Commits · 32850b937426db543f12fd077a56db5f50fc383c · gaoqiong / composable_kernel

09 Oct, 2019 1 commit
- Ported xdlops kernels to debug bwdwrw fp32/fp16/bfp16 issue. Verified atleast fwd data fp32 works. · 32850b93
  Wen-Heng (Jack) Chung authored Oct 09, 2019
  
  32850b93
29 Jul, 2019 1 commit
- Add fp16 support in implicit gemm · 583755a7
  Tejash Shah authored Jul 29, 2019
  
  583755a7
27 Jun, 2019 1 commit
- do more benchmark · 85ae70d3
  Chao Liu authored Jun 26, 2019
  
  85ae70d3
24 Jun, 2019 2 commits
- debugging vector load for generic tensor copy · e55cfe15
  Chao Liu authored Jun 24, 2019
  
  e55cfe15
- enabling vector load on merged dim · df29a7e0
  Chao Liu authored Jun 24, 2019
  
  df29a7e0
17 Jun, 2019 1 commit
- refactoring for miopen · 33d1e0e2
  Chao Liu authored Jun 17, 2019
  
  33d1e0e2
13 Jun, 2019 2 commits
- added strides and dilations suppport to implicit gemm v4 · b1cb48a0
  Chao Liu authored Jun 13, 2019
  
  b1cb48a0
- reorginzed files · 1566b317
  Chao Liu authored Jun 13, 2019
  
  1566b317