Chao Liu
|
42e9613bb8
|
fixed build issue
[ROCm/composable_kernel commit: f2b92ba945]
|
2019-06-12 09:59:28 -05:00 |
|
Chao Liu
|
14451df1fd
|
reorginze files
[ROCm/composable_kernel commit: 81497a93a0]
|
2019-06-11 23:49:51 -05:00 |
|
Chao Liu
|
3dc2ff612c
|
rename files, added header guard, added namespace
[ROCm/composable_kernel commit: 88b77181aa]
|
2019-06-11 14:35:59 -05:00 |
|
Chao Liu
|
88d7d31a76
|
remove .hip extension
[ROCm/composable_kernel commit: 05e046654c]
|
2019-06-11 09:12:41 -05:00 |
|
Chao Liu
|
6fa187c7ac
|
refactor
[ROCm/composable_kernel commit: a68b16a5d9]
|
2019-06-06 19:49:40 -05:00 |
|
Chao Liu
|
d141b32528
|
use more constexpr for Array
[ROCm/composable_kernel commit: 0a386c46a9]
|
2019-06-06 19:26:08 -05:00 |
|
Chao Liu
|
07ad21131b
|
fix build issue
[ROCm/composable_kernel commit: 08c692433e]
|
2019-06-06 17:13:15 -05:00 |
|
Chao Liu
|
5e7dff691c
|
refactor
[ROCm/composable_kernel commit: 7a89684f92]
|
2019-06-06 16:50:35 -05:00 |
|
Chao Liu
|
92ae4c49bc
|
fix build issue
[ROCm/composable_kernel commit: eafdabba77]
|
2019-06-04 20:11:04 -05:00 |
|
Chao Liu
|
c26ccd3d07
|
use more constexpr
[ROCm/composable_kernel commit: 709f13a6d7]
|
2019-06-04 20:00:48 -05:00 |
|
Chao Liu
|
f11222a3ac
|
try using more constexpr
[ROCm/composable_kernel commit: 498e71b098]
|
2019-06-04 17:02:49 -05:00 |
|
Chao Liu
|
da9ea59483
|
use vectorized read and write for threadwise generic tensor copy
[ROCm/composable_kernel commit: 917d7a2b1d]
|
2019-06-03 18:58:01 -05:00 |
|
Chao Liu
|
60ca6e4bcc
|
tuned
[ROCm/composable_kernel commit: 33c38e2a23]
|
2019-06-03 10:50:25 -05:00 |
|
Chao Liu
|
dcf6a21e9c
|
refactor
[ROCm/composable_kernel commit: 7cdb665a50]
|
2019-06-01 10:48:48 -05:00 |
|
Chao Liu
|
29acbaeb77
|
refactor
[ROCm/composable_kernel commit: 97ba755f2f]
|
2019-05-31 22:28:32 -05:00 |
|
Chao Liu
|
fc219b5877
|
refactor
[ROCm/composable_kernel commit: 8d4607403e]
|
2019-05-31 09:59:00 -05:00 |
|
Chao Liu
|
e4c69380ee
|
refactor
[ROCm/composable_kernel commit: 3a6044aa84]
|
2019-05-31 01:25:14 -05:00 |
|
Chao Liu
|
aa9f3a41ca
|
fix build issue
[ROCm/composable_kernel commit: 9ef124cc15]
|
2019-05-30 21:29:08 -05:00 |
|
Chao Liu
|
3e7d31c326
|
added lds doubl buffer for implicit gemm v4 (nchw, kcyx)
[ROCm/composable_kernel commit: ad5db169b6]
|
2019-05-30 20:55:21 -05:00 |
|
Chao Liu
|
be1b3be83d
|
added implicit gemm v4 (nchw, kcyx)
[ROCm/composable_kernel commit: cb78cc74b7]
|
2019-05-30 20:28:01 -05:00 |
|
Chao Liu
|
db876ea7ec
|
adding implicit gemm v4 (nchw, kcyx)
[ROCm/composable_kernel commit: b2439ec9dd]
|
2019-05-30 17:50:49 -05:00 |
|
Chao Liu
|
fa3cbd975d
|
adding implicit gemm v4 (nchw, kcyx)
[ROCm/composable_kernel commit: 0a2657312e]
|
2019-05-30 00:07:39 -05:00 |
|
Chao Liu
|
b9f18c06d4
|
implicit gemm v3 added lds double buffer
[ROCm/composable_kernel commit: 8c385cf5cf]
|
2019-05-24 17:17:22 -05:00 |
|
Chao Liu
|
8d89b3c212
|
refactor
[ROCm/composable_kernel commit: e17b495dff]
|
2019-05-24 16:43:29 -05:00 |
|
Chao Liu
|
002c0abaf8
|
added implicit GEMM v3
[ROCm/composable_kernel commit: 8fcf3f1e68]
|
2019-05-23 23:54:31 -05:00 |
|
Chao Liu
|
55cd4a3aa0
|
refactor
[ROCm/composable_kernel commit: 8ce1480430]
|
2019-05-23 22:53:35 -05:00 |
|
Chao Liu
|
39b829b919
|
adding implicit gemm v3
