The NVIDIA application of DIGITS, release 18.02, is available.
NVIDIA DIGITS application version 18.02 is based on DIGITS 6.1.
Contents of DIGITS
This application contains the complete source of the version of DIGITS in /opt/digits
. It is pre-built and installed into the /usr/local/python/
directory in the application.
The container also includes the following:
- Ubuntu 16.04 including Python 2.7
- NVIDIA CUDA 9.0.176 including:
- CUDA® Basic Linear Algebra Subroutines library™ (cuBLAS) 9.0.282 Patch 2 which is installed by default
- cuBLAS 9.0.234 Patch 1 as a debian file. Installing Patch 1 by issuing the
dpkg -i /opt/cuda-cublas-9-0_9.0.234-1_amd64.deb
command is the workaround for the known issue described below.
- NVIDIA CUDA® Deep Neural Network library™ (cuDNN) 7.0.5
- NCCL 2.1.2 (optimized for NVLink™ )
- NVCaffe™ 0.16.5
- Torch™ 7
- TensorFlow™ 1.4.0
Driver Requirements
Release 18.02 is based on CUDA 9, which requires NVIDIA Driver release 384.xx.
Key Features and Enhancements
This DIGITS container image version includes the following key features and enhancements.
- NVIDIA DIGITS application version 18.02 is based on DIGITS version 6.1
- Allow WebSocket to automatically detect browser scheme
- Fixed missing
URL_PREFIX
in html files - Fixed
dtype
error due to TensorFlow API change - Dependency update to match three frameworks
- Latest version of cuBLAS
- Ubuntu 16.04 with January 2018 updates
Known Issues
cuBLAS 9.0.282 regresses RNN seq2seq FP16 performance for a small subset of input sizes. This issue should be fixed in the next update. As a workaround, install cuBLAS 9.0.234 Patch 1 by issuing the dpkg -i /opt/cuda-cublas-9-0_9.0.234-1_amd64.deb
command.