Commits · no_raw_index_calculation · gaoqiong / composable_kernel

10 May, 2021 4 commits
- v5r1 test · 6aa275b4
  Jing Zhang authored May 10, 2021
  
  6aa275b4
- refactor · 385f5513
  Chao Liu authored May 10, 2021
  
  385f5513
- refactor · cd100621
  Chao Liu authored May 10, 2021
  
  cd100621
- refactor driver for conv · d8c89b68
  Chao Liu authored May 10, 2021
  
  d8c89b68
07 May, 2021 1 commit
- add docker script · fd160c63
  Chao Liu authored May 07, 2021
  
  fd160c63
04 May, 2021 1 commit
- update v4r4 nhwc · 457ee9a1
  Chao Liu authored May 04, 2021
  
  457ee9a1
30 Apr, 2021 2 commits
- Merge remote-tracking branch 'origin/master' into no_raw_index_calculation · fe325666
  Chao Liu authored Apr 30, 2021
  
  fe325666
- fix bug in tensor adaptor · 65a00640
  Chao Liu authored Apr 30, 2021
  
  65a00640
29 Apr, 2021 2 commits
- added back pipelined 2x2 to blockwise gemm · fc148cef
  Chao Liu authored Apr 29, 2021
  
  fc148cef
- blockwise gemm does 3d*3d=4d · 0374f8de
  Chao Liu authored Apr 29, 2021
  
  0374f8de
28 Apr, 2021 3 commits
- Use Tuple and vector_type instead of Array for holding tensor data (#30) · d075adf1
  Chao Liu authored Apr 28, 2021
```
* replacing array with tuple and vector for tensor data
```
  d075adf1
- updated block-cluster in gridwise gemm and thread-cluster in blockwise copy to... · 4a661578
  Chao Liu authored Apr 28, 2021
```
updated block-cluster in gridwise gemm and thread-cluster in blockwise copy to use cluster descriptor
```
  4a661578
- added TensorAdaptor class and use it to implement cluster descriptor · 8b306478
  Chao Liu authored Apr 28, 2021
  
  8b306478
26 Apr, 2021 1 commit
- Merge remote-tracking branch 'origin/no_array' into no_raw_index_calculation · 2178d1d8
  Chao Liu authored Apr 26, 2021
  
  2178d1d8
24 Apr, 2021 1 commit
- update v5r1 · 7484a103
  Chao Liu authored Apr 24, 2021
  
  7484a103
23 Apr, 2021 6 commits
- clean up · ba31eb3e
  Chao Liu authored Apr 23, 2021
  
  ba31eb3e
- clean up · 08bb4372
  Chao Liu authored Apr 23, 2021
  
  08bb4372
- updating v5r1 · 905f5a3f
  Chao Liu authored Apr 23, 2021
  
  905f5a3f
- updating v5r1 · 474733b5
  Chao Liu authored Apr 23, 2021
  
  474733b5
- updating v5r1 · 415a4a5b
  Chao Liu authored Apr 23, 2021
  
  415a4a5b
- refactor DynamicBuffer · 32d485dd
  Chao Liu authored Apr 23, 2021
  
  32d485dd
22 Apr, 2021 5 commits
- bug fix · b6e43b25
  Chao Liu authored Apr 22, 2021
  
  b6e43b25
- added back amd_assembly_outer_product_1x2 and amd_assembly_outer_product_1x4 · f5654649
  Chao Liu authored Apr 22, 2021
  
  f5654649
- updating v5r1 · 9d5d6afa
  Chao Liu authored Apr 22, 2021
  
  9d5d6afa
- using tuple (instead of vector) for holding C thread matrix data to solve... · dcee43fe
  Chao Liu authored Apr 22, 2021
```
using tuple (instead of vector) for holding C thread matrix data to solve register over-allocation issue
```
  dcee43fe
- use vector type for holding C thread matrix data, but it cause register over-allocation · aeb05cc4
  Chao Liu authored Apr 22, 2021
  
  aeb05cc4
21 Apr, 2021 6 commits
- clean · d990eff6
  Chao Liu authored Apr 21, 2021
  
  d990eff6
- use StaticBuffer for thread matrix A/B in blockwise GEMM · 437c996a
  Chao Liu authored Apr 21, 2021
  
  437c996a
- fix bug · 36de63ff
  Chao Liu authored Apr 21, 2021
  
  36de63ff
- replace raw pointer with DynamicBuffer in blockwise and threadwise gemm · 888f1d68
  Chao Liu authored Apr 21, 2021
  
  888f1d68
- replacing array with vector for tensor data · 35d68cf8
  Chao Liu authored Apr 21, 2021
  
  35d68cf8
- replacing array with vector for tensor data · 712babe4
  Chao Liu authored Apr 21, 2021
  
  712babe4
20 Apr, 2021 2 commits
- replacing array with vector for tensor data · 03f7892a
  Chao Liu authored Apr 20, 2021
  
  03f7892a
- replacing array with vector for tensor data · e8421cca
  Chao Liu authored Apr 20, 2021
  
  e8421cca
19 Apr, 2021 1 commit
- replacing array with vector for tensor data · 4978c9e7
  Chao Liu authored Apr 19, 2021
  
  4978c9e7
17 Apr, 2021 3 commits
- replacing array with vector for tensor data · e38c1b73
  Chao Liu authored Apr 17, 2021
  
  e38c1b73
- replacing array with vector for tensor data · 841b1480
  Chao Liu authored Apr 17, 2021
  
  841b1480
- replacing raw index calculation with coordinate transformation · fa163f3b
  Chao Liu authored Apr 16, 2021
  
  fa163f3b
14 Apr, 2021 1 commit
- replacing raw index calculation with coordinate transformation · 82696a73
  Chao Liu authored Apr 14, 2021
  
  82696a73
13 Apr, 2021 1 commit
- Overhaul vector_type and use real vector for int8x4_t instead of aliasing from int32_t (#29) · e4790c25
  Chao Liu authored Apr 12, 2021
```
* overhaul vector_type, make int8x4_t real vector instead of aliasing from int32_t
```
  e4790c25