TensorFlow Release 18.01

The NVIDIA container image of TensorFlow, release 18.01, is available.

TensorFlow container image version 18.01 is based on TensorFlow 1.4.0.

Contents of TensorFlow

This container image contains the complete source of the version of NVIDIA TensorFlow in /opt/tensorflow. It is pre-built and installed as a system Python module.

To achieve optimum TensorFlow performance, for image based training, the container includes a sample script that demonstrates efficient training of convolutional neural networks (CNNs). The sample script may need to be modified to fit your application.

The container also includes the following:

• Ubuntu 16.04
  Note:

  Container image 18.01-py2 contains Python 2.7; 18.01-py3 contains Python 3.5.

• NVIDIA CUDA 9.0.176 including CUDA^® Basic Linear Algebra Subroutines library™ (cuBLAS) 9.0.282
• NVIDIA CUDA^® Deep Neural Network library™ (cuDNN) 7.0.5
• NVIDIA^® Collective Communications Library ™ (NCCL) 2.1.2 (optimized for NVLink™ )
• Horovod™ 0.11.2

Driver Requirements

Release 18.01 is based on CUDA 9, which requires NVIDIA Driver release 384.xx.

Key Features and Enhancements

This TensorFlow release includes the following key features and enhancements.

• Addition of Python 3 package
• Horovod is now pre-installed in the container
• Updated OpenSeq2Seq example to include latest bug fixes
• Latest version of cuBLAS
• Latest version of cuDNN
• Latest version of NCCL
• Ubuntu 16.04 with December 2017 updates

Known Issues

cuBLAS 9.0.282 regresses RNN seq2seq FP16 performance for a small subset of input sizes. As a workaround, revert back to the 11.12 container.