Luwei Zhou
fa38c36b2f
Update ONEDNN version to fix AVX2 bug.
2022-05-26 19:59:11 +08:00
Li, Tingqian
c533aa6b8b
fix cpplint
2022-05-26 07:32:18 -04:00
Li, Tingqian
5b6827649c
fix binary prelu post Ops
2022-05-26 06:21:55 -04:00
Li, Tingqian
f037208b3f
Generate exec graph in cpu dump check tool
2022-05-26 05:52:41 -04:00
Li, Tingqian
a1ccf115df
Add verbose log
2022-05-26 02:46:15 -04:00
Li, Tingqian
8b7f7d6391
Add CPU dump check tool
2022-05-25 21:07:02 -04:00
Zhang Yi3
6a8b37fe4f
[CPU] Add BF16 AMX test for Matmul
2022-05-25 17:28:57 +08:00
Luo Cheng
f326f0b824
make nightly case run. tested on amx/avx512/avx2.
2022-05-25 13:25:56 +08:00
Luo Cheng
ecdda2c909
mix legacy/new binary postops
2022-05-24 20:21:42 +08:00
Zhang Yi3
c9304c5c7f
[WA] skip fakequantize fusing in bf16
2022-05-24 17:26:46 +08:00
Luwei Zhou
04ba093dca
Compiling issue fix.
2022-05-23 16:01:29 +08:00
Luwei Zhou
d81896502f
WR enforce reorder bug and add NSPC into deconv supported list.
2022-05-23 14:53:35 +08:00
Luo Cheng
26a20e5875
testcase for conv brgconv avx512/amx
2022-05-23 09:48:50 +08:00
Luo Cheng
d7d4097418
testcase for conv brgconv avx512/amx
2022-05-22 21:10:08 +08:00
Luo Cheng
91ec6b0806
testcase init support amx check
2022-05-20 22:30:27 +08:00
Luo Cheng
841b2dc47a
disable BF16 fusing fakequant testcase
2022-05-20 20:57:30 +08:00
Luo Cheng
b53bc6bac5
MemoryInput/Output support bf16; Enforce bf16 'NO' should enable snipptes
2022-05-20 20:01:19 +08:00
Li, Tingqian
f5af87f810
Fix cpplint error
2022-05-20 11:59:55 +03:00
Li, Tingqian
c4021f8e6e
Fix primType check issue
2022-05-20 11:15:55 +03:00
Luo Cheng
a1f21f3797
fix clang check
2022-05-19 20:12:50 +08:00
Luo Cheng
b08bf8a9de
testcase subgraph sets default ENFORCE_BF16 to NO
2022-05-19 19:55:35 +08:00
Luo Cheng
cecd11457e
OVClassBasicTest case typo
2022-05-19 19:55:35 +08:00
Luo Cheng
d9d934bf86
[WA] bf16 crash due to MemoryInput/Output
2022-05-19 19:55:35 +08:00
Li, Tingqian
c302bc94da
Add cpuDebugFuncTests target
2022-05-19 03:59:19 +03:00
Li, Tingqian
c20d762af8
Fix ConcatConvSumInPlaceTest
2022-05-19 02:55:58 +03:00
Luo Cheng
f5ea549d97
fix gemm bf16 fail
2022-05-18 10:40:35 +08:00
Luo Cheng
b2ba3d5055
add gemm int8 binary postops to fix GroupConvolutionQDqTransformation fail
2022-05-18 08:08:49 +08:00
Luo Cheng
a9104e1a88
add gemm int8 binary postops to fix GroupConvolutionQDqTransformation fail
2022-05-17 19:51:24 +08:00
Luwei Zhou
fbf241d9d8
WR to disable the LPT multiplyToGroupConv test because the transformation was disabled in d5e16f
2022-05-16 11:27:33 +08:00
Zhang Yi3
63b283fe88
[WA] remove invalid FQ tests
2022-05-16 15:43:26 +08:00
Luo Cheng
167f74a7bc
fix avx2 groupconv accuracy problem
2022-05-16 07:24:33 +08:00
Luo Cheng
36be40c7e9
fix gemm bf16 win crash
2022-05-14 11:58:25 +08:00
Luo Cheng
543963bf8d
test subgraph added nhwc format check
2022-05-13 19:07:23 +08:00
Luo Cheng
901210fa6c
testcase conv maxpool will check brgconv instead of jit
2022-05-13 18:50:39 +08:00
Luo Cheng
eb87aacc49
[WA] Remove the moc fail case by #af4731a1
2022-05-13 18:17:10 +08:00
Luo Cheng
a56e1a9c4b
group deconv may crash on memory out of bound
2022-05-13 17:26:16 +08:00
Luo Cheng
11238a504c
Merge pull request #43 from liubo-intel/liubo/onednn_2.6_bugfix
...
