platform/upstream/caffeonacl.git
8 years agotranpose parameter added to IP layer to support tied weights in an autoencoder. Argum...
Youssef Kashef [Fri, 29 Jan 2016 18:21:48 +0000 (19:21 +0100)]
tranpose parameter added to IP layer to support tied weights in an autoencoder. Arguments to matrix multiplication function are conditioned on this parameter, no actual transposing takes place.

test ip gradient computation with transpose on

8 years agoMerge pull request #3545 from gdh1995/relative-path-for-soft-links
Jeff Donahue [Fri, 29 Jan 2016 08:59:23 +0000 (00:59 -0800)]
Merge pull request #3545 from gdh1995/relative-path-for-soft-links

use relative paths on making build/tools/ links

8 years agouse relative paths on making build/tools/ links
gdh1995 [Wed, 13 Jan 2016 10:20:41 +0000 (18:20 +0800)]
use relative paths on making build/tools/ links

The old uses `abspath`, which I think is so harmful:
* If I `cp -a` the whole project, `build/tools/caffe` still refer to
    the old file, until `make clean`, making debugging very hard
* For `tar` and `scp`, the soft links can not work
    unless the target project folder has the same path

8 years agoMerge pull request #2917 from ajkl/patch-2
Jeff Donahue [Wed, 27 Jan 2016 18:43:29 +0000 (10:43 -0800)]
Merge pull request #2917 from ajkl/patch-2

fixing the database param

8 years agoMerge pull request #3022 from jeffdonahue/expose-param-display-names
Jeff Donahue [Wed, 27 Jan 2016 18:37:48 +0000 (10:37 -0800)]
Merge pull request #3022 from jeffdonahue/expose-param-display-names

Net: expose param_display_names_

8 years agoMerge pull request #3605 from jeffdonahue/bias-gemm-type-fix
Jeff Donahue [Wed, 27 Jan 2016 18:34:20 +0000 (10:34 -0800)]
Merge pull request #3605 from jeffdonahue/bias-gemm-type-fix

Remove incorrect cast of gemm int arg to Dtype in BiasLayer

8 years agoRemove incorrect cast of gemm int arg to Dtype in BiasLayer
Jeff Donahue [Wed, 27 Jan 2016 07:09:27 +0000 (23:09 -0800)]
Remove incorrect cast of gemm int arg to Dtype in BiasLayer

8 years agoMerge pull request #3591 from jeffdonahue/scale-bias-layer
Evan Shelhamer [Wed, 27 Jan 2016 01:43:14 +0000 (17:43 -0800)]
Merge pull request #3591 from jeffdonahue/scale-bias-layer

Scale and Bias Layers

8 years agoMerge pull request #2651 from keir/master
Jeff Donahue [Tue, 26 Jan 2016 23:34:42 +0000 (15:34 -0800)]
Merge pull request #2651 from keir/master

8 years agoMake the two separate build systems clearer in the documentation
Keir Mierle [Fri, 26 Jun 2015 07:10:21 +0000 (00:10 -0700)]
Make the two separate build systems clearer in the documentation

8 years agoMerge pull request #3602 from jeffdonahue/rm-cuda-props
Jeff Donahue [Tue, 26 Jan 2016 23:20:07 +0000 (15:20 -0800)]
Merge pull request #3602 from jeffdonahue/rm-cuda-props

Remove unnecessary CAFFE_TEST_CUDA_PROP declarations

8 years agoMerge pull request #2144 from tishibas/load_image-improved
Jeff Donahue [Tue, 26 Jan 2016 23:17:00 +0000 (15:17 -0800)]
Merge pull request #2144 from tishibas/load_image-improved

improved to load RGB image as grayscale image

8 years agoMerge pull request #2810 from madan-ram/patch-2
Jeff Donahue [Tue, 26 Jan 2016 23:07:43 +0000 (15:07 -0800)]
Merge pull request #2810 from madan-ram/patch-2

