yeyunpeng2020
|
813d12fac6
|
post training support kirin quant
|
4 years ago |
yeyunpeng2020
|
3afdd358b6
|
QuantStrategy add set support op list.
|
4 years ago |
yeyunpeng2020
|
a2b2bd516f
|
fix matmul int8 bug && support activation symmetry
|
4 years ago |
i-robot
|
c1b1324ca4
|
!26815 310 ocr model support dynamic hw
Merge pull request !26815 from zhengyuanhua/model_br1
|
4 years ago |
i-robot
|
9ed9f21077
|
!26771 [MS][LITE][develop] template DelegateModel and modify c api
Merge pull request !26771 from sunsuodong/delegate_model_and_fix_c_api
|
4 years ago |
i-robot
|
3a5487c1bb
|
!26818 activation support asymmetry & weight support symmetry
Merge pull request !26818 from yeyunpeng2020/quant_bak
|
4 years ago |
yeyunpeng2020
|
1b3d4de742
|
activation support asymmetry & weight support symmetry
|
4 years ago |
zhengyuanhua
|
9f1c444178
|
310 ocr suport dynamic hw
|
4 years ago |
i-robot
|
a2403caa3c
|
!26797 [MSLITE] bechmark supoort mpirun
Merge pull request !26797 from ling/op
|
4 years ago |
ling
|
da403ad1a3
|
[MSLITE] bechmark test
|
4 years ago |
sunsuodong
|
ef80c94aac
|
template DelegateModel and fix c api
|
4 years ago |
i-robot
|
13a2d09e1a
|
!26661 310 support dynamic image size
Merge pull request !26661 from zhengyuanhua/model_br1
|
4 years ago |
i-robot
|
c2af5258cc
|
!26769 optimize the Quantizer base class
Merge pull request !26769 from yeyunpeng2020/quant_bak2
|
4 years ago |
i-robot
|
d0792507bd
|
!26736 [MS][LITE]support parameter cache and distribution predict
Merge pull request !26736 from 张学同/distribution_cache_dev
|
4 years ago |
yeyunpeng2020
|
8aaff90f7e
|
optimize the Quantizer base class and optimize min_quant_weight_channel strategy
|
4 years ago |
i-robot
|
289ff6518b
|
!26681 [MS][LITE] fix npu compatibility bugs and strided_slice-transpose decrease bugs
Merge pull request !26681 from XianglongZeng/myms_new
|
4 years ago |
zhangxuetong
|
9ef0fa84bb
|
support parameter cache and distribution predict
|
4 years ago |
zhengyuanhua
|
c0332b4fc7
|
310 support image size
|
4 years ago |
yeyunpeng2020
|
35e0526b51
|
full quant support skip node
|
4 years ago |
zengxianglong
|
229b94e8a7
|
fix npu compatibility bugs and strided_slice-transpose decrease bugs
|
4 years ago |
i-robot
|
354b7d44df
|
!26703 fix converter input shape
Merge pull request !26703 from yeyunpeng2020/converter
|
4 years ago |
yeyunpeng2020
|
dc44f76ac2
|
fix converter input shape
|
4 years ago |
yeyunpeng2020
|
dd1f9eec30
|
Calibrator support setting symmetry
|
4 years ago |
i-robot
|
2e7c2ef782
|
!26653 [MSLITE][DEVELOP] add onnx hard sigmoid parser
Merge pull request !26653 from yangruoqi713/master
|
4 years ago |
i-robot
|
8d38efb2c4
|
!26657 abstract QuantStrategy && optimize full quant prepare
Merge pull request !26657 from yeyunpeng2020/quant_bak
|
4 years ago |
yeyunpeng2020
|
ad53cd005a
|
abstract QuantStrategy && optimize full quant prepare
|
4 years ago |
i-robot
|
6b6bdf5d1d
|
!26507 [MSLITE] Add scale fusion.
Merge pull request !26507 from wangshaocong/scale_fusion
|
4 years ago |
yangruoqi713
|
b55e565f27
|
[MSLITE][DEVELOP] add onnx hard sigmoid parser
|
4 years ago |
i-robot
|
ddc5399ca7
|
!26617 [MS][LITE]infershape support non tail call
Merge pull request !26617 from mengyuanli/non_tail_call
|
4 years ago |
mengyuanli
|
36e9a0deb7
|
modify infershape to support non tail call
|
4 years ago |
wang_shaocong
|
ba0d056735
|
[MSLITE] Add scale_scale fusion
|
4 years ago |
yeyunpeng2020
|
3b33952146
|
fix image preprocess bug && remove unused param
|
4 years ago |
i-robot
|
e8226f8f4f
|
!26578 Optimize full quantization code structure
Merge pull request !26578 from yeyunpeng2020/full_quant
|
4 years ago |
yeyunpeng2020
|
730678e7ab
|
Optimize full quantization code structure
|
4 years ago |
i-robot
|
62ad43e660
|
!26488 cpplint check
Merge pull request !26488 from wangyanling/cpplint
|
4 years ago |
i-robot
|
a148442099
|
!26509 post training bind thread
Merge pull request !26509 from yeyunpeng2020/quant
|
4 years ago |
i-robot
|
a7ac06071a
|
!26425 [MSLITE] Add scale activation fusion.
Merge pull request !26425 from wangshaocong/scale_act_fusion
|
4 years ago |
yeyunpeng2020
|
a107abd292
|
post training quant bind thread
|
4 years ago |
wangyanling
|
ca2c686ccb
|
cpplint check
|
4 years ago |
i-robot
|
d936efbb39
|
!26476 optimize auto_tune and mixed bit quant
Merge pull request !26476 from yeyunpeng2020/debug
|
4 years ago |
yeyunpeng2020
|
e89d048199
|
converter mindir support inputshape
|
4 years ago |
yeyunpeng2020
|
b443593c65
|
optimize statistic time
|
4 years ago |
i-robot
|
a826b14d05
|
!26427 benchmark support cos-sim
Merge pull request !26427 from yeyunpeng2020/benchmark
|
4 years ago |
i-robot
|
8268eb3068
|
!26465 [MS][LITE][ToD] Fix Fp16 bug
Merge pull request !26465 from Nizzan/export_nizzan
|
4 years ago |
nizzan
|
60195d9dc9
|
Fix Fp16 bug
|
4 years ago |
i-robot
|
0e6264e036
|
!26303 adjust div op and parameter format for ocr precision
Merge pull request !26303 from zhengyuanhua/model_br1
|
4 years ago |
yeyunpeng2020
|
7c8abe6004
|
benchmark support cos-sim
|
4 years ago |
wang_shaocong
|
5531a50bf7
|
[MSLITE] Add scale_act fusion
|
4 years ago |
i-robot
|
77367282be
|
!26243 [MSLITE] Complete mul_add fusion
Merge pull request !26243 from wangshaocong/mul_add_fusion
|
4 years ago |
i-robot
|
f5865d0ea1
|
!26333 [MSLITE] add get rank id in tensorrt delegate
Merge pull request !26333 from Liu_Xuu/trt_1112_rankid
|
4 years ago |