Programmers should spend gracefully

Valentine's Day is just a normal overtime Thursday for programmers, but if you are thinking about "how to spend Valentine's Day", congratulations! You have transcended the realm of programming languages ​​and IDEs, standing at the top of the programmer's contempt chain!

So, how do programmers spend Valentine's Day gracefully?

0. A wave of code love poems confession

If you are still single, when someone goes on a date, we can live this gracefully:

1. Read a book
A must-read for developers in 2019! 20 Alibaba tech giants have helped you make a list of classic books! - Yunqi Community - Alibaba Cloud

10 Free Machine Learning and Data Science Books

10 Free Books You Must Read in Machine Learning and Data Science

Seven Books to Lay the Mathematical Foundation of Machine Learning and Data Science Develop Paper

The Best Deep Learning Books of 2018

2. Watch brain-burning movies on Youku

3. Listen to songs on Xiami, and the efficiency of typing code will fly
The following songs are recommended by Yunqi Weibo developer fans:

1).The Scientist

2). Cowboys are busy

3). Taoist pure heart formula

4). One of my Taoist nun friends

5). Pantheon

6). Tick

7). Great Compassion Mantra

8). Small half

9). The piano song of the night

10).Dream wedding

4. Learn to program
Free programming language lecture series:

[Wei Wei Python Sharing Collection] How to quickly master the basics of Python programming? Here is your key to mastering the world of Python programming! - Yunqi Community - Alibaba Cloud

[Dry Goods Collection] I heard that the city of Hangzhou belongs to Java, it is better to listen to Alibaba experts talk about Java - Programmer Sought

[Playback video + PPT download and organize] Programming Language Lecture Series: Deep Learning JavaScript and React Technology - Yunqi Community - Alibaba Cloud

[Playback video + PPT download and organize] Programming Language Lecture Series: Getting Started with C++ Quickly

5. Follow the latest open source projects
At the end of 2018 and the beginning of 2019, Alibaba released several open source projects:

1). Deep learning framework X-Deep Learning

Open source time: December 2018

Project Introduction:

Artificial intelligence technology with deep learning as the core has achieved great success in speech recognition, computer vision, natural language processing and other fields in the past few years. Among them, the hardware computing power represented by GPU and the excellent open source deep learning framework start to a huge boost.

Although open source frameworks represented by TensorFlow, PyTorch, MxNet, etc. have achieved great success, when we apply deep learning technology to large-scale industrial-level scenarios such as advertising, recommendation, and search, we find that these frameworks are not very good. meet our needs. The contradiction is that most open source frameworks are designed for low-dimensional continuous data such as images and speech, while many core application scenarios of the Internet (such as advertising/recommendation/search) are often faced with high-dimensional, sparse and discrete heterogeneous data, and the scale of parameters varies Tens of billions or even hundreds of billions. Further, many product applications require real-time training and updating of large-scale deep models. Existing open source frameworks are often difficult to meet the requirements of industrial-level production applications in terms of distributed performance, computing efficiency, horizontal scalability, and real-time system adaptability. need.

X-DeepLearning is an industrial-grade deep learning framework designed and optimized for such scenarios. After being tempered by Alibaba's advertising business, XDL has performed well in terms of training scale, performance, and horizontal scalability. Industrial-grade algorithmic solutions in the recommendation/search domain.

github address:
https://github.com/alibaba/x-deeplearning

More introduction:

Ali open source the first deep learning framework X-Deep Learning! - Yunqi Community - Alibaba Cloud

2). Self-developed scientific computing engine Mars

Open source time: January 2019

Project Introduction:

Scientific computing is numerical computing, which refers to the application of computers to deal with mathematical computing problems encountered in scientific research and engineering technology. Scientific computing is used in many fields such as image processing, machine learning, and deep learning. There are many languages ​​and libraries that provide tools for scientific computing. Among them, Numpy has become the leader with its concise and easy-to-use syntax and powerful performance, and has formed a huge technology stack based on this.


Numpy's core concept of multidimensional arrays is the basis for various upper-level tools. Multidimensional arrays are also called tensors. Compared with two-dimensional tables/matrices, tensors have more powerful expressive power. Therefore, the popular deep learning frameworks are also widely based on tensor data structures.


With the boom in machine learning/deep learning, the concept of tensors has gradually become familiar, and the scale requirements for general-purpose computations on tensors are also increasing. But the reality is that excellent scientific computing libraries such as Numpy are still stuck in the stand-alone era and cannot break through the scale bottleneck. The current popular distributed computing engines are not born for scientific computing. The mismatch of upper-level interfaces makes it difficult to write scientific computing tasks with traditional SQL/MapReduce. The execution engine itself is not optimized for scientific computing, which makes the computing efficiency unsatisfactory.

