Google released the TensorFlow machine learning framework specifically for graphical data
Google has recently officially introduced Neural Structured Learning (NSL), an open source machine learning framework, using Neural Graph Learning method to train neural networks based on a number of variety of charts and different structured data.
In particular, NSL is built for great compatibility with TensorFlow - an open source software library for machine learning in many types of language awareness and understanding tasks - and is specifically designed for both Experienced and inexperienced machine learning practitioners can use it. Basically, NSL can create model models for computer vision, deploy NLP, and run predictions from graphical data sets such as medical records or charts. knowledge (knowledge graphs).
- Google releases an urgent update for Chrome, users should update immediately
Neural Structured Learning is an open source machine learning framework
'Utilizing specific structured signals during training allows developers to obtain models with greater accuracy, especially when the amount of data that has been labeled is relative. small. Training based on structured signals also helps create stronger models. These types of training techniques have been widely used internally within Google to improve the performance of models in a more positive and faster way, '' TensorFlow's engineering team said in a blog post.
NSL can train machine learning models through supervised learning, semi-supervised learning or unsupervised learning, thereby creating models. The figure uses graphical signals to standardize the training process, in some cases less than 5 lines of code.
- Google launched Google Go worldwide, extremely light version with many useful features, can replace the Googe application
The transport structure of the Neural Structured Learning framework
In addition, the new framework also comes with a number of tools that can help developers structure their data and APIs to create training models as opposed to minimal code lines.
Earlier in April, Google Cloud also introduced several training solutions for structured data, such as linked sheets in BigQuery and AutoML Tables.
In another news related to artificial intelligence, last week Google AI (formerly known as Google Research) also released the open source tool SM3 - an optimizer dedicated to training models. large-scale languages like BERT from Google and OpenAI by GPT2.
- Google Photos now allows searching for images by text, extracting text in images
SM3 open source tool page on Github
You can find more information about the NSL machine learning framework at the following addresses:
- Google officially launches TensorBoard.dev and TensorFlow Enterprise
- Google launched TensorFlow Lite 1.0 for mobile devices and embedded devices
- Google researchers for gaming AI to improve enhanced learning ability
- How to install TensorFlow in Linux and Raspberry Pi
- Google introduced Coral, the platform that supports building AI IoT hardware integration
- 3.5 million WSL users can now use GPU Compute from Linux right on Windows
- Google released the API set for the .NET Framework
- The best Python tools for Machine Learning and Data Science
- Google released a huge AI training data warehouse with over 5 million photos of 200,000 locations worldwide
- Free online learning about AI and Machine learning on Google website