What is the name of Google's deep learning framework released in 2015?

Google’s deep-learning framework released in 2015 was TensorFlow, an open-source library created by the Google Brain team.

TensorFlow grew out of DistBelief, Google’s earlier internal system for training neural networks. Google released TensorFlow under the Apache 2.0 license on November 9, 2015, making the technology available to researchers, developers, and companies outside Google.

The name refers to tensors, multidimensional arrays of data that flow through a computational graph as a model performs operations. TensorFlow supports training and inference on CPUs, GPUs, and Google’s tensor processing units, and it can run from large server clusters to mobile and edge devices.

TensorFlow is sometimes confused with Keras, Caffe, or PyTorch. Keras is a high-level neural-network API that became closely integrated with TensorFlow, while PyTorch and Caffe are separate frameworks. TensorFlow 2.0, released in 2019, emphasized a simpler, more intuitive programming approach.