Megvii Engine Team
|
086ee045ba
|
fix(mge): fix error when transform model to nchw4/32/64 tensor format
GitOrigin-RevId: 34be9c7844
|
2 years ago |
Megvii Engine Team
|
21a975f8fd
|
feat(gopt): add channel padding pass when enable nchw44/nchw88/nchw44-dot
GitOrigin-RevId: fbe0516ffc
|
2 years ago |
Megvii Engine Team
|
b82e8f007c
|
refactor(gopt): refact the padding channel opt pass
GitOrigin-RevId: ee3f55aa66
|
2 years ago |
Megvii Engine Team
|
fac67e7c2b
|
feat(gopt): support nchw44 global pooling with fuse_grain
GitOrigin-RevId: 4c43a149f8
|
2 years ago |
Megvii Engine Team
|
c2e9860feb
|
chore(license): remove all license in file header
GitOrigin-RevId: a0e31247a6
|
3 years ago |
Megvii Engine Team
|
87de704a46
|
feat(gopt): fuse conv h_swish
GitOrigin-RevId: a3d12991fb
|
3 years ago |
Megvii Engine Team
|
3726f5cc92
|
feat(gopt): merger consecutive relayout and dimshuffle to one relayout to optimize CD4 performarce
GitOrigin-RevId: a058776be3
|
3 years ago |
Megvii Engine Team
|
1fead9b6b0
|
feat(gopt): merge consecutive dimshuffle and relayout to one relayout to optimize CD4 performace
GitOrigin-RevId: 16f22baa80
|
3 years ago |
Megvii Engine Team
|
26d1e4f7ed
|
feat(gopt): optimize cd4 pass rule for elemwise and typecvt to let cd4 start as soon as possible
GitOrigin-RevId: 6580dedca7
|
3 years ago |
Megvii Engine Team
|
5f4501e0f3
|
fix(gopt): fix conv bias fuse 2 noline
GitOrigin-RevId: a6ab9f4e5e
|
3 years ago |
Megvii Engine Team
|
e715423f20
|
feat(src/gopt): add optpass on arm for fusing typecvt and elemwise to elemwise multi type
GitOrigin-RevId: e6bcbbf91b
|
3 years ago |
Megvii Engine Team
|
34773ba37b
|
fix(mgb/gopt): tensorcore pass replace BatchConvBias inputs to nchw4
GitOrigin-RevId: 3ff3c422fb
|
3 years ago |
Megvii Engine Team
|
369c2ccc5a
|
style(all): reformat c++ code
GitOrigin-RevId: 3ffd1b211f
|
3 years ago |
Megvii Engine Team
|
88c1eedbd7
|
feat(mgb/gopt): enable reduce for nchw44
GitOrigin-RevId: fce59d0762
|
3 years ago |
Megvii Engine Team
|
c0ccd0ea7e
|
feat(mge/bn): add NHWC support for bn
GitOrigin-RevId: 0a5bb6f72d
|
3 years ago |
Megvii Engine Team
|
d7cc4628f6
|
perf(gopt): opt concat for OpenCL
GitOrigin-RevId: 9bb226d4b1
|
3 years ago |
Megvii Engine Team
|
a3cd3fc74f
|
test(mgb/gopt): add testcase for global layout transform
GitOrigin-RevId: f9669e1ba0
|
3 years ago |
Megvii Engine Team
|
3eb0505f9b
|
feat(imperative): add support for quantized conv transpose2d
GitOrigin-RevId: ffd6431299
|
3 years ago |
Megvii Engine Team
|
869a03271b
|
perf(mgb): disable FoldingConvBiasDimshufflePass in cuda10 for performance
GitOrigin-RevId: d1b95a6f01
|
3 years ago |
Megvii Engine Team
|
239916a997
|
fix(mgb/gopt): fix testcase for enable nchw64 pass
GitOrigin-RevId: 2ae8d1608d
|
4 years ago |
Megvii Engine Team
|
009c90a2fe
|
feat(mgb/gopt): modify padding policy for 4bit conv bias oprs
GitOrigin-RevId: 188a2c3728
|
4 years ago |
Megvii Engine Team
|
b4687ce8da
|
feat(dnn/cuda): add convolution with i8 input and u4 output
GitOrigin-RevId: 8be439abf1
|
4 years ago |
Megvii Engine Team
|
bba04f02e5
|
feat(mgb/gopt): add fusion support for conv, astype(s4) and reformat
GitOrigin-RevId: 6329ca2c5f
|
4 years ago |
Megvii Engine Team
|
7d3df995cb
|
feat(gopt/inference): allow Float32 output dtype in EnableNCHW4Pass
GitOrigin-RevId: 81100dbaf7
|
4 years ago |
Megvii Engine Team
|
47dcdf3e17
|
