AWS Deep Learning AMIs Now Include Chainer 4.2, Keras 2.2, and MXNet Model Server 0.3

Posted on: Jul 5, 2018

The AWS Deep Learning AMIs for Ubuntu and Amazon Linux now come with the latest deep learning framework support for Chainer 4.2, Keras 2.2 as well as the Apache MXNet Model Server 0.3. This release also includes upgrades of the NVIDIA stack in the AMIs including the latest NVIDIA GPU driver 390.46, CUDA 8.0 and 9.0, as well as cuDNN 7.1.4 and NCCL 2.2.13.  

The AMIs also support TensorFlow, PyTorch, Apache MXNet and Gluon, CNTK, Caffe, and Theano, pre-installed and fully-configured with the latest NVIDIA CUDA stack. With the AMIs you can get started with writing your deep learning code in minutes and leverage the computational power of Amazon EC2 GPU instances such as Amazon EC2 P2 and P3.

For developers who want to experiment with NVIDIA CUDA 9.1 stack, the Deep Learning Base AMI for Ubuntu and Amazon Linux now comes with NVIDIA CUDA 9.1, cuDNN 7.0.5 and NCCL 2.1.15. You can switch the default CUDA 9.0 configuration on the AMI to the CUDA 9.1 stack by following the instructions on our developer guide tutorial.

The latest AWS Deep Learning AMIs are now available on the AWS Marketplace. You can learn more about the AMIs by using our getting started tutorial or our developer guide for more tutorials, resources, and release notes. You can also subscribe to our discussion forum to get new launch announcements and post your questions.