google/gemmlowp

Issues

• Suggestions for resources to understand gemmlowp

  #210 opened 3 years ago by mrtpk123
  2

• is D in your fomation (2) calculated by this?

  #204 opened 4 years ago by mittkongg
  0

• int8*int8 -> float?

  #203 opened 4 years ago by XapaJIaMnu
  3

• core dump in the neon-gemm-kernel-benchmark.cc.

  #138 opened 7 years ago by liyancas
  5

• How to use gemmlowp in C project?

  #200 opened 5 years ago by ArtyZe
  0

• Issues compiling for bare metal application

  #196 opened 5 years ago by dav1d-wright
  2

• SIMD back-end for IBM Power and Z

  #195 opened 5 years ago by geert56
  3

• How to quantize accumulator from int32 to uint8

  #188 opened 5 years ago by kmhatre14
  4

• error result when W and X don't range from -1 to 1

  #167 opened 6 years ago by pingfengluo
  2

• How to calculate non-linear function after 8bit quantization ?

  #192 opened 5 years ago by youck9224
  1

• what is "ab_x2_high32" in <func::SaturatingRoundingDoublingHighMul> stand for?

  #191 opened 5 years ago by jiaheni1127
  0

• no such package '@remotejdk_linux//':

  #190 opened 5 years ago by yue1-li
  0

• Quantize matmul in CPU avx2 have effect？

  #189 opened 5 years ago by zhoujianqian
  0

• There is a problem about how to quantize the accumulator(int32) into uint8

  #144 opened 6 years ago by sxsxsx
  2

• How can I use a new gemm-kernel in tensorflow or other machine learning framework?

  #185 opened 6 years ago by knjwhn
  1

• Python 3 support

  #182 opened 6 years ago by dibya001
  2

• Error compilation for Windows x64 using MingGW 64

  #180 opened 6 years ago by Arritmic
  1

• is only conv layer supported?

  #158 opened 6 years ago by yfyan0204
  2

• Is this product range of int8*int8 in comment document expected?

  #174 opened 6 years ago by zhenhuaw-me
  3

• CellFormat in AVX2 kernel incorrect? Question for clarification

  #159 opened 6 years ago by SuperFluffy
  6

• Add two feature maps

  #170 opened 6 years ago by Aspirinkb
  1

• run dotprod instruction failed on apple A12 and qualcomm 845

  #175 opened 6 years ago by xyoungli
  1

• A problem with the design of kernel in Arm64

  #162 opened 6 years ago by knjwhn
  1

• eight_bit_int_gemm get all zero output and segment fault

  #160 opened 6 years ago by lcxywfe
  4

• there is a problem about the result_scale and result_zero_point?

  #137 opened 7 years ago by victorygogogo
  9

• Document code style/formatting

  #154 opened 6 years ago by alexfru
  1

• [BUG] Failed to build on several architectures

  #136 opened 6 years ago by cdluminate
  5

• Bug in neon-gemm-kernel-benchmark.cc?

  #133 opened 6 years ago by flyingforyou
  5

• Mixing openmp with gemmlowp multithreading causes low performance

  #148 opened 6 years ago by liyinhgqw
  1

• Does avx2 feature of gemmlowp support gcc4.8.5?

  #141 opened 6 years ago by Qoboty
  4

• run ./correctness_meta_gemm ...............................Bus error

  #145 opened 6 years ago by sxsxsx
  1

• iOS gemmlowp_test failed with linker

  #140 opened 7 years ago by harouwu
  4

• issue with aligned_alloc on macOS 10.12 with clang 802.0.42 cannot find <malloc.h>

  #134 opened 7 years ago by samhodge
  2

• Reconsider GEMMLOWP_ALLOW_SLOW_SCALAR_FALLBACK

  #117 opened 7 years ago by jart
  2

• Recent fix causes 2.5x performance regression

  #108 opened 7 years ago by andreasgal
  17

• Compile-time warning: implicit conversion loses integer precision: 'unsigned long' to 'int' [-Wshorten-64-to-32]

  #106 opened 7 years ago by Arcank
  1