fix CPU 'ConvolutionBackpropDataLayerTest' fail issue
2022-05-13 13:59:24 +08:00
Luwei Zhou
6955c389d6
[WR] Removed failed the ReduceMean tests caused by 21f3555.
2022-05-13 11:42:38 +08:00
liubo-intel
9e9e3dd01b
reopen 'FuseDeconvolutionAndSimpleOperation' Transform to fix CPU 'ConvolutionBackpropDataLayerTest' fail issue
2022-05-13 11:31:21 +08:00
Zhang Yi3
3bfc042a35
fix xmm zero check
2022-05-12 23:52:31 +08:00
Luo Cheng
4696b9e58a
[WA] make cpu case to run completed
2022-05-12 15:35:16 +08:00
Luo Cheng
e17179d795
Merge remote-tracking branch 'upstream/master' into luocheng/onednn_2.6
2022-05-06 13:38:39 +08:00
yanlan song
912f40e74d
stateful inferface impl for AUTO/HETERO ( #11590 )
...
* CPU for stateful model
Signed-off-by: fishbell <bell.song@intel.com >
* log
Signed-off-by: fishbell <bell.song@intel.com >
* hetero impl
Signed-off-by: fishbell <bell.song@intel.com >
* enable tests
Signed-off-by: fishbell <bell.song@intel.com >
2022-05-06 12:43:56 +08:00
guozhong wang
870455675c
add cumulative_throughput for python ( #11195 )
2022-05-06 12:43:41 +08:00
guozhong wang
0a65f5f607
selectdevice returns MULTI:device in cumulative_throughput ( #11367 )
...
* selectdevice returns MULTI:device in cumulative_throughput
* load multi with throughput and disable cpu helper in cumulative
* disable cpu helper in cumulative_throughput
* add cumulative to bechmark_app help message
* modify benchmark_app.hpp clang-format
2022-05-06 12:42:59 +08:00
Andrew Kwangwoong Park
caccde6a82
[GPU] Add TC for MXNet-style NMS model(decrease_label_id) ( #10624 )
...
- Add TC for decrease_label_id=true to cover MXNet-style NMS models
- Fix segfault issue that occurs when data precision is fp16
Signed-off-by: Andrew Kwangwoong Park <andrew.kwangwoong.park@intel.com >
Signed-off-by: Andrew Park <andrew.park@intel.com >
2022-05-06 10:36:46 +09:00
Artur Kulikowski
1331eabbed
Getting of python version depend on the used shell ( #11614 )
2022-05-05 23:42:54 +02:00
Katarzyna Mitrus
50287625d7
[Eye-9] Python API for Eye-9 ( #11552 )
2022-05-05 16:16:36 +02:00
mei, yang
2d0ffd8fe5
Specify GenerateProposals-9 ( #11004 )
2022-05-05 10:28:18 +02:00
Bo Liu
d560cf19a3
Specify ROIAlign-9 ( #11067 )
2022-05-05 10:27:58 +02:00
