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
BaiXuePrincess
Paddle
提交
793c35ef
P
Paddle
项目概览
BaiXuePrincess
/
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
0
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
0
Issue
0
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
未验证
提交
793c35ef
编写于
11月 08, 2022
作者:
S
shentanyue
提交者:
GitHub
11月 08, 2022
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix npu:0 stage (#47729)
上级
caca5687
变更
1
显示空白变更内容
内联
并排
Showing
1 changed file
with
36 addition
and
26 deletion
+36
-26
python/paddle/device/__init__.py
python/paddle/device/__init__.py
+36
-26
未找到文件。
python/paddle/device/__init__.py
浏览文件 @
793c35ef
...
...
@@ -249,28 +249,6 @@ def _convert_to_place(device):
avaliable_xpu_device
=
re
.
match
(
r
'xpu:\d+'
,
lower_device
)
avaliable_npu_device
=
re
.
match
(
r
'npu:\d+'
,
lower_device
)
avaliable_mlu_device
=
re
.
match
(
r
'mlu:\d+'
,
lower_device
)
if
(
not
avaliable_gpu_device
and
not
avaliable_xpu_device
and
not
avaliable_npu_device
and
not
avaliable_mlu_device
):
device_info_list
=
device
.
split
(
':'
,
1
)
device_type
=
device_info_list
[
0
]
if
device_type
in
core
.
get_all_custom_device_type
():
device_id
=
device_info_list
[
1
]
device_id
=
int
(
device_id
)
place
=
core
.
CustomPlace
(
device_type
,
device_id
)
else
:
raise
ValueError
(
"The device must be a string which is like 'cpu', {}"
.
format
(
', '
.
join
(
"'{}', '{}:x'"
.
format
(
x
,
x
)
for
x
in
[
'gpu'
,
'xpu'
,
'npu'
,
'mlu'
]
+
core
.
get_all_custom_device_type
()
)
)
)
if
avaliable_gpu_device
:
if
not
core
.
is_compiled_with_cuda
():
raise
ValueError
(
...
...
@@ -293,9 +271,19 @@ def _convert_to_place(device):
place
=
core
.
XPUPlace
(
device_id
)
if
avaliable_npu_device
:
if
not
core
.
is_compiled_with_npu
():
device_info_list
=
device
.
split
(
':'
,
1
)
device_type
=
device_info_list
[
0
]
if
device_type
in
core
.
get_all_custom_device_type
():
device_id
=
device_info_list
[
1
]
device_id
=
int
(
device_id
)
place
=
core
.
CustomPlace
(
device_type
,
device_id
)
return
place
else
:
raise
ValueError
(
"The device should not be {}, since PaddlePaddle is "
"not compiled with NPU"
.
format
(
avaliable_npu_device
)
"not compiled with NPU or compiled with custom device"
.
format
(
avaliable_npu_device
)
)
device_info_list
=
device
.
split
(
':'
,
1
)
device_id
=
device_info_list
[
1
]
...
...
@@ -311,6 +299,28 @@ def _convert_to_place(device):
device_id
=
device_info_list
[
1
]
device_id
=
int
(
device_id
)
place
=
core
.
MLUPlace
(
device_id
)
if
(
not
avaliable_gpu_device
and
not
avaliable_xpu_device
and
not
avaliable_npu_device
and
not
avaliable_mlu_device
):
device_info_list
=
device
.
split
(
':'
,
1
)
device_type
=
device_info_list
[
0
]
if
device_type
in
core
.
get_all_custom_device_type
():
device_id
=
device_info_list
[
1
]
device_id
=
int
(
device_id
)
place
=
core
.
CustomPlace
(
device_type
,
device_id
)
else
:
raise
ValueError
(
"The device must be a string which is like 'cpu', {}"
.
format
(
', '
.
join
(
"'{}', '{}:x'"
.
format
(
x
,
x
)
for
x
in
[
'gpu'
,
'xpu'
,
'npu'
,
'mlu'
]
+
core
.
get_all_custom_device_type
()
)
)
)
return
place
...
...
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