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
wmsofts
Paddle
提交
783dbe9a
P
Paddle
项目概览
wmsofts
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
0
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
体验新版 GitCode,发现更多精彩内容 >>
提交
783dbe9a
编写于
1月 15, 2019
作者:
X
Xin Pan
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
more doc
test=develop
上级
f997109b
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
45 addition
and
6 deletion
+45
-6
paddle/fluid/imperative/README.md
paddle/fluid/imperative/README.md
+45
-6
未找到文件。
paddle/fluid/imperative/README.md
浏览文件 @
783dbe9a
# Overview
Imperative Programming
Imperative Programming
is easier to learn, debug and try new ideas.
# Related Works
...
...
@@ -37,12 +37,38 @@ class PyLayer(core.PyLayer):
@
staticmethod
def
backward
(
inputs
):
# any backward logic implemented with numpy io.
```
## Tracer
Python Variable -> C++ VarBase -> C++ Variable -> C++ Tensor
Current: Python Variable -> C++ VarBase -> C++ Variable -> C++ Tensor
Longer term.
```
python
# Parent class.
class
PyVarBase
(
object
):
pass
# Current python variable.
class
Variable
(
PyVarBase
):
pass
class
IVariable
(
PyVarBase
):
def
__init__
(
self
):
self
.
_ivar
=
core
.
VarBase
()
def
to
(
device
):
pass
def
value
():
pass
def
backward
():
pass
def
gradient_value
():
pass
# operators to override.
```
```
cpp
...
...
@@ -62,10 +88,21 @@ class Tracer {
};
```
*
Trace forward operations
*
Perform simple python level infer and return to user.
*
Perform autograd to generate gradients.
*
Clear trace.
*
Apply gradients with optimizers
## Autodiff
Lots of research already.
https://autodiff-workshop.github.io/
https://en.wikipedia.org/wiki/Automatic_differentiation
## Execution Engine
Lazy execution of pushed C++ operations.
## Tests
...
...
@@ -76,7 +113,6 @@ https://autodiff-workshop.github.io/
*
All function layers with parameters converted to class Layers.
*
Models converted to imperative mode.
# Examples
```
python
...
...
@@ -131,6 +167,10 @@ class MLP(fluid.imperative.Layer):
```
## Save/Load Models
TODO
# Plan
2.
1,3 fulltime, Can run a few simple models. (Currently, 2 20% engs)
...
...
@@ -139,10 +179,9 @@ class MLP(fluid.imperative.Layer):
6.
1, 5 fulltime, Performance close to Pytorch, can run multi-devices. Release Beta.
8.
1, 5 fulltime, Works in general. Covert current models to use imperative mode.
12.
1, 5 fulltime, Can compile to static graph, support more optimizations.
8.
1, 5 fulltime, Works in general. Update existing models. Can compile to static graph, support more optimizations.
12.
1 Done.
# Discussion
...
...
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