive_shape_index_
[
i
]
!=
-
1
)
{
if
(
inductive_shape_index_
[
i
]
!=
-
1
)
{
use_slots_shape_
[
i
][
inductive_shape_index_
[
i
]]
=
use_slots_shape_
[
i
][
inductive_shape_index_
[
i
]]
=
...
...
paddle/fluid/framework/data_feed.h
浏览文件 @
d1e1d858
...
@@ -232,6 +232,9 @@ class DataFeed {
...
@@ -232,6 +232,9 @@ class DataFeed {
std
::
vector
<
std
::
string
>
ins_id_vec_
;
std
::
vector
<
std
::
string
>
ins_id_vec_
;
std
::
vector
<
std
::
string
>
ins_content_vec_
;
std
::
vector
<
std
::
string
>
ins_content_vec_
;
platform
::
Place
place_
;
platform
::
Place
place_
;
// The input type of pipe reader, 0 for one sample, 1 for one batch
int
input_type_
;
};
};
// PrivateQueueDataFeed is the base virtual class for ohther DataFeeds.
// PrivateQueueDataFeed is the base virtual class for ohther DataFeeds.
...
...
paddle/fluid/framework/data_feed.proto
浏览文件 @
d1e1d858
...
@@ -32,4 +32,5 @@ message DataFeedDesc {
...
@@ -32,4 +32,5 @@ message DataFeedDesc {
optional
int32
thread_num
=
5
;
optional
int32
thread_num
=
5
;
optional
string
rank_offset
=
6
;
optional
string
rank_offset
=
6
;
optional
int32
pv_batch_size
=
7
[
default
=
32
];
optional
int32
pv_batch_size
=
7
[
default
=
32
];
optional
int32
input_type
=
8
[
default
=
0
];
}
}
python/paddle/fluid/dataset.py
浏览文件 @
d1e1d858
...
@@ -221,6 +221,9 @@ class DatasetBase(object):
...
@@ -221,6 +221,9 @@ class DatasetBase(object):
self
.
dataset
.
set_filelist
(
filelist
)
self
.
dataset
.
set_filelist
(
filelist
)
self
.
filelist
=
filelist
self
.
filelist
=
filelist
def
set_input_type
(
self
,
input_type
):
self
.
proto_desc
.
input_type
=
input_type
def
set_use_var
(
self
,
var_list
):
def
set_use_var
(
self
,
var_list
):
"""
"""
Set Variables which you will use.
Set Variables which you will use.
...
...
python/paddle/fluid/tests/unittests/test_dataset.py
浏览文件 @
d1e1d858
...
@@ -601,6 +601,63 @@ class TestDataset(unittest.TestCase):
...
@@ -601,6 +601,63 @@ class TestDataset(unittest.TestCase):
os
.
remove
(
"./test_queue_dataset_run_a.txt"
)
os
.
remove
(
"./test_queue_dataset_run_a.txt"
)
os
.
remove
(
"./test_queue_dataset_run_b.txt"
)
os
.
remove
(
"./test_queue_dataset_run_b.txt"
)
def
test_queue_dataset_run_3
(
self
):
"""
Testcase for QueueDataset from create to run.
Use CUDAPlace
Use float type id
"""
with
open
(
"test_queue_dataset_run_a.txt"
,
"w"
)
as
f
:
data
=
"2 1 2 2 5 4 2 2 7 2 1 3
\n
"
data
+=
"2 6 2 2 1 4 2 2 4 2 2 3
\n
"
data
+=
"2 5 2 2 9 9 2 2 7 2 1 3
\n
"
data
+=
"2 7 2 2 1 9 2 3 7 2 5 3
\n
"
f
.
write
(
data
)
with
open
(
"test_queue_dataset_run_b.txt"
,
"w"
)
as
f
:
data
=
"2 1 2 2 5 4 2 2 7 2 1 3
\n
"
data
+=
"2 6 2 2 1 4 2 2 4 2 2 3
\n
"
data
+=
"2 5 2 2 9 9 2 2 7 2 1 3
\n
"
data
+=
"2 7 2 2 1 9 2 3 7 2 5 3
\n
"
f
.
write
(
data
)
slots
=
[
"slot1"
,
"slot2"
,
"slot3"
,
"slot4"
]
slots_vars
=
[]
for
slot
in
slots
:
var
=
fluid
.
data
(
name
=
slot
,
shape
=
[
None
,
1
],
dtype
=
"int64"
,
lod_level
=
1
)
slots_vars
.
append
(
var
)
dataset
=
fluid
.
DatasetFactory
().
create_dataset
(
"InMemoryDataset"
)
dataset
.
set_input_type
(
1
)
dataset
.
set_batch_size
(
1
)
dataset
.
set_thread
(
2
)
dataset
.
set_filelist
(
[
"test_queue_dataset_run_a.txt"
,
"test_queue_dataset_run_b.txt"
])
dataset
.
set_pipe_command
(
"cat"
)
dataset
.
set_use_var
(
slots_vars
)
dataset
.
load_into_memory
()
exe
=
fluid
.
Executor
(
fluid
.
CPUPlace
()
if
not
core
.
is_compiled_with_cuda
(
)
else
fluid
.
CUDAPlace
(
0
))
exe
.
run
(
fluid
.
default_startup_program
())
if
self
.
use_data_loader
:
data_loader
=
fluid
.
io
.
DataLoader
.
from_dataset
(
dataset
,
fluid
.
cpu_places
(),
self
.
drop_last
)
for
i
in
range
(
self
.
epoch_num
):
for
data
in
data_loader
():
exe
.
run
(
fluid
.
default_main_program
(),
feed
=
data
)
else
:
for
i
in
range
(
self
.
epoch_num
):
try
:
exe
.
train_from_dataset
(
fluid
.
default_main_program
(),
dataset
)
except
Exception
as
e
:
self
.
assertTrue
(
False
)
os
.
remove
(
"./test_queue_dataset_run_a.txt"
)
os
.
remove
(
"./test_queue_dataset_run_b.txt"
)
class
TestDatasetWithDataLoader
(
TestDataset
):
class
TestDatasetWithDataLoader
(
TestDataset
):
"""
"""
...
...
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