Skip to content

Latest commit

 

History

History
61 lines (49 loc) · 2.48 KB

RELEASE.md

File metadata and controls

61 lines (49 loc) · 2.48 KB
Raw

Release 1.0.5

Major Features and Improvements

  • Updated the Tensorflow docker to 19.09
  • Added SSD with ResNet34 as backbone in PyTorch framework
  • Removed SSD with VGG16 in Caffe framework
  • Added support of running benchmarks in detached docker mode
  • Moved large files to Zenodo and reduced repository size

Release 1.0.4

Major Features and Improvements

  • Add TF inference test with integration of TensorRT
  • Add DIEN model from Alimama
  • Add BERT model, a GOOGLE version and an NVIDIA version
  • Add cascaded pyramid network
  • Add convolutional recurrent neural network
  • Add SSD with resnet18 as backbone uisng Caffe framework
  • Add SSD with VGG18 as backbone using Caffe framework
  • Add TensorRT implementation of SSD with VGG18 as backbone
  • Add Faster RCNN
  • Add Graph Convolutional Network
  • Add NMT TensorRT implementation
  • Add SegLink model
  • Add Wide & Deep model
  • Change unit of results to measurable metrics such as images/s, recommendations/s, sentences/s
  • Fix many miscellaneous issues
  • Refine scripts for accuracy tests

Release 1.0.3

Major Features and Improvements

  • Add new models: NCF for recommendation class, DSSD for object detection class
  • Nvidia docker has license issue on distribution, users have to download by themselves. Add script to install some dependencies
  • Add md5 checksum for some big files to help us spot the download issues
  • Add multi-card training for DIN model
  • Add accuracy test cases in Caffe CNN models. Inference engine from different vendor could compare not only performance number but also accuracy loss

Release 1.0.2

Major Features and Improvements

  • Add the trained checkpoint file for googlenet, resnet50, resnet152, densenet121
  • Add multi-card training in for CNN-Tensorflow, SSD, MaskRCNN, NMT

Release 1.0.1

Major Features and Improvements

  • Reorganize the automation workflow to improve the running scripts quality.
  • Users can choose run all of application in a few scripts or each application separately.
  • Add preprocessing script to extract and save data to csv file.
  • Remove Alexnet as it is out of date.
  • Remove Vgg16 as it is repeatedly used in SSD test.
  • Add TensorRT-5 inference script for Caffe model.
  • Add Tensorcore FP16 GEMM in micro tests.

Bug Fixes and Other Changes

DIN model:

  • Change the inference workload to apply 100 items for each user to recommend.
  • The inference batch size is based on number of users. It is set to 1, 32, and 64. Iteration of 1000 is applied to minimize the overhead.