PyTorch Release 19.07

The NVIDIA container image for PyTorch, release 19.07, 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 18.04 including Python 3.6 environment
• NVIDIA CUDA 10.1.168 including cuBLAS 10.2.0.168
• NVIDIA cuDNN 7.6.1
• NVIDIA NCCL 2.4.7 (optimized for NVLink™ )
• APEX
• MLNX_OFED +3.4
• OpenMPI 3.1.3
• TensorBoard 1.14.0+nv
• TensorRT 5.1.5
• DALI 0.11.0 Beta
• Tensor Core optimized examples:
  • BERT
  • Mask R-CNN
  • Tacotron 2 and WaveGlow v1.1
  • SSD300 v1.1
  • Neural Collaborative Filtering (NCF)
  • Transformer
  • ResNet50 v1.5
  • GNMT v2
• Jupyter and JupyterLab:
  • Jupyter Client 5.3.1
  • Jupyter Core 4.5.0
  • Jupyter Notebook 5.7.8
  • JupyterLab 1.0.2
  • JupyterLab Server 1.0.0
  • Jupyter-TensorBoard

Driver Requirements

Release 19.07 is based on NVIDIA CUDA 10.1.168, which requires NVIDIA Driver release 418.67. 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.

GPU Requirements

Release 19.07 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

This PyTorch release includes the following key features and enhancements.

• PyTorch container image version 19.07 is based on PyTorch 1.2.0a0 including upstream commits up through commit f6aac41 from June 19, 2019.
• Latest version of NVIDIA cuDNN 7.6.1
• Latest version of MLNX_OFED +3.4
• Added TensorBoard 1.14.0+nv to the container.
• Latest versions of Jupyter Client 5.3.1, Jupyter Core 4.5.0, JupyterLab 1.0.2 and JupyterLab Server 1.0.0, including Jupyter-TensorBoard integration.
• Latest version of DALI 0.11.0 Beta
• Latest version of Ubuntu 18.04

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.

• BERT model. BERT, or Bidirectional Encoder Representations from Transformers, is a new method of pre-training language representations which obtains state-of-the-art results on a wide array of Natural Language Processing (NLP) tasks. This model is based on theBERT: Pre-training of Deep Bidirectional Transformers for Language Understanding paper. NVIDIA's implementation of BERT is an optimized version of the Hugging Face implementation, leveraging mixed precision arithmetic and Tensor Cores on V100 GPUs for faster training times while maintaining target accuracy.
• 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.
• 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.
• 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.
• NCF model. The Neural Collaborative Filtering (NCF) 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.
• Transformer model. 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.
• ResNet50 v1.5 model. The ResNet50 v1.5 model is a slightly modified version of the original ResNet50 v1 model that trains to a greater accuracy.
• 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 3 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.

Known Issues

• Performance of Mask R-CNN in FP32 precision is up to 20% slower in the 19.07 container compared to the previous release. For best performance on Mask R-CNN, it is recommended to use automatic mixed precision training. This can easily be done using the float16 option with the MaskRCNN example included in this container.
• Due to recent changes on batch norm multiplier initialization (PyTorch commit: c60465873c5cf8f1a36da39f7875224d4c48d7ca), all batch norm multiplier is initialized as constant 1, instead of uniformly distributed between 0 and 1, as it was previously. This has caused accuracy issue for our TACOTRON2 model. If similar accuracy regression is observed during an update from 19.06 to 19.07, we recommend to re-initialize the batch norm multiplier using uniformed distribution. This could be done by passing your model to the following function:
  Copy

  Copied!

              
  
              
  def init_bn(module):
      if isinstance(module, torch.nn.modules.batchnorm._BatchNorm):
    	  if module.affine:
    	      module.weight.data.uniform_()
      for child in module.children():
    	  init_bn(child)