TensorFlow on AWS

Enhance and visualize your deep learning applications with ML tools

Fine-tune applications with visualization tools, including histograms and graphs, to quickly train deep neural networks.

Train and deploy your deep learning models on AWS securely with optimal performance and currency.

Access documentation and tutorials to accelerate your artificial intelligence (AI) development and join an active community on GitHub.

How it works

Researchers and developers can use TensorFlow to help enhance their applications with machine learning (ML). AWS provides broad support for TensorFlow, helping customers develop and serve their own models across computer vision (CV), natural language processing (NLP), speech translation, and more.

Diagram shows how you can train models in TensorFlow; start the server; apply tools to understand, debug, and enhance your applications; and then update versioning.

Use cases

Access state-of-the-art models

Get distributed training in the latest NLP and CV models with tf.distribute.strategy.

Deploy enhanced models

Deploy NLP and CV models with TensorFlow Serving—a flexible, high-performance serving system for ML models.

Visualize training and performance

Enhance your ML models with TensorBoard, a visualization toolkit used to host, track, and share your ML experiments.

How to get started

Explore TensorFlow on AWS

Start using TensorFlow with SageMaker, AWS Deep Learning AMIs, and more.

Create a free account

Instantly get access to the AWS Free Tier

Start building

Get started building in the AWS Management Console.

Explore more of AWS