In 2019, these 45 AI open source tools, you still don’t know?
Editor’s note: This article is from WeChat public account “AI Technology Base Camp “(ID: rgznai100), author: Jane.
A good tool that improves development efficiency and optimizes the project development process. Both the enterprise and the developer are looking for a development tool that suits them. However, choosing the right tool is not easy, and sometimes it is even a daunting task.
At the end of 2018, We have compiled the top 100 Python open source tools and projects that we received last year,Github open source project summary and the most popular open source project Top200. Today, AI Technology Base Camp (ID: rgznai100) selected 45 preparations by collecting new AI tools from home and abroad this year (2018.10–2019.10). Popular open source tools, I hope you will not miss it.
Jax collection Autograd and XLA’s high-performance machine learning research tools, contributed by Google Open Source. Many people say that Jax is a replacement for TF, which is simpler and easier to use.
Open source address: https://github.com/google/jax
2, AdaNet [Stars: 2.9k]
AdaNet is a lightweight framework based on TensorFlow. Automated learning of high-quality models with minimal expert intervention, AdaNet provides a common framework that can be used not only to learn neural network architectures, but also to learn to integrate and get better models.
Open Source Address:https ://github.com/tensorflow/adanet
3, TensorFlowExtended(TFX)[Stars:720]
TFX is Machine learning tools for production environments. With TFX, you can create a production-level machine learning pipeline for the many needs of production application deployment and best practices. TFX begins with the extraction of data and then provides services through data validation, feature engineering, training, and evaluation.
Open source address: https://github.com/tensorflow/tfx
TensorFlowFederated(TFF ) frameworks can be used for decentralized data learning and computational experiments. It implements the Federated Learning (FL) approach, which will provide developers with distributed machine learning to train shared ML models on multiple devices without data leaving the device. Among them, it provides a layer of privacy protection through encryption, and the weight of model training on the device is shared with the central model for continuous learning.
Open Source Address:https ://github.com/tensorflow/federated