SLIDE

The SLIDE package contains the source code for reproducing the main experiments in this paper.

Dataset

The Datasets can be downloaded in Amazon-670K. Note that the data is sorted by labels so please shuffle at least the validation/testing data.

TensorFlow Baselines

We suggest directly get TensorFlow docker image to install TensorFlow-GPU. For TensorFlow-CPU compiled with AVX2, we recommend using this precompiled build.

Also there is a TensorFlow docker image specifically built for CPUs with AVX-512 instructions, to get it use:

docker pull clearlinux/stacks-dlrs_2-mkl    

config.py controls the parameters of TensorFlow training like learning rate. example_full_softmax.py, example_sampled_softmax.py are example files for Amazon-670K dataset with full softmax and sampled softmax respectively.

Build/Run on Intel platform

Prerequisites:

CMake >= 3.0 Intel Compiler (ICC) >= 19

Build with ICC compiler

source /opt/intel/compilers_and_libraries/linux/bin/compilervars.sh -arch intel64 -platform linux
cd /path/to/slide-root
mkdir -p bin && cd bin 
# BDW (AVX2)
cmake .. -DCMAKE_CXX_COMPILER=icpc -DCMAKE_C_COMPILER=icc
# SKX/CLX (AVX512)
cmake .. -DCMAKE_CXX_COMPILER=icpc -DCMAKE_C_COMPILER=icc -DOPT_AVX512=1
# CPX (AVX512 + BF16)
cmake .. -DCMAKE_CXX_COMPILER=icpc -DCMAKE_C_COMPILER=icc -DOPT_AVX512=1 -DOPT_AVX512_BF16=1
make -j

Run on Intel SKX/CLX/CPX

cd bin
OMP_NUM_THREADS= KMP_HW_SUBSET=s,c,t KMP_AFFINITY=compact,granularity=fine KMP_BLOCKTIME=200 ./runme ../SLIDE/Config_amz.csv
For example, on CLX8280 2Sx28c:
OMP_NUM_THREADS=112 KMP_HW_SUBSET=2s,28c,2t KMP_AFFINITY=compact,granularity=fine KMP_BLOCKTIME=200 ./runme ../SLIDE/Config_amz.csv

For best performance please set Batchsize=multiple-of-logic-core-number from SLIDE/Config_amz.csv.

Results can be checked from the log file under dataset:

tail -f dataset/log.txt