[ROCm/composable_kernel commit: 1cc683a3a3]
|
2019-05-23 22:10:40 -05:00 |
|
Chao Liu
|
979dc4da2e
|
adding implicit gemm v3
[ROCm/composable_kernel commit: 8a4b59785b]
|
2019-05-22 19:39:56 -05:00 |
|
Chao Liu
|
c562401581
|
behavior has changed (better and worse), figuring out why
[ROCm/composable_kernel commit: 2a48812edb]
|
2019-05-21 16:43:56 -05:00 |
|
Chao Liu
|
45e1ad4dea
|
adding ConstantMergedTensorDescriptor, refactering ConstantTensorDescriptor, Sequence
[ROCm/composable_kernel commit: acd7082fe1]
|
2019-05-21 16:17:58 -05:00 |
|
Chao Liu
|
2c97c114cf
|
refactor
[ROCm/composable_kernel commit: cd29b09a82]
|
2019-05-19 11:41:00 -05:00 |
|
Chao Liu
|
75df8d78a0
|
rework sequence
[ROCm/composable_kernel commit: a6b95c393b]
|
2019-05-18 23:21:02 -05:00 |
|
Chao Liu
|
4e619d117b
|
rework sequence
[ROCm/composable_kernel commit: df73287b82]
|
2019-05-17 14:56:39 -05:00 |
|
Chao Liu
|
dec8c3ebdd
|
adding implicit gemm v3
[ROCm/composable_kernel commit: 33b5a8556b]
|
2019-05-16 22:23:18 -05:00 |
|
Chao Liu
|
ffd172378a
|
adding implicit gemm v3
[ROCm/composable_kernel commit: 5e5c27a63b]
|
2019-05-16 13:22:40 -05:00 |
|
Chao Liu
|
ac7741cc7c
|
adding implicit gemm v3
[ROCm/composable_kernel commit: b7d052459d]
|
2019-05-15 09:58:17 -05:00 |
|
Chao Liu
|
aeeefc1de3
|
refactored
[ROCm/composable_kernel commit: 4957d5a399]
|
2019-05-02 14:49:20 -05:00 |
|
Chao Liu
|
eb289d06a3
|
remove dead code
[ROCm/composable_kernel commit: 4a99f54c31]
|
2019-05-02 11:09:42 -05:00 |
|
Chao Liu
|
537c363633
|
v1r3 nchw*cyxk=nkhw lds double buffer
[ROCm/composable_kernel commit: d2436a58c4]
|
2019-04-27 15:06:15 -05:00 |
|
Chao Liu
|
ba02fcbfe0
|
fix v1r3 output reorder bug
[ROCm/composable_kernel commit: 63cdc6d2a4]
|
2019-04-27 00:58:49 -05:00 |
|
Chao Liu
|
72f3c44073
|
nchw*cyxk*nkhw on AMD
[ROCm/composable_kernel commit: c138e2126d]
|
2019-04-26 18:03:55 -05:00 |
|
Jing Zhang
|
b6d7e5fd48
|
ds_read_offset
[ROCm/composable_kernel commit: 49d5af1002]
|
2019-04-26 15:55:26 -05:00 |
|
Chao Liu
|
f5b53b6f69
|
debugging ds_read asm
[ROCm/composable_kernel commit: 3ce77700b6]
|
2019-04-26 15:34:55 -05:00 |
|
Chao Liu
|
6bf82635a4
|
fix batch gemm asm bug
[ROCm/composable_kernel commit: b93d2e1b57]
|
2019-04-26 14:40:19 -05:00 |
|
Chao Liu
|
87e3e0dd9a
|
trying gemm asm
[ROCm/composable_kernel commit: 46a0aec185]
|
2019-04-26 10:20:02 -05:00 |
|
Chao Liu
|
04e99df5df
|
tuning on vega 20
[ROCm/composable_kernel commit: 2603bb0fe3]
|
2019-04-25 17:28:59 -05:00 |
|
Chao Liu
|
868068eee1
|
implicit gemm v1r3 nchw_cyxk_nkhw
[ROCm/composable_kernel commit: a903146427]
|
2019-04-25 15:14:39 -05:00 |
|
Chao Liu
|
21988c32b4
|
added implicit gemm v1r3 lds_double_buffer NCHW * CYXK = KNHW, reworked static functionals
[ROCm/composable_kernel commit: 569ad66e2a]
|
2019-04-23 17:51:14 -05:00 |
|
Chao Liu
|
fe17969c81
|
added lds double buffer (on C dimension) for implicit gemm v1r3, as a result, it should achieve 90% of peak for all filter sizes, on CHWN format
[ROCm/composable_kernel commit: 87d8740bf5]
|
2019-04-19 17:49:25 -05:00 |
|
Chao Liu
|
26fab71682
|
added implicit gemm v1r3, refactored decomposition of wei tensor (loop over y, x first, and C second) to allow easy lds double buffer on C
[ROCm/composable_kernel commit: 6d066ede00]
|
2019-04-19 16:46:29 -05:00 |
|