Code Monkey home page Code Monkey logo

yousan.ai's Introduction

yansan.ai

What is yousan.ai

It is a project which provides a lot of resources you may need in your 'deeplearning' study,including projects, ebooks books and courses.

Supported framework

  • caffe
  • tensorflow
  • pytorch
  • mxnet
  • paddlepaddle,
  • darknet
  • deeplearning4j
  • matconvnet
  • keras
  • chainer
  • cntk
  • lasadge
  • and so on.

Task

  • Computer vision: image classification,image segmentation,object detection and so on.
  • Speech processing: speech recognition,speech synthesis and so on.
  • Natural Language Processing: knowledge graph,automatic question answering and so on.
  • most importantly, you can always contact me from email or github issues for any problems you meet.

books

  • 深度学习之图像识别.
  • 深度学习之模型设计.
  • 深度学习之人脸图像.
  • 深度学习之摄影图像.
  • 生成对抗网络GAN.
  • 深度学习之图像识别(全彩版).

ebooks

  • 有三AI视觉算法工程师成长指导手册.
  • 有三AI深度学习开源框架实践指导手册.

course

  • 深度学习之数据使用.
  • 深度学习之图像分类.
  • 深度学习之图像分割.
  • 深度学习之目标检测.
  • 深度学习之图像生成GAN.
  • 深度学习之图像翻译GAN.
  • 深度学习之图像增强GAN.
  • 深度学习之模型分析.
  • 深度学习之模型设计.
  • 深度学习之模型优化.
  • 深度学习之模型部署.
  • 深度学习之视频分类.
  • 深度学习之人脸检测与识别.
  • 深度学习之人脸属性编辑.

Problems

yousan.ai's People

Contributors

longpeng2008 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

yousan.ai's Issues

如何下载

你好,请问这些论文如何下载呢?只能全部download吗?

Exit

Background

Back in #3943, we started rendering text editor contents inside a shadow DOM boundary to prevent outer style sheets from altering the editor inner contents. Although this introduced many benefits, it carried several tradeoffs too.

In particular, we started duplicating some of the style sheets ({context: 'atom-text-editor'}) inside each shadow root so that package and theme authors could still be able to explicitly target the editor visual elements. An alternative solution to this problem that we suggested (before the shadow DOM spec was finalized) was to use /deep/ and ::shadow pseudo selectors, which unfortunately are now being deprecated.

Even more importantly, this architecture was getting in the way of making atom-text-editor a reusable component, because it made it even more coupled to Atom and its ecosystem. Other problems involved things like #4590, and a more complex user experience for package and theme authors in general. Ultimately, we started feeling like the extra complexity of the shadow DOM wasn't worth its benefits.

Removing shadow DOM from atom-text-editor

This pull request aims at removing the shadow DOM in a way that prevents elements within atom-text-editor that we don't fully control (i.e. the grammar scopes applied as CSS classes to every line's inner <span>) from being mistakenly styled from the outside. Specifically, every syntactic class name will now be prepended with syntax--; styling a JavaScript operator, for example, should now look like the following:

.syntax--source.syntax--js .syntax--operator {
  color: green;
}

Given the wide scope of this change, we are trying to reduce package breakage to a minimum by transforming deprecated selectors automatically and emitting a deprecation in deprecation-cop. Applying these transformations can be quite onerous and therefore we have introduced a layer of caching in 07d56b2 to ensure their performance impact stays low after the first time Atom is launched.

Other notable changes include:

  • Block decorations rendering has been rewritten from scratch (it previously worked only with the shadow DOM).

  • When the editor is focused, document.activeElement will now return the hidden <input> inside atom-text-editor, instead of atom-text-editor itself. I could find only one third party package relying on this behavior, and I am going to submit a pull request shortly to use document.activeElement.closest('atom-text-editor') instead.

  • Any Grim or style sheet deprecation will now cause specs to fail, see ensureNoDeprecatedFunctionCalls and ensureNoDeprecatedStylesheets.

  • Rendering flashes for decorations has been broken for a long time, but removing the shadow DOM highlighted the cause of the issue, and so they have been fixed in 1091b0e.

    flash

Conclusion

Removing the shadow DOM is an important step for the editor's future, as doing so will allow us to extract it, clean it up and optimize it. Considering the invasive nature of this change, we are planning to 🚢 it after rolling the railcars next week, so that everyone can build it from master and extensively test it for an entire release cycle.