Based on the above status of scientific computing, Alibaba unified big data computing platform MaxCompute R&D team, after more than a year of research and development, broke the boundaries of big data and scientific computing fields, completed the first version and opened it up. Mars, a unified distributed computing framework based on tensors. Using Mars for scientific computing not only reduces the completion of large-scale scientific computing tasks from thousands of lines of code in MapReduce to a few lines of code in Mars, but also greatly improves performance. At present, Mars implements the tensor part, that is, numpy distributed, and implements 70% of the common numpy interfaces. In the follow-up, in the version of Mars 0.2, pandas is being distributed, and a fully compatible interface with pandas will be provided to build the entire ecosystem.

Mars, as a new generation of ultra-large-scale scientific computing engine, not only enters the distributed era of inclusive scientific computing, but also makes efficient scientific computing possible with big data.

github address:
https://github.com/mars-project/mars

More introduction:

Ali's first self-developed scientific computing engine, Mars, is heavily open-sourced, revealing the secrets of ultra-large-scale scientific computing - Programmer Sought

Mars - A Matrix-Based Unified Distributed Computing Framework - Yunqi Community - Alibaba Cloud

What Mars is, what it can do, and how to do it - A note about Mars' sharing at PyCon China 2018 - Yunqi Community - Alibaba Cloud

How does Mars execute distributedly? - Alibaba Cloud Community

Mars Algorithm Practice - Face Recognition - Yunqi Community - Alibaba Cloud

3). Graph deep learning framework Euler

Open source time: January 2019

Project Introduction:

This is the first open-source graph deep learning framework in China after large-scale application of core business. This open source, Euler has built-in a large number of algorithms for users to use directly, and the relevant code can be downloaded on GitHub.

Graph learning and deep learning are both branches of artificial intelligence. As a big data marketing platform under Alibaba, Alimama innovatively combines graph learning and deep learning to launch Euler, which can help greatly improve marketing efficiency. Euler has been tempered and verified in Alimama's core business scenarios. At the same time, it also has high application value in scenarios involving complex network analysis such as finance, telecommunications, and medical care. For example, users can use Euler to learn and reason about complex heterogeneous graphs constructed based on user transactions and other financial data, and then apply them to scenarios such as financial anti-fraud.

github address:
https://github.com/alibaba/euler

More introduction:

Euler is out today! The first industrial-grade open source framework for deep learning of graphs in China, made by Alimama - Yunqi Community - Alibaba Cloud

4). Distributed transaction solution Fescar

Open source time: January 2019

Project Introduction:

Ali is one of the first companies in China to carry out distributed (microservice) transformation of applications, so it has long encountered the problem of distributed transactions under the microservice architecture.

In 2014, the Alibaba middleware team released TXC (Taobao Transaction Constructor) to provide distributed transaction services for applications within the group.

In 2016, TXC was transformed into a product and landed on Alibaba Cloud as GTS (Global Transaction Service), becoming the only cloud-based distributed transaction product in the industry at that time. It was among the public cloud and proprietary cloud solutions in Alibaba Cloud. , began to serve many external customers.

Since 2019, based on the technical accumulation of TXC and GTS, the Alibaba middleware team has launched the open source project Fescar (Fast & EaSy Commit And Rollback, FESCAR) to build this distributed transaction solution with the community.

TXC/GTS/Fescar are in the same line, and have delivered a unique answer to solving the distributed transaction problem under the microservice architecture.

github address:
https://github.com/alibaba/fescar

More introduction:

Alibaba Open Source Distributed Transaction Solution Fescar Full Analysis - Yunqi Community - Alibaba Cloud

5). Real-time computing platform Blink

Open source time: January 2019

Project Introduction:

Blink is a typical open-source technology of "passing fire through salary", which is inherited from the Flink open-source framework, which was first suitable for data processing in small-traffic Internet scenarios.

Because of its optimism about real-time computing, Alibaba took the lead in transforming Flink and pushed Flink's computing power to the peak. It launched the internal version of Blink, which reduced the computing delay to the millisecond level that humans cannot perceive: when browsing the web, you just blinked your eyes. , but the information processed on Taobao and Tmall has been refreshed 1.7 billion times.

How important is real-time computing? In 2004, Google opened the era of offline computing, which can perform timing calculations for massive data. However, with the development of e-commerce, finance and other industries, new demands have been placed on big data computing. In financial transaction scenarios, if the risk control system cannot By observing the behavior dynamics of each account in real time, it may be impossible to block dangerous transactions due to the risk of omission of 1 second delay, resulting in consumer losses. But real-time computing breaks through this technical bottleneck.

Now, all core businesses of Alibaba Group have used Blink. In addition to the double 11 technical exam, ET City Brain calculates 1,300 signal intersections and 4,500 videos in Hangzhou in real time, ensuring the smooth flow of traffic arteries; Taobao and Tmall display real-time "exclusive" pages for hundreds of millions of users every day.

In the future, this technology will also be applied to various scenarios in society: IoT devices in factory equipment can analyze data in real time and improve production yield; logistics delivery platforms can place orders immediately and deliver goods to consumers in time; The navigation software follows the travel trajectory at all times, and no longer misses the turning point...