提交 · 4435cd45afcaede9300dbe4b320aa97d048d7c53 · PaddlePaddle / Paddle-Lite

03 6月, 2019 2 次提交

update proposal kernel and other kernels in FPGA v2 track fixed#1676 (#1677) · 4435cd45

由 qnqinan 提交于 6月 03, 2019

* update concat and split kernel and related files in FPGA v2(v3) track

* update

* update

* update kernel and related files in FPGA v2 track

* update

* update

* update kernel and related files for static quantization in FPGA v2 track

* update

* update feed and fetch kernel in FPGA v2 track

* update io file

* update feed fetch and softmax kernel in FPGA v2 track

* update proposal kernel and other kernels in FPGA v2 track

4435cd45

Z

add density_prior_box op (#1674) · f7247e36
由 zp7 提交于 6月 03, 2019

f7247e36

31 5月, 2019 1 次提交
- A
  
  Update README contents (#1667) · b8a14527
  由 adaxi123 提交于 5月 31, 2019
  
  b8a14527
30 5月, 2019 2 次提交
- Z
  add reshape2, transpose2, split GPU operator (#1664) · d1bb62e6
  由 zp7 提交于 5月 30, 2019
```
* add reshape2, transpose2, split GPU operator

* fix reshape2op && delete log
```
  d1bb62e6
- S
  [TEST] Add gesture model for test (#1670) · 737c6041
  由 Shuai Yuan 提交于 5月 30, 2019
```
* [TEST] add gesture model for test
```
  737c6041
29 5月, 2019 1 次提交
- J
  
  fix infershape of flatten op (#1669) · b62964c5
  由 Jiaying Zhao 提交于 5月 29, 2019
  
  b62964c5
28 5月, 2019 2 次提交
- Y
  
  1. lazy load for mpsimage. (#1666) · d6c4c922
  由 Yanzhan Yang 提交于 5月 28, 2019
  
  d6c4c922
- Y
  1. enable conv mps with texture bias. 2. add scale mps. 3. remove duplicate... · fcfeac88
  由 Yanzhan Yang 提交于 5月 28, 2019
```
1. enable conv mps with texture bias. 2. add scale mps. 3. remove duplicate computation in fetch kernel. 4. fix typo. (#1665)
```
  fcfeac88
27 5月, 2019 2 次提交
- Y
  1.enable both tensor and texture as bias of conv_add_relu. 2. add con… (#1662) · c1ffeb30
  由 Yanzhan Yang 提交于 5月 27, 2019
```
* 1.enable both tensor and texture as bias of conv_add_relu. 2. add conv_relu fusion op.

* winograd could be used when. out channel is less than or equal to 16
```
  c1ffeb30
- Z
  
  add fusion_conv_relu op for CPU and GPU_CL · b56e720e
  由 zhaojiaying01 提交于 5月 27, 2019
  
  b56e720e
24 5月, 2019 1 次提交

1.update CPUContext to probe CPU info 2.improve the performance of... · 164da673

由 hong19860320 提交于 5月 24, 2019

1.update CPUContext to probe CPU info 2.improve the performance of SlidingwindowConv3x3s1 and SlidingwindowConv3x3s2 (#1655)

164da673

23 5月, 2019 3 次提交
- Y
  
  1.merge conv, conv add, conv add relu into one implementation. (#1661) · 349857ce
  由 Yanzhan Yang 提交于 5月 23, 2019
  
  349857ce
- J
  
  support element_add in opencl fusion conv op (#1660) · d6c620b9
  由 Jiaying Zhao 提交于 5月 23, 2019
  
  d6c620b9
- J
  fix opencl no persistable bias in fusion conv op (#1657) · 21d8b64b
  由 Jiaying Zhao 提交于 5月 23, 2019
```
* fix opencl no persistable bias in fusion conv op

* delete log
```
  21d8b64b
22 5月, 2019 8 次提交
- J
  
  marker1-anchor update (#1652) · 7d8589c8
  由 jameswu2014 提交于 5月 22, 2019
  
  7d8589c8
- J
  fix cpu conv_add and conv_add_relu (#1654) · 111ca78b
  由 Jiaying Zhao 提交于 5月 22, 2019
```
* fix cpu conv_add and conv_add_relu

* fix cpu conv_add and conv_add_relu
```
  111ca78b
- Y
  1. add optimize parameter to load function. 2. add transpose operation if the... · 66ad725e
  由 Yanzhan Yang 提交于 5月 22, 2019
```
1. add optimize parameter to load function. 2. add transpose operation if the input of fetch op is not [0, 2, 3, 1]. (#1651)
```
  66ad725e
- J
  
  delete test commit, open openmp switch (#1650) · 9e63d954
  由 Jiaying Zhao 提交于 5月 22, 2019
  