8 years agoUpdate mnist readme.md: scale moved to transform_param
Madan Ram [Thu, 23 Jul 2015 11:12:15 +0000 (16:42 +0530)]
Update mnist readme.md: scale moved to transform_param

8 years agoMerge pull request #3297 from sjbrown/patch-1
Jeff Donahue [Tue, 26 Jan 2016 22:32:39 +0000 (14:32 -0800)]
Merge pull request #3297 from sjbrown/patch-1

Add a -c to wget so that it continues interrupted downloads

8 years agoMerge pull request #3132 from bwilbertz/LastIterationLoss
Jeff Donahue [Tue, 26 Jan 2016 22:27:27 +0000 (14:27 -0800)]
Merge pull request #3132 from bwilbertz/LastIterationLoss

Fix loss of last iteration when average_loss > 1

8 years agoRemove unnecessary CAFFE_TEST_CUDA_PROP declarations
Jeff Donahue [Tue, 26 Jan 2016 21:58:58 +0000 (13:58 -0800)]
Remove unnecessary CAFFE_TEST_CUDA_PROP declarations

8 years agoMerge pull request #3393 from kkhoot/fix-reshape
Jeff Donahue [Tue, 26 Jan 2016 18:01:34 +0000 (10:01 -0800)]
Merge pull request #3393 from kkhoot/fix-reshape

8 years agoPrevent in-place computation in ReshapeLayer and FlattenLayer
Kang Kim [Fri, 27 Nov 2015 12:57:51 +0000 (21:57 +0900)]
Prevent in-place computation in ReshapeLayer and FlattenLayer

8 years agoMerge pull request #3496 from jeffdonahue/fix-testdatatransformer-leaks
Jeff Donahue [Tue, 26 Jan 2016 17:55:08 +0000 (09:55 -0800)]
Merge pull request #3496 from jeffdonahue/fix-testdatatransformer-leaks

TestDataTransformer: fix some memory leaks

8 years agoMerge pull request #3581 from Austriker/draw-python3
Jeff Donahue [Tue, 26 Jan 2016 17:38:20 +0000 (09:38 -0800)]
Merge pull request #3581 from Austriker/draw-python3

Updated draw.py to make it work under Python 3

8 years agoUpdated import to make it work with pydotplus
Hugo Serrat [Thu, 21 Jan 2016 13:34:01 +0000 (14:34 +0100)]
Updated import to make it work with pydotplus

8 years agoMerge pull request #3593 from ronghanghu/matcaffe-versioning
Ronghang Hu [Sat, 23 Jan 2016 20:15:46 +0000 (12:15 -0800)]
Merge pull request #3593 from ronghanghu/matcaffe-versioning

show Caffe's version from MatCaffe

8 years agoshow Caffe's version from MatCaffe
Ronghang Hu [Sat, 23 Jan 2016 09:22:04 +0000 (01:22 -0800)]
show Caffe's version from MatCaffe

8 years agoSeparation and generalization of ChannelwiseAffineLayer into BiasLayer
Jeff Donahue [Fri, 22 Jan 2016 23:57:47 +0000 (15:57 -0800)]
Separation and generalization of ChannelwiseAffineLayer into BiasLayer
and ScaleLayer.  The behavior of ChannelwiseAffineLayer can be
reproduced by a ScaleLayer with `scale_param { bias_term: true }`.

BiasLayer and ScaleLayer each take 1 or 2 bottoms, with the output having
the same shape as the first.  The second input -- either another  bottom or a
learned parameter -- will have its axes (virtually) broadcast and tiled to have
the same shape as the first, after which elementwise addition (Bias) or
multiplication (Scale) is performed.

