Skip to content

  • 体验新版
    • 正在加载...
  • 登录
  • PaddlePaddle
  • Paddle
  • Issue
  • #10460

P
Paddle
  • 项目概览

PaddlePaddle / Paddle
大约 2 年 前同步成功

通知 2325
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
  • 动态
  • 分支图
  • 创建新Issue
  • 提交
  • Issue看板
已关闭
开放中
Opened 5月 07, 2018 by saxon_zh@saxon_zhGuest

Some issues in the `framework.py`

Created by: JiayiFeng

Last weekend I tried to implement the trainer.test() in our newest design (#10248 (closed) ). During this, I noticed that a few functions and interfaces of framework.py are implemented in quite twisty ways. Some of them only make users confused while the others may result in serious bugs. Since framework.py is the pivot of our Python API, I propose to polish framework.py first before further development on Python API.

Here are some issues I have noticed in the framework.py:

  1. Variable and Parameter are using **kwarg as it's __init__ inputs. With **kwarg, users are likely to be confused about what is exactly required to create a Variable and Parameter.

https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/python/paddle/fluid/framework.py#L152-L163

https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/python/paddle/fluid/framework.py#L1182

  1. Some code in the __init__ of Variable is used to handle the situation that the variable to be created has existed. In this part of code, we check each attribute of Variable one by one to make sure the existing variable is consistent with the one to be created, which makes the code extremely long. Is there any way to simplify this part?

https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/python/paddle/fluid/framework.py#L167-L236

  1. The function of var() is opposite in Block and Scope. If the queried variable not exists, the var() of Block will throw an exception while the one of Scope will create it.

The var() of Block: https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/python/paddle/fluid/framework.py#L717-L723

The var() of Scope: https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/paddle/fluid/framework/scope.cc#L51-L59

  1. The clone_variable() of Block does not clone all attributes of variables correctly. capacity, error_clip and stop_gradient are omitted. And why all cloned Variable's persistable are True?

https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/python/paddle/fluid/framework.py#L960-L990

  1. In the Program's clone interface. The new Program is created with the old one's C++ desc instead of the Python object itself. However, some information only exists in Python(e.g., error_clip, stop_gradient, ...). It mean a Program cannot be cloned entirely with this interface.

https://github.com/PaddlePaddle/Paddle/blob/f43b71b242467d665c134262c2b7167cef622757/python/paddle/fluid/framework.py#L1030-L1054

指派人
分配到
无
里程碑
无
分配里程碑
工时统计
无
截止日期
无
标识: paddlepaddle/Paddle#10460
渝ICP备2023009037号

京公网安备11010502055752号

网络110报警服务 Powered by GitLab CE v13.7
开源知识
Git 入门 Pro Git 电子书 在线学 Git
Markdown 基础入门 IT 技术知识开源图谱
帮助
使用手册 反馈建议 博客
《GitCode 隐私声明》 《GitCode 服务条款》 关于GitCode
Powered by GitLab CE v13.7