Before merging this, we will also need to merge the following core package pull requests, so that we stop using any deprecated API or selector, and make the test suite green again:

/cc: exit

__Originally posted by @as-cii in https://github.com/atom/atom/pull/12903__start

a bug

encoding should be 'gbk'

Start

Background

Back in #3943, we started rendering text editor contents inside a shadow DOM boundary to prevent outer style sheets from altering the editor inner contents. Although this introduced many benefits, it carried several tradeoffs too.

In particular, we started duplicating some of the style sheets ({context: 'atom-text-editor'}) inside each shadow root so that package and theme authors could still be able to explicitly target the editor visual elements. An alternative solution to this problem that we suggested (before the shadow DOM spec was finalized) was to use /deep/ and ::shadow pseudo selectors, which unfortunately are now being deprecated.

Even more importantly, this architecture was getting in the way of making atom-text-editor a reusable component, because it made it even more coupled to Atom and its ecosystem. Other problems involved things like #4590, and a more complex user experience for package and theme authors in general. Ultimately, we started feeling like the extra complexity of the shadow DOM wasn't worth its benefits.

Removing shadow DOM from atom-text-editor

This pull request aims at removing the shadow DOM in a way that prevents elements within atom-text-editor that we don't fully control (i.e. the grammar scopes applied as CSS classes to every line's inner <span>) from being mistakenly styled from the outside. Specifically, every syntactic class name will now be prepended with syntax--; styling a JavaScript operator, for example, should now look like the following:

.syntax--source.syntax--js .syntax--operator {
  color: green;
}

Given the wide scope of this change, we are trying to reduce package breakage to a minimum by transforming deprecated selectors automatically and emitting a deprecation in deprecation-cop. Applying these transformations can be quite onerous and therefore we have introduced a layer of caching in 07d56b2 to ensure their performance impact stays low after the first time Atom is launched.

Other notable changes include:

  • Block decorations rendering has been rewritten from scratch (it previously worked only with the shadow DOM).

  • When the editor is focused, document.activeElement will now return the hidden <input> inside atom-text-editor, instead of atom-text-editor itself. I could find only one third party package relying on this behavior, and I am going to submit a pull request shortly to use document.activeElement.closest('atom-text-editor') instead.

  • Any Grim or style sheet deprecation will now cause specs to fail, see ensureNoDeprecatedFunctionCalls and ensureNoDeprecatedStylesheets.

  • Rendering flashes for decorations has been broken for a long time, but removing the shadow DOM highlighted the cause of the issue, and so they have been fixed in 1091b0e.

    flash

Conclusion

Removing the shadow DOM is an important step for the editor's future, as doing so will allow us to extract it, clean it up and optimize it. Considering the invasive nature of this change, we are planning to 🚢 it after rolling the railcars next week, so that everyone can build it from master and extensively test it for an entire release cycle.

Before merging this, we will also need to merge the following core package pull requests, so that we stop using any deprecated API or selector, and make the test suite green again:

/cc: @atom/core

__Originally posted by @as-cii in https://github.com/atom/atom/pull/12903__exit

书中的一个IndexError

def getshow():
    # Positive sample
    all_x = X[:, 0]
    all_y = X[:, 1]
    # Negative sample
    all_negative_x = [1, 0]
    all_negative_y = [1, 1]
    # Calculate the slope and intercept of the dividing line
    k = -W[2] / W[3]
    b = -(W[0] + W[1]) / W[3]
    xdata = np.linspace(0, 5)
    plt.figure()
    plt.plot(xdata, xdata * k + b, 'r')
    plt.plot(all_x, all_y, 'bo')
    plt.plot(all_negative_x, all_negative_y, 'yo')
    plt.show()

在书中第15页。其中不存在W[2]和W[3],所以k和b无法求解

矩阵shape不匹配

请问一下,运行similarity.py脚本出现如下错误:ValueError: Layer weight shape (512, 128) not compatible with provided weight shape (512, 312),请问是哪里的问题?谢谢

ner: IDCNN+CRF--报错:InvalidArgumentError (see above for traceback): ConcatOp : Dimensions

InvalidArgumentError (see above for traceback): ConcatOp : Dimensions of inputs should match: shape[0] = [1,3,100] vs. shape[1] = [1,1,20]
[[node char_embedding/concat (defined at /home/admin/workspace/NER_IDCNN_CRF/model.py:155) = ConcatV2[N=2, T=DT_FLOAT, Tidx=DT_INT32, _device="/job:localhost/replica:0/task:0/device:CPU:0"](char_embedding/embedding_lookup, char_embedding/seg_embedding/embedding_lookup, char_embedding/concat/axis)]]

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.