github.com/marian-nmt/FBGEMM.git - Unnamed repository; edit this file 'description' to name the repository.

Age	Commit message (Expand)	Author
2020-01-14	Fix compile errorsyouki/merge-win-int8	Young Jin Kim
2020-01-14	Update jit parameters for windows	Young Jin Kim
2020-01-14	Fix undefined reference error on macos (#247)	Hongzhang Shan
2020-01-14	Cmake changes for MSVC build (#242)	Daya Khudia
2020-01-14	Allocate memory instead of using variable length arrays (#241)	Daya Khudia
2020-01-14	More fixes related to MSVC build (#240)	Daya Khudia
2020-01-14	fix perf regression in resnet50 and resnet101 (#239)	Daya Khudia
2020-01-13	Fix multi-instance benchmarking for fp16 (#246)	Daya Khudia
2020-01-10	attributes not at the right place	Daya Khudia
2020-01-10	4-bit SLS with emb dim not a multiple of 2 (#243)	Jongsoo Park
2020-01-09	Githubaction (#237)	Hongzhang Shan
2020-01-07	4bit JIT'ed SLS kernel (#233)	Jongsoo Park
2020-01-05	start to add github action file	Hongzhang Shan
2020-01-04	Handle remainder loop with masked SIMD instructions (#235)	Daya Khudia
2020-01-04	Restructure code gen for FBGEMMFP16 (#236)	Evgeny Fiksman
2020-01-03	Adding missing copyright message	Daya Khudia
2020-01-02	simplify SparseAdagrad (#227)	Jongsoo Park
2020-01-02	simplify EmbeddingSpMDM (#228)	Jongsoo Park
2019-12-27	add more comments on handling small scale handling; remove redundant if (#232)	Jongsoo Park
2019-12-26	Fixing cblas.h for mac build in fp16 benchmark	Hongzhang Shan
2019-12-26	refine special handling of small quant scale (#231)	Jongsoo Park
2019-12-20	Fix ISA detection mechanism (#225)	Evgeny Fiksman
2019-12-20	Adding FP32 JIT SparseAdagrad and Rowwise Adagrad (#224)	Protonu Basu
2019-12-20	Using import/export correctly	Daya Khudia
2019-12-19	cleaning up some code in referene implementations WRT SLS that is not needed	Protonu Basu
2019-12-19	Avoid GNU compiler specific C++ extension	Daya Khudia
2019-12-18	Streamline usage of attributes (#222)	Daya Khudia
2019-12-18	Cast instead of reinterpret_cast in recently added code	Daya Khudia
2019-12-17	Fix build with C++11 (#223)	Evgeny Fiksman
2019-12-17	avoid enforcment of unsupported ISA (#221)	Evgeny Fiksman
2019-12-13	Use cast intrinsics	Daya Khudia
2019-12-12	Suitable declaration for cblas_gemm_compute	Daya Khudia
2019-12-12	Explicit template parameters	Daya Khudia
2019-12-12	Things like iota/accumulate are defined in numeric header	Daya Khudia
2019-12-12	fbgemm API decorator only needed at declaration	Daya Khudia
2019-12-12	Aligned memory allocation that works with MSVC as well	Daya Khudia
2019-12-11	Back out "Revert D18138146: [caffe2] Add support for AVX512-256(YMM) in FBGEM...	Evgeny Fiksman
2019-12-11	Revert D18138146: Add support for AVX512-256(YMM) in FBGEMM16	Jianyu Huang
2019-12-11	Add support for AVX512-256(YMM) in FBGEMM16 (#209)	Evgeny Fiksman
2019-12-11	Reapply fix for make_unique not supported by C++11 (#217)	Evgeny Fiksman
2019-12-11	Fix non-MKL build in CI (#216)	Evgeny Fiksman
2019-12-10	FBGEMM CI avoid call to C++14 (#215)	Evgeny Fiksman
2019-12-10	Fix FBGEMM OSS CI (#212)	Evgeny Fiksman
2019-12-10	Fix pytorch build (#214)	Evgeny Fiksman
2019-12-10	Fix FBGEMM OSS CI (#211)	Jianyu Huang
2019-12-10	Add additional execution arguments to the benchmark (#207)	Evgeny Fiksman
2019-12-10	Pre-transpose Weight matrix (B) before sending to MKL routine (#205)	Evgeny Fiksman
2019-12-10	Enhance benchmark to include cache eviction and multi instance with OMP (#204)	Evgeny Fiksman
2019-12-09	Adding FP 32 SLS, and unifying it with 8 Bit SLS (#206)	Protonu Basu
2019-12-09	Add the the FP16 <-> FP32 conversion benchmark (#210)	Jianyu Huang