Some other benchmarks

[mrgloom]

Here is some other benchmarks:

CNN
https://github.com/soumith/convnet-benchmarks
https://github.com/jcjohnson/cnn-benchmarks
https://github.com/ducha-aiki/caffenet-benchmark
https://github.com/baidu-research/DeepBench

RNN
https://github.com/glample/rnn-benchmarks

[gfursin]

Thanks @mrgloom .
Right now we focus on stabilizing our universal engine to be able to build DNN frameworks such as Caffe in different flavours (CPU,CUDA,OpenCL) across Linux, Windows and Android together with various sub-packages (BLAS) in a unified and reproducible way. Another focus is to provide customizable experimentation framework to be able to run and time various tests (such as object classification) also in a unified and reproducible way with JSON meta information, so that other researchers can reuse results and build upon them. Whenever ready, we will start adding other benchmarks or converting them into CK to be able to run across diverse hardware and data sets...

[mrgloom]

Just another benchmark from some Belarusian group of researchers (more focused on framework comparision, FCNN networks, also including distributed setting):
https://github.com/gakarak/DeepLearning_Frameworks_Benchmark

[mrgloom]

DAWNBench An End-to-End Deep Learning Benchmark and Competition
https://dawn.cs.stanford.edu/benchmark/