8 years agoMerge pull request #3311 from lukeyeager/bvlc/versioning
Evan Shelhamer [Sat, 23 Jan 2016 02:54:17 +0000 (18:54 -0800)]
Merge pull request #3311 from lukeyeager/bvlc/versioning

Add versioning to build and mark version 1.0.0-rc3

8 years agoMerge pull request #3584 from graphaelli/python3-config
Philipp Krähenbühl [Sat, 23 Jan 2016 00:59:01 +0000 (16:59 -0800)]
Merge pull request #3584 from graphaelli/python3-config

Add makefile config option for linking Python 3 libraries

8 years agoVersion 1.0.0-rc3
Luke Yeager [Fri, 22 Jan 2016 23:30:35 +0000 (15:30 -0800)]
Version 1.0.0-rc3

8 years agoAdd ChannelwiseAffine for batch norm
Dmytro Mishkin [Thu, 14 Jan 2016 15:10:11 +0000 (17:10 +0200)]
Add ChannelwiseAffine for batch norm

8 years agoMerge pull request #3388 from mohomran/exponential_linear_units
Evan Shelhamer [Fri, 22 Jan 2016 18:49:48 +0000 (10:49 -0800)]
Merge pull request #3388 from mohomran/exponential_linear_units

Exponential Linear Units

8 years agoAdd makefile config option for linking Python 3 libraries
Robbie Cooper [Thu, 21 Jan 2016 19:11:00 +0000 (14:11 -0500)]
Add makefile config option for linking Python 3 libraries

8 years agoMerge pull request #3536 from intelcaffe/im2col-speedup
Jon Long [Wed, 20 Jan 2016 21:38:10 +0000 (13:38 -0800)]
Merge pull request #3536 from intelcaffe/im2col-speedup

Performance related update of im2col() and col2im() functions

8 years agoMerge pull request #3574 from tahtguymike/maxpool_bug
Evan Shelhamer [Wed, 20 Jan 2016 18:09:59 +0000 (10:09 -0800)]
Merge pull request #3574 from tahtguymike/maxpool_bug

  Substitute Caffe engine for cuDNN max pooling
  as workaround for in-place after max pooling issue

8 years agoWorkaround for inplace max pooling issue
thatguymike [Wed, 20 Jan 2016 01:01:34 +0000 (17:01 -0800)]
Workaround for inplace max pooling issue

8 years agoPerformance related update of im2col() and col2im() functions
Mariusz Moczala [Wed, 20 Jan 2016 08:28:35 +0000 (09:28 +0100)]
Performance related update of im2col() and col2im() functions

8 years agoMerge pull request #3525 from philkr/cmake_python3
Jon Long [Fri, 8 Jan 2016 02:48:55 +0000 (18:48 -0800)]
Merge pull request #3525 from philkr/cmake_python3

CMake python version fix

8 years agoMerge pull request #2865 from philkr/top_bottom_names
Jon Long [Fri, 8 Jan 2016 02:42:40 +0000 (18:42 -0800)]
Merge pull request #2865 from philkr/top_bottom_names

Exposing layer top and bottom names to python

8 years agoCMake python version fix
philkr [Wed, 6 Jan 2016 15:23:35 +0000 (07:23 -0800)]
CMake python version fix

8 years agoMerge pull request #3519 from philkr/faster_solver
Jeff Donahue [Wed, 6 Jan 2016 05:20:54 +0000 (21:20 -0800)]
Merge pull request #3519 from philkr/faster_solver

Speeding up ADAM (and maybe other solvers too)

8 years agoSpeeding up the GPU solvers
philkr [Tue, 5 Jan 2016 20:45:52 +0000 (12:45 -0800)]
Speeding up the GPU solvers

8 years agoExposing layer top and bottom names to python
philkr [Wed, 5 Aug 2015 18:54:08 +0000 (11:54 -0700)]
Exposing layer top and bottom names to python

8 years agoTestDataTransformer: fix some memory leaks caused by use of 'new'
Jeff Donahue [Wed, 30 Dec 2015 05:10:14 +0000 (21:10 -0800)]
TestDataTransformer: fix some memory leaks caused by use of 'new'

8 years agoMerge pull request #3490 from fyu/spaceplus
Jeff Donahue [Tue, 29 Dec 2015 06:27:34 +0000 (22:27 -0800)]
Merge pull request #3490 from fyu/spaceplus

remove extra space before +

8 years agoremove extra space before +
Fisher Yu [Tue, 29 Dec 2015 03:46:49 +0000 (22:46 -0500)]
remove extra space before +

8 years agoMerge pull request #3487 from longjon/dilation
Jon Long [Mon, 28 Dec 2015 23:56:46 +0000 (15:56 -0800)]
Merge pull request #3487 from longjon/dilation

An early implementation of this functionality for Caffe was written by @gpapan, which was extended and improved by @tamakoji in a previous implementation of this branch.

