Horovod
Simplifying and accelerating distributed deep learning with Open-source framework for easy multi-GPU/machine training across TensorFlow, Keras, PyTorch, and Apache MXNet.
Horovod is an open-source distributed deep learning training framework for TensorFlow, Keras, PyTorch, and Apache MXNet, simplifying and accelerating the process with easy-to-use functionality for horizontal data parallelism across multiple GPUs or machines. It reduces training time through its DistributedOptimizer and offers utilities for parameter broadcasting and gradient synchronization, with noteworthy achievements including integrations on Microsoft Azure and support for various deep learning frameworks. Horovod is licensed under the Apache 2.0 license, fostering a strong community presence and open-source collaboration.
About Horovod
Horovod is a open-source distributed deep learning training framework, developed for TensorFlow, Keras, PyTorch, and Apache MXNet. The mission of Horovod is to simplify and accelerate the process of distributed deep learning. Horovod offers easy-to-use functionality that allows users to run deep learning models across multiple GPUs or machines with minimal configuration.
Horovod provides a DistributedOptimizer for TensorFlow, which enables training on large datasets using multiple GPUs or machines. By leveraging the power of horizontal data parallelism, Horovod can significantly reduce the time required to train deep learning models. The framework also includes utilities for broadcasting model parameters and synchronizing gradients across workers.
Some notable achievements of Horovod include its integration with popular cloud platforms such as Microsoft Azure and its support for a variety of deep learning frameworks including TensorFlow, Keras, PyTorch, and Apache MXNet. The project has a strong community presence with active Slack channels, mailing lists, and public discussion forums.
Horovod is licensed under the Apache 2.0 license and is open source, allowing anyone to contribute to its development and use it for their deep learning projects.