Commits · fa6d8037905626d6252789f91358b260c8f5831b · gaoqiong / composable_kernel

29 Jul, 2019 1 commit
- Changed scalar type to vector type for threadwise gemm for fp16 and bfloat16 data types · fa6d8037
  Tejash Shah authored Jul 29, 2019
  
  fa6d8037
26 Jul, 2019 1 commit
- Added fp16 support in implicit gemm · 2185affb
  Tejash Shah authored Jul 26, 2019
  
  2185affb
28 Jun, 2019 1 commit
- some benchmark on vega 7 · f0716f5b
  Chao Liu authored Jun 28, 2019
  
  f0716f5b
27 Jun, 2019 2 commits
- tested on P100 · dab29389
  Chao Liu authored Jun 27, 2019
  
  dab29389
- do more benchmark · 85ae70d3
  Chao Liu authored Jun 26, 2019
  
  85ae70d3
26 Jun, 2019 1 commit
- add more test · 35269cf7
  Chao Liu authored Jun 26, 2019
  
  35269cf7
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
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