  9e63d954
- develop nearestinterp op fix bug in element add close #1647 (#1648) · 9eec8e6b
  由 xiebaiyuan 提交于 5月 22, 2019
```
* develop nearestinterp op  fix bug in element add close #1647

* remove log
remove log
```
  9eec8e6b
- Z
  
  test1 · 0c9d9192
  由 zhaojiaying01 提交于 5月 22, 2019
  
  0c9d9192
- Y
  1. reshape output transpose should be 0, 1, 2, 3. 2.mps conv should be... · 4ef42f5f
  由 Yanzhan Yang 提交于 5月 22, 2019
```
1. reshape output transpose should be 0, 1, 2, 3. 2.mps conv should be disabled when groups is larger than 1. 3. slice should re-tranpose its input if needed. (#1646)
```
  4ef42f5f
- J
  add opencl Scale op (#1645) · 452900fd
  由 Jiaying Zhao 提交于 5月 22, 2019
```
* fix memcpy size in opencl fetch kernel

* add opencl Scale op
```
  452900fd
21 5月, 2019 6 次提交
- Y
  1.add groups to conv3x3. 2.add original size to tensor and tenture and adjust... · 6b32f9ad
  由 Yanzhan Yang 提交于 5月 21, 2019
```
1.add groups to conv3x3. 2.add original size to tensor and tenture and adjust concat layer to handle tensor truncation logic. (#1643)
```
  6b32f9ad
- Add leakyrelu arm cpu op close #1636 (#1638) · 75efc594
  由 xiebaiyuan 提交于 5月 21, 2019
  
  75efc594
- N
  
  method to adjust texture transpose (#1642) · ad098db1
  由 NazgulLee 提交于 5月 21, 2019
  
  ad098db1
- N
  
  add name protection for protobuf-c (#1641) · 6d761d3c
  由 NazgulLee 提交于 5月 21, 2019
  
  6d761d3c
- Y
  
  1.check slice and nearest op result. 2.reshape output should have same layout as its input. (#1640) · ce00384f
  由 Yanzhan Yang 提交于 5月 21, 2019
  
  ce00384f
- J
  opendevice bug fixed for FPGA V1 track (#1635) · 2b5bf193
  由 jameswu2014 提交于 5月 21, 2019
```
* V2-conv-hellocase pass & V1 verify-pass

* V1 paralvl equal to 2

* softmax channel =2 bug

* softmaxbug update2

* io update2

* V1-dwconv relu bug

* opendevice bug fixed
```
  2b5bf193
20 5月, 2019 2 次提交
- Y
  1.add flatten2, slice and nearest_interp op. 2.truncate tensor with more than... · 20f5ae65
  由 Yanzhan Yang 提交于 5月 20, 2019
```
1.add flatten2, slice and nearest_interp op. 2.truncate tensor with more than 4 dimensions to 4. (#1634)
```
  20f5ae65
- Y
  
  add exp, sigmoid and leaky_relu op (#1633) · 1b7329e2
  由 Yanzhan Yang 提交于 5月 20, 2019
  
  1b7329e2
18 5月, 2019 1 次提交
- Y
  1.parse tensor desc info from params file. 2.ensure tensor desc is consistent... · ef827162
  由 Yanzhan Yang 提交于 5月 18, 2019
```
1.parse tensor desc info from params file. 2.ensure tensor desc is consistent between params file and model file. (#1631)
```
  ef827162
17 5月, 2019 2 次提交
- J
  
  fix memcpy size in opencl fetch kernel (#1630) · 5945175f
  由 Jiaying Zhao 提交于 5月 17, 2019
  
  5945175f
- add notice for android gpu doc (#1627) · b05774ca
  由 xiebaiyuan 提交于 5月 17, 2019
  
  b05774ca
16 5月, 2019 3 次提交
- N
  
  add reshape2, transpose2, relu6 and scale operator (#1628) · 857ffcbc
  由 NazgulLee 提交于 5月 16, 2019
  
  857ffcbc
- X
  
  fix crash when close optimise in gpu test close #1626 · 7deff6de
  由 xiebaiyuan 提交于 5月 16, 2019
  
  7deff6de
- J
  V1-dwconv relu bug for FPGA track (#1624) · 43e040dc
  由 jameswu2014 提交于 5月 16, 2019
```
* V2-conv-hellocase pass & V1 verify-pass

* V1 paralvl equal to 2

* softmax channel =2 bug

* softmaxbug update2

* io update2

* V1-dwconv relu bug
```
  43e040dc
15 5月, 2019 2 次提交

Y

add nslock to load and clear method of runner (#1623) · fceff4ff
由 Yanzhan Yang 提交于 5月 15, 2019

fceff4ff

remove dafault opencl kernel path (#1622) · 17fa2ebd

由 Jiaying Zhao 提交于 5月 15, 2019

* optimize GPU conv performance and structure

* add CL macro for test_conv_gpu

* fix build failure

* change funtion name

* change funtion name

* fix load_when_predict mode in opencl

* remove dafault opencl kernel path

17fa2ebd

13 5月, 2019 2 次提交

fix load_when_predict mode in opencl (#1619) · 4f7e31d9

由 Jiaying Zhao 提交于 5月 13, 2019

* optimize GPU conv performance and structure

* add CL macro for test_conv_gpu

* fix build failure

* change funtion name

* change funtion name

* fix load_when_predict mode in opencl

4f7e31d9

update feed fetch and softmax kernel in FPGA v2 track fixed#1615 (#1616) · b942f2de

由 qnqinan 提交于 5月 13, 2019

* update concat and split kernel and related files in FPGA v2(v3) track

* update

* update

* update kernel and related files in FPGA v2 track

* update

* update

* update kernel and related files for static quantization in FPGA v2 track

* update

* update feed and fetch kernel in FPGA v2 track

* update io file

* update feed fetch and softmax kernel in FPGA v2 track

b942f2de