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
MegEngine 天元
MegEngine
提交
171c683d
MegEngine
项目概览
MegEngine 天元
/
MegEngine
1 年多 前同步成功
通知
410
Star
4707
Fork
583
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
DevOps
流水线
流水线任务
计划
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
MegEngine
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
DevOps
DevOps
流水线
流水线任务
计划
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
流水线任务
提交
Issue看板
提交
171c683d
编写于
6月 20, 2022
作者:
M
Megvii Engine Team
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
docs(api/lite): add struct python api desc
GitOrigin-RevId: 82f873b498497eafbf40eed6aa78ef036650cd89
上级
3cd54dd6
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
48 addition
and
18 deletion
+48
-18
lite/pylite/megenginelite/struct.py
lite/pylite/megenginelite/struct.py
+48
-18
未找到文件。
lite/pylite/megenginelite/struct.py
浏览文件 @
171c683d
...
...
@@ -6,10 +6,21 @@ from enum import Enum, IntEnum
class
LiteBackend
(
IntEnum
):
"""
The backend type enum, default only
"""
LITE_DEFAULT
=
0
class
LiteDeviceType
(
IntEnum
):
"""
The backend device type enum
Note:
compute and storage will base on the device
"""
LITE_CPU
=
0
LITE_CUDA
=
1
LITE_ATLAS
=
3
...
...
@@ -19,6 +30,13 @@ class LiteDeviceType(IntEnum):
class
LiteDataType
(
IntEnum
):
"""
The tensor data type enum
Note:
half for float16, int for int32
"""
LITE_FLOAT
=
0
LITE_HALF
=
1
LITE_INT
=
2
...
...
@@ -29,6 +47,12 @@ class LiteDataType(IntEnum):
class
LiteTensorPhase
(
IntEnum
):
"""
The tensor type enum
Note:
LITE_IO for both LITE_INPUT and LITE_OUTPUT
"""
LITE_IO
=
0
LITE_INPUT
=
1
LITE_OUTPUT
=
2
...
...
@@ -36,7 +60,7 @@ class LiteTensorPhase(IntEnum):
class
LiteIOType
(
IntEnum
):
"""
the input and output type
, include SHAPE and VALUE
The input and output type enum
, include SHAPE and VALUE
sometimes user only need the shape of the output tensor
"""
...
...
@@ -46,26 +70,26 @@ class LiteIOType(IntEnum):
class
LiteAlgoSelectStrategy
(
IntEnum
):
"""
operation algorithm seletion strategy type
, some operations have
Operation algorithm seletion strategy type enum
, some operations have
multi algorithms, different algorithm has different attribute, according to
the strategy, the best algorithm will be selected.
Note:
These strategies can be combined
LITE_ALGO_HEURISTIC | LITE_ALGO_PROFILE means: if profile cache not valid,
use heuristic instead
Note:
These strategies can be combined
LITE_ALGO_HEURISTIC | LITE_ALGO_PROFILE means: if profile cache not valid,
use heuristic instead
LITE_ALGO_HEURISTIC | LITE_ALGO_REPRODUCIBLE means: heuristic choice the
reproducible algo
LITE_ALGO_HEURISTIC | LITE_ALGO_REPRODUCIBLE means: heuristic choice the
reproducible algo
LITE_ALGO_PROFILE | LITE_ALGO_REPRODUCIBLE means: profile the best
algorithm from the reproducible algorithms set
LITE_ALGO_PROFILE | LITE_ALGO_REPRODUCIBLE means: profile the best
algorithm from the reproducible algorithms set
LITE_ALGO_PROFILE | LITE_ALGO_OPTIMIZED means: profile the best
algorithm form the optimzed algorithms, thus profile will process fast
LITE_ALGO_PROFILE | LITE_ALGO_OPTIMIZED means: profile the best
algorithm form the optimzed algorithms, thus profile will process fast
LITE_ALGO_PROFILE | LITE_ALGO_OPTIMIZED | LITE_ALGO_REPRODUCIBLE means:
profile the best algorithm form the optimzed and reproducible algorithms
LITE_ALGO_PROFILE | LITE_ALGO_OPTIMIZED | LITE_ALGO_REPRODUCIBLE means:
profile the best algorithm form the optimzed and reproducible algorithms
"""
LITE_ALGO_HEURISTIC
=
1
...
...
@@ -76,10 +100,16 @@ class LiteAlgoSelectStrategy(IntEnum):
class
LiteLogLevel
(
IntEnum
):
"""
DEBUG: The most verbose level, printing debugging info
INFO: The default level
WARN: Printing warnings
ERROR: The least verbose level, printing errors only
Log level enum
Note:
DEBUG: The most verbose level, printing debugging info
INFO: The default level
WARN: Printing warnings
ERROR: The least verbose level, printing errors only
"""
DEBUG
=
0
...
...
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