[New]TensorFlow Core API Documentation

[Goldie Gadde]

bcc: anno...@tensorflow.org

Hello TensorFlow Developers,

We are happy to announce the launch of a new Build with Core section in the TensorFlow guide ! This initiative provides users with a new entry point into the TensorFlow ecosystem where they can learn to use the TensorFlow Core APIs for building highly configurable, performant, and scalable computational workflows from the ground up.

The TensorFlow Core APIs provide a set of comprehensive, composable, and extensible low-level APIs primarily aimed at building machine learning (ML) models with fine-grained control as well as authoring ML workflow tools and frameworks within the TensorFlow platform. The Core APIs can also be applied for non-ML scientific computing purposes such as physics simulations and graphics. The target audience for these Core APIs consists of researchers, library builders, framework authors, and high-level API users interested in adding additional functionalities to their ML workflows. 

The Build with Core section contains the following guides:

• Overview 

• Quickstart for Core

• Logistic regression for binary classification 

• Multilayer perceptrons for digit recognition

• Matrix approximation for image compression

• Custom optimizers

• DTesnor with Core APIs

Build with Core will now be the designated section for tutorials that cover building from scratch using Core APIs. It will continue to be updated with more content over time. 

Best,

Goldie Gadde (On behalf of TensorFlow Team)