PyTorch Release 19.05
The NVIDIA container image for PyTorch, release 19.05, is available on NGC.
Contents of the PyTorch container
This container image contains the complete source of the version of PyTorch in
/opt/pytorch. It is pre-built and installed in Conda default environment (
/opt/conda/lib/python3.6/site-packages/torch/) in the container image.
The container also includes the following:
- Ubuntu 16.04 including Python 3.6 environment
- NVIDIA CUDA 10.1 Update 1 including cuBLAS 10.1 Update 1
- NVIDIA cuDNN 7.6.0
- NVIDIA NCCL 2.4.6 (optimized for NVLink™ )
- OpenMPI 3.1.3
- TensorRT 5.1.5
- DALI 0.9.1 Beta
- Tensor Core optimized examples:
- Jupyter and JupyterLab:
Release 19.05 is based on CUDA 10.1 Update 1, which requires NVIDIA Driver release 418.xx. However, if you are running on Tesla (Tesla V100, Tesla P4, Tesla P40, or Tesla P100), you may use NVIDIA driver release 384.111+ or 410. The CUDA driver's compatibility package only supports particular drivers. For a complete list of supported drivers, see the CUDA Application Compatibility topic. For more information, see CUDA Compatibility and Upgrades.
Release 19.05 supports CUDA compute capability 6.0 and higher. This corresponds to GPUs in the Pascal, Volta, and Turing families. Specifically, for a list of GPUs that this compute capability corresponds to, see CUDA GPUs. For additional support details, see Deep Learning Frameworks Support Matrix.
Key Features and Enhancements
- PyTorch container image version 19.05 is based on PyTorch 1.0.1commit 828a6a3b from March 31, 2019
- Latest version of NVIDIA CUDA 10.1 Update 1 including cuBLAS 10.1 Update 1
- Latest version of NVIDIA cuDNN 7.6.0
- Latest version of TensorRT 5.1.5
- Latest version of DALI 0.9.1 Beta
- Ubuntu 16.04 with April 2019 updates
Tensor Core Examples
These examples focus on achieving the best performance and convergence from NVIDIA Volta Tensor Cores by using the latest deep learning example networks for training. Each example model trains with mixed precision Tensor Cores on Volta, therefore you can get results much faster than training without tensor cores. This model is tested against each NGC monthly container release to ensure consistent accuracy and performance over time. This container includes the following tensor core examples.
- An implementation of the Mask R-CNN model. Mask R-CNN is a convolution based neural network for the task of object instance segmentation. The paper describing the model can be found here. NVIDIA’s Mask R-CNN model is an optimized version of Facebook’s implementation, leveraging mixed precision arithmetic using Tensor Cores on NVIDIA Tesla V100 GPUs for 1.3x faster training time while maintaining target accuracy.
- An implementation of the Tacotron 2 and WaveGlow v1.1 model. This text-to-speech (TTS) system is a combination of two neural network models: a modified Tacotron 2 model from the Natural TTS Synthesis by Conditioning WaveNet on Mel Spectrogram Predictions paper and a flow-based neural network model from the WaveGlow: A Flow-based Generative Network for Speech Synthesis paper.
- An implementation of the SSD300 v1.1 model. The SSD300 v1.1 model is based on the SSD: Single Shot MultiBox Detector paper. The main difference between this model and the one described in the paper is in the backbone. Specifically, the VGG model is obsolete and is replaced by the ResNet50 model.
- An implementation of the Neural Collaborative Filtering (NCF) model. The NCF model focuses on providing recommendations, also known as collaborative filtering; with implicit feedback. The training data for this model should contain binary information about whether a user interacted with a specific item. NCF was first described by Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu and Tat-Seng Chua in the Neural Collaborative Filtering paper.
- An implementation of the Transformer model architecture. The Transformer model is based on the optimized implementation in Facebook's Fairseq NLP Toolkit and is built on top of PyTorch. The original version in the Fairseq project was developed using Tensor Cores, which provides significant training speedup. Our implementation improves the performance and is tested on a DGX-1V 16GB.
- An implementation of the ResNet50 model. The ResNet50 v1.5 model is a slightly modified version of the original ResNet50 v1 model that trains to a greater accuracy.
- An implementation of the GNMT v2 model. The GNMT v2 model is similar to the one discussed in the Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation paper.
Automatic Mixed Precision (AMP)
NVIDIA’s Automatic Mixed Precision (AMP) for PyTorch is available in this container through a preinstalled release of Apex. AMP enables users to try mixed precision training by adding only 2 lines of Python to an existing FP32 (default) script. AMP will choose an optimal set of operations to cast to FP16. FP16 operations require 2X reduced memory bandwidth (resulting in a 2X speedup for bandwidth-bound operations like most pointwise ops) and 2X reduced memory storage for intermediates (reducing the overall memory consumption of your model). Additionally, GEMMs and convolutions with FP16 inputs can run on Tensor Cores, which provide an 8X increase in computational throughput over FP32 arithmetic.
Comprehensive guidance and examples demonstrating AMP for PyTorch can be found in the documentation.
For more information about AMP, see the Training With Mixed Precision Guide.
- Persistent batch normalization kernels are enabled by default in this build. This will provide a performance boost to many networks, but in rare cases may cause a network to fail to train properly. We expect to address this in the 19.06 container.