xrosliang/oneflow-benchmark

OneFlow models for benchmarking.

OneFlow Deep Learning Benchmarks

Introduction

This repository provides OneFlow deep learning benchmark examples for CV, CTR and NLP, and more models are on the way and will be provided here when ready.

Convolutional Networks for Computer Vision Classification

• ResNet-50
• ResNeXt-50-32*4d
• VGG-16
• Inception-V3
• AlexNet
• MobileNet-V2

Wide Deep Learning for Click-Through-Rate (CTR) Recommender Systems

• OneFlow-WDL

BERT for Nature Language Process

• BERT Pretrain for Language Modeling
• SQuAD for Question Answering
• CoLA and MRPC of GLUE

OneFlow Benchmark Test Reports

Model	DType	XLA	Throughput	Speedup on 32 devices
ResNet50-V1.5	Float32	No	11.6k imges/sec	30.4
BERT base Pretrain	Float32	No	530k tokens/sec	28.54