index
:
platform/upstream/caffeonacl
accepted/tizen_5.0_unified
accepted/tizen_5.5_unified
accepted/tizen_5.5_unified_mobile_hotfix
accepted/tizen_5.5_unified_wearable_hotfix
accepted/tizen_unified
armcl-v18.11
master
sandbox/daeinki/armcl-v18.08
sandbox/nmerinov/llvm
tizen
tizen_5.0
tizen_5.5
tizen_5.5_mobile_hotfix
tizen_5.5_tv
tizen_5.5_wearable_hotfix
Domain: Machine Learning / ML Framework; Licenses: BSD-2-Clause;
Inki Dae <inki.dae@samsung.com>
summary
refs
log
tree
commit
diff
log msg
author
committer
range
path:
root
/
src
/
caffe
Age
Commit message (
Expand
)
Author
Files
Lines
2014-07-23
add gflags dependency to caffe.
Yangqing Jia
1
-22
/
+35
2014-07-18
reapply namespace change
Yangqing Jia
35
-94
/
+20
2014-07-18
cosmetics: add syntax = proto2
Yangqing Jia
2
-0
/
+4
2014-07-17
Merge pull request #716 from Yangqing/dev
Jeff Donahue
11
-27
/
+17
2014-07-17
relax benchmark test timing for cheap hardware / CI build
Evan Shelhamer
1
-4
/
+4
2014-07-17
collect CUDA includes and calls, separate from CPU-only mode, leave out
Evan Shelhamer
48
-238
/
+276
2014-07-17
add guards to drop GPU code in CPU-only mode
Evan Shelhamer
3
-80
/
+93
2014-07-17
stub out GPU layer methods to crash loudly in CPU-only mode
Evan Shelhamer
26
-0
/
+102
2014-07-16
move using statements inside namespace caffe to avoid polluting the whole nam...
Yangqing Jia
11
-27
/
+17
2014-07-15
Another bugfix related to my CPU/GPU test changes: make NetTest a
Jeff Donahue
1
-1
/
+1
2014-07-15
Add Net Test to verify correct param_propagate_down behavior.
Jeff Donahue
1
-8
/
+111
2014-07-15
Use Blobs instead of SyncedMemorys for the bias_multiplier_'s.
Jeff Donahue
4
-20
/
+12
2014-07-15
Make ConvolutionLayer and InnerProductLayer abide by param_propagate_down_
Jeff Donahue
4
-79
/
+112
2014-07-15
Add param_propagate_down_ vector to layer, populate according to
Jeff Donahue
1
-1
/
+11
2014-07-15
use layer_param instead of layers_[layer_id]->layer_param()
Jeff Donahue
1
-3
/
+2
2014-07-14
Fix SoftmaxLayerTest: forgot to change this one to use DtypesAndDevices;
Jeff Donahue
1
-8
/
+11
2014-07-14
Seed HingeLossLayerTest; bad values can cause test (and therefore Travis
Jeff Donahue
1
-0
/
+1
2014-07-14
Move test headers to include/.
Jeff Donahue
42
-2240
/
+616
2014-07-10
Replace cudaMemcpy with caffe_gpu_memcpy in SyncedMemory per @longjon
Kai Li
3
-10
/
+6
2014-07-10
Implement @Yangqing's solution to copy memory in the SyncedMemory
Kai Li
1
-4
/
+2
2014-07-10
Switch to GPU mode when pointer is move to or from GPU in SyncedMemory
Kai Li
1
-0
/
+2
2014-07-10
Check the GPU mode to decide which memcpy to use
Kai Li
1
-6
/
+6
2014-07-10
Avoid using cudaMemcpy for memcpy when there is no GPU and CUDA driver
Kai Li
2
-2
/
+12
2014-07-07
Merge pull request #614 from ronghanghu/rectangular_pooling
Jeff Donahue
4
-174
/
+526
2014-07-07
added gradient check for non-square pooling
Ronghang Hu
1
-86
/
+128
2014-07-07
fixed style errors
Ronghang Hu
3
-21
/
+23
2014-07-05
add tests for rectangular pooling regions
Ronghang Hu
4
-55
/
+313
2014-07-05
fixing pooling SetUp() to allow default values for stride and pad
Ronghang Hu
1
-10
/
+10
2014-07-03
Update pooling_layer.cu
Ronghang Hu
1
-53
/
+64
2014-07-03
Update pooling_layer.cpp
Ronghang Hu
1
-24
/
+57
2014-07-03
Update caffe.proto
Ronghang Hu
1
-5
/
+11
2014-07-03
fix casts (static for void*)
Evan Shelhamer
1
-12
/
+12
2014-07-03
reduce caffe_copy to instantiations, split off caffe_memcpy for void*
Evan Shelhamer
4
-28
/
+14
2014-07-03
replace all memset with caffe_set() / caffe_gpu_set()
Evan Shelhamer
8
-28
/
+18
2014-07-03
replace all memcpy by caffe_copy
Evan Shelhamer
21
-80
/
+66
2014-07-03
do all caffe_copy() as UVA mem copy, and drop caffe_gpu_copy()
Evan Shelhamer
1
-8
/
+13
2014-07-03
replace softmax cudaMemcpy with caffe_gpu_copy
Evan Shelhamer
1
-4
/
+2
2014-07-03
switch to unified virtual addressing CUDA memcpy
Evan Shelhamer
8
-16
/
+16
2014-07-03
report UVA in platform test
Evan Shelhamer
1
-0
/
+2
2014-07-03
ConvolutionLayer can take N bottom blobs and N top blobs
Jeff Donahue
3
-130
/
+195
2014-06-29
Merge pull request #545 from jamt9000/im2col-kernel-test
Evan Shelhamer
1
-0
/
+125
2014-06-28
lint
Evan Shelhamer
2
-1
/
+2
2014-06-28
Remove Cuda.major >= 2 check on Dropout test
Sergio
1
-8
/
+0
2014-06-27
Check that pointers are different before copying in caffe_copy and caffe_gpu_...
Sergio
1
-4
/
+12
2014-06-27
Added test to Dropout to check gradients during Test phase
Sergio
2
-1
/
+27
2014-06-27
Fix var names in Dropout.cu
Sergio
1
-1
/
+1
2014-06-27
Modify Dropout to allow backward pass in TEST phase
Sergio
2
-13
/
+19
2014-06-27
Comment-fix.
Rob Hess
1
-1
/
+1
2014-06-27
Update name of last added param.
Rob Hess
1
-1
/
+1
2014-06-27
Add unit test for accuracy layer.
Rob Hess
1
-0
/
+90
[next]