fix(mgb/core): fix dtype and resize modifiers for tensor
GitOrigin-RevId: a9d95a4cd8
|
4 years ago |
Megvii Engine Team
|
0fb9cc41e4
|
fix(gopt): fix nchw64 opt pass
GitOrigin-RevId: dec18d1ab1
|
4 years ago |
Megvii Engine Team
|
86b69cacd0
|
fix(dnn): fixes for int4
GitOrigin-RevId: 845e164fd3
|
4 years ago |
Megvii Engine Team
|
8da2f698a3
|
feat(dnn/cuda): support warp perspective/pooling op when channel not aligned to 64
GitOrigin-RevId: 39f29ec990
|
4 years ago |
Megvii Engine Team
|
ae6ff2c5a6
|
feat(mgb/gopt): add opt pass for nchw64 layout transform
GitOrigin-RevId: adede7cef6
|
4 years ago |
Megvii Engine Team
|
63a9bd30a8
|
feat(mgb/gopt): add an opt pass for padding channels to enable fast int8/int4 support on GPU
GitOrigin-RevId: 94c719bb5c
|
4 years ago |
Megvii Engine Team
|
36b1ba052f
|
fix(mgb/dnn): fix cudnn8.0.4 convbias with z
GitOrigin-RevId: 09453d8a12
|
4 years ago |
Megvii Engine Team
|
2d18074a70
|
fix(mgb): fix spell error
GitOrigin-RevId: acae00e0a5
|
4 years ago |
Megvii Engine Team
|
a437ec8e88
|
fix(src/gopt): add replace func of typecvt opr for nhwcd4 pass
GitOrigin-RevId: 801eb1dab3
|
4 years ago |
Megvii Engine Team
|
04b1a45af4
|
fix(dnn): fix cudnn crash when finalize called after cudnn dtor
GitOrigin-RevId: b0ad639921
|
4 years ago |
Megvii Engine Team
|
14a089c49d
|
fix(dnn): change ci to cudnn804, reopen testcase
GitOrigin-RevId: 90713a801b
|
4 years ago |
Megvii Engine Team
|
ba2ad46e54
|
feat(gopt): add deconv nchw4 int8 opt pass, add deconv nchw int8
GitOrigin-RevId: c0530a949e
|
4 years ago |
Megvii Engine Team
|
a3ea1f153c
|
feat(mgb/opr): add fast profile and combined Execution strategy
GitOrigin-RevId: 843dc3a790
|
4 years ago |
Megvii Engine Team
|
c82d88751a
|
fix(dnn/cuda): add cuda nchw int8 conv impl with nchw4 to fix cu111 compatibility
GitOrigin-RevId: 771968f9ac
|
4 years ago |
Megvii Engine Team
|
51868533c8
|
fix(mgb/gopt): fix opt pass elementwise operation shape issue at tranform to NCHW4
GitOrigin-RevId: c0c4e3f82e
|
4 years ago |
Megvii Engine Team
|
cf27dd642c
|
fix(cuda): use cudnn8.0.4 as cu111 default libs
GitOrigin-RevId: 721ca73bae
|
4 years ago |
Megvii Engine Team
|
649e4dd750
|
test(cuda): fix test for cu111
GitOrigin-RevId: 04fe5eb23f
|
4 years ago |
Megvii Engine Team
|
2e4b9a42f7
|
fix(mgb/gopt): fix folding conv dimshuffle opt pass
GitOrigin-RevId: 878b7de9de
|
4 years ago |
Megvii Engine Team
|
364afec033
|
chore(mge): update copyright years
GitOrigin-RevId: 3c0690bcc1
|
4 years ago |
Megvii Engine Team
|
1b1ad56a82
|
fix(mgb/gopt): fix warp fusion opt pass
GitOrigin-RevId: a40bbcd719
|
4 years ago |
Megvii Engine Team
|
4e9be159f7
|
feat(mgb/gopt): add opt pass for fusing convolution and reformat
GitOrigin-RevId: d0c5deace2
|
4 years ago |
Megvii Engine Team
|
61f917fb8e
|
feat(dnn/cuda): add impl for fusing warp perspective and dimshuffle
GitOrigin-RevId: 51e025973f
|
4 years ago |
Megvii Engine Team
|
3bf73ff16f
|
feat(dnn): add cuda preprocess fusion
GitOrigin-RevId: d789c99e59
|
4 years ago |
Megvii Engine Team
|
5f171298aa
|
feat(mgb/gopt): add AxisAddRemove opr support for cd4 opt pass
GitOrigin-RevId: 85218ee0c4
|
4 years ago |
Megvii Engine Team
|
6f5d0febf1
|
perf(dnn/cuda): enhance performance for pooling forward
GitOrigin-RevId: 55fb2a9b25
|
4 years ago |
Megvii Engine Team
|
7cd71c3102
|
fix(mgb/gopt): fix cd4 elewise transform
GitOrigin-RevId: 027d5e53e4
|
4 years ago |