8 years agoenable dilated deconvolution
Jonathan L Long [Mon, 28 Dec 2015 04:56:24 +0000 (20:56 -0800)]
enable dilated deconvolution

Since the underlying routines are shared, we need only upgrade
compute_output_shape.

8 years agoadd short description of dilation to caffe.proto
Jonathan L Long [Sat, 26 Dec 2015 21:10:11 +0000 (13:10 -0800)]
add short description of dilation to caffe.proto

8 years agodisable dilated deconvolution
Fisher Yu [Sat, 26 Dec 2015 21:10:02 +0000 (13:10 -0800)]
disable dilated deconvolution

8 years agoadd and improve tests for dilated convolution/im2col
Fisher Yu [Sat, 26 Dec 2015 21:04:25 +0000 (13:04 -0800)]
add and improve tests for dilated convolution/im2col

8 years agoadd support for N-D dilated convolution
Fisher Yu [Mon, 28 Dec 2015 04:48:30 +0000 (20:48 -0800)]
add support for N-D dilated convolution

8 years agoadd support for 2D dilated convolution
Fisher Yu [Mon, 7 Dec 2015 01:04:43 +0000 (20:04 -0500)]
add support for 2D dilated convolution

8 years agoMerge pull request #3468 from mnogu/blobs_lr-to-lr_mult
Jon Long [Sun, 20 Dec 2015 23:49:58 +0000 (15:49 -0800)]
Merge pull request #3468 from mnogu/blobs_lr-to-lr_mult

Replace blobs_lr with lr_mult in readme.md.

8 years agoReplace blobs_lr with lr_mult in readme.md.
Muneyuki Noguchi [Sun, 20 Dec 2015 10:12:09 +0000 (19:12 +0900)]
Replace blobs_lr with lr_mult in readme.md.

models/finetune_flickr_style/deploy.prototxt uses lr_mult now.

8 years agoMerge pull request #3439 from flx42/cudnn_v4
Evan Shelhamer [Sat, 12 Dec 2015 04:00:00 +0000 (01:00 -0300)]
Merge pull request #3439 from flx42/cudnn_v4

Add ifdef in CuDNNConvolutionLayer for cuDNN v4

8 years agoFix CuDNNConvolutionLayer for cuDNN v4
Felix Abecassis [Thu, 10 Dec 2015 23:11:51 +0000 (15:11 -0800)]
Fix CuDNNConvolutionLayer for cuDNN v4

Add a macro to check the current cuDNN version

8 years agoMerge pull request #3432 from ianfhunter/patch-1
Evan Shelhamer [Wed, 9 Dec 2015 20:31:45 +0000 (12:31 -0800)]
Merge pull request #3432 from ianfhunter/patch-1

[docs] fix typo in interfaces.md

8 years agoUpdate interfaces.md
Ian Hunter [Wed, 9 Dec 2015 13:29:01 +0000 (13:29 +0000)]
Update interfaces.md

typo

8 years agoMerge pull request #3090 from longjon/summarize-tool
Yangqing Jia [Wed, 9 Dec 2015 01:38:00 +0000 (17:38 -0800)]
Merge pull request #3090 from longjon/summarize-tool

A Python script for at-a-glance net summary

8 years agoMerge pull request #3395 from BonsaiAI/extract_features-compatibility
Jon Long [Fri, 4 Dec 2015 21:11:30 +0000 (13:11 -0800)]
Merge pull request #3395 from BonsaiAI/extract_features-compatibility

Fix compatibility issues with extract_features

8 years agoELU layer with basic tests
Mohamed Omran [Thu, 26 Nov 2015 00:46:42 +0000 (01:46 +0100)]
ELU layer with basic tests

8 years agoCorrect type of device_id; disambiguate shared_ptr
T.E.A de Souza [Sun, 29 Nov 2015 06:24:09 +0000 (14:24 +0800)]
Correct type of device_id; disambiguate shared_ptr

8 years agoMerge pull request #3404 from BonsaiAI/remove-hamming-dist
Jon Long [Wed, 2 Dec 2015 08:11:26 +0000 (00:11 -0800)]
Merge pull request #3404 from BonsaiAI/remove-hamming-dist

Remove hamming_distance and popcount

8 years agoRemove hamming_distance and popcount
Tea [Wed, 2 Dec 2015 07:39:19 +0000 (15:39 +0800)]
Remove hamming_distance and popcount

8 years agoMerge pull request #3313 from gdh1995/master
Evan Shelhamer [Wed, 2 Dec 2015 06:17:36 +0000 (22:17 -0800)]
Merge pull request #3313 from gdh1995/master

don't divide by 0 duration when downloading model binary

8 years agoMerge pull request #3267 from cbalint13/cmake-cudnn
Evan Shelhamer [Wed, 2 Dec 2015 06:05:37 +0000 (22:05 -0800)]
Merge pull request #3267 from cbalint13/cmake-cudnn

[build] Display and store cuDNN version numbers for CMake

8 years agoMerge pull request #3285 from longjon/cuda-dead-cpp
Evan Shelhamer [Wed, 2 Dec 2015 06:04:34 +0000 (22:04 -0800)]
Merge pull request #3285 from longjon/cuda-dead-cpp

Remove dead preprocessor code for number of CUDA threads

8 years agoMerge pull request #3315 from shelhamer/layer-headers
Evan Shelhamer [Wed, 2 Dec 2015 05:30:35 +0000 (21:30 -0800)]
Merge pull request #3315 from shelhamer/layer-headers

Dismantle Layer Headers

8 years agodismantle layer headers
Evan Shelhamer [Tue, 10 Nov 2015 08:22:58 +0000 (00:22 -0800)]
dismantle layer headers

No more monolithic includes: split layers into their own headers for modular inclusion and build.

8 years agoMerge pull request #3378 from BonsaiAI/fix-temp-dir-creation-failure
Ronghang Hu [Sat, 28 Nov 2015 04:53:46 +0000 (20:53 -0800)]
Merge pull request #3378 from BonsaiAI/fix-temp-dir-creation-failure

Safely create temporary files and directories

8 years agoSecure temporary file creation
Tea [Wed, 25 Nov 2015 03:43:45 +0000 (11:43 +0800)]
Secure temporary file creation

8 years agoSecure implementation of MakeTempDir
T.E.A de Souza [Tue, 24 Nov 2015 06:33:27 +0000 (14:33 +0800)]
Secure implementation of MakeTempDir

8 years agoMerge pull request #3320 from BonsaiAI/disambiguate-dtype
Ronghang Hu [Fri, 27 Nov 2015 19:29:15 +0000 (11:29 -0800)]
Merge pull request #3320 from BonsaiAI/disambiguate-dtype

Cast std::max args to Dtype

8 years agoMerge pull request #2693 from AdamStelmaszczyk/patch-1
Ronghang Hu [Fri, 27 Nov 2015 19:25:52 +0000 (11:25 -0800)]
Merge pull request #2693 from AdamStelmaszczyk/patch-1

Deprecated OpenCV consts leading to compilation error

8 years agoMerge pull request #3361 from BonsaiAI/avoid-snprintf
Ronghang Hu [Fri, 27 Nov 2015 18:50:52 +0000 (10:50 -0800)]
Merge pull request #3361 from BonsaiAI/avoid-snprintf

replace snprintf with a C++98 equivalent

8 years agoMerge pull request #3352 from lukeyeager/bvlc/pytest-python-layer
Ronghang Hu [Thu, 26 Nov 2015 18:45:19 +0000 (10:45 -0800)]
Merge pull request #3352 from lukeyeager/bvlc/pytest-python-layer

Skip python layer tests if WITH_PYTHON_LAYER unset

8 years agoMerge pull request #3389 from ronghanghu/remove-stepearly
Ronghang Hu [Thu, 26 Nov 2015 18:34:09 +0000 (10:34 -0800)]
Merge pull request #3389 from ronghanghu/remove-stepearly

Remove bogus stepearly in MNIST example

8 years agoRemove bogus stepearly in MNIST example
Ronghang Hu [Thu, 26 Nov 2015 05:02:02 +0000 (21:02 -0800)]
Remove bogus stepearly in MNIST example

This `examples/lenet/lenet_stepearly_solver.prototxt` is introduced in #190 by mistake, since stepearly is never actually merged.

8 years agoreplace snprintf with a C++98 equivalent
Tea [Tue, 17 Nov 2015 09:05:56 +0000 (17:05 +0800)]
replace snprintf with a C++98 equivalent

8 years agoMerge pull request #3296 from cdoersch/normalize_batch
Jeff Donahue [Mon, 23 Nov 2015 03:33:36 +0000 (19:33 -0800)]
Merge pull request #3296 from cdoersch/normalize_batch

Better normalization options for SoftmaxWithLoss layer

8 years agoBetter normalization options for SoftmaxWithLoss layer.
Carl Doersch [Fri, 6 Nov 2015 22:41:30 +0000 (14:41 -0800)]
Better normalization options for SoftmaxWithLoss layer.

8 years agoMerge pull request #3363 from BonsaiAI/remove-opencv-include
Ronghang Hu [Fri, 20 Nov 2015 11:08:03 +0000 (03:08 -0800)]
Merge pull request #3363 from BonsaiAI/remove-opencv-include

Exclude core.hpp when building without OpenCV

8 years agoMerge pull request #3362 from eelstork/patch-2
Ronghang Hu [Fri, 20 Nov 2015 10:51:31 +0000 (02:51 -0800)]
Merge pull request #3362 from eelstork/patch-2

GetDB must return a value.

8 years agoExclude core.hpp when building without OpenCV
T.E.A de Souza [Fri, 20 Nov 2015 08:52:25 +0000 (16:52 +0800)]
Exclude core.hpp when building without OpenCV

8 years agoFunction must return a value
T.E.A de Souza [Fri, 20 Nov 2015 08:36:29 +0000 (16:36 +0800)]
Function must return a value

Currently compilation will fail with some compilers when LevelDB and LMDB are disabled.
Very similar to a recently fixed issue.

8 years agoConvert std::max args to Dtype
Tea [Thu, 12 Nov 2015 07:15:22 +0000 (15:15 +0800)]
Convert std::max args to Dtype

8 years agoMerge pull request #3359 from ronghanghu/fix-accuracy-layer-top
Evan Shelhamer [Thu, 19 Nov 2015 20:07:32 +0000 (12:07 -0800)]
Merge pull request #3359 from ronghanghu/fix-accuracy-layer-top

Fix MaxTopBlobs in Accuracy Layer

8 years agoFix MaxTopBlobs in Accuracy Layer
Ronghang Hu [Thu, 19 Nov 2015 18:05:48 +0000 (10:05 -0800)]
Fix MaxTopBlobs in Accuracy Layer

Fix the typo "MaxTopBlos" to "MaxTopBlobs". This typo causes maximum top number to be incorrect.

8 years agoSkip python layer tests if WITH_PYTHON_LAYER unset
Luke Yeager [Wed, 18 Nov 2015 18:38:32 +0000 (10:38 -0800)]
Skip python layer tests if WITH_PYTHON_LAYER unset

8 years agoMerge pull request #3321 from BonsaiAI/synced-mem-functions-will-return-a-value
Evan Shelhamer [Mon, 16 Nov 2015 07:19:58 +0000 (23:19 -0800)]
Merge pull request #3321 from BonsaiAI/synced-mem-functions-will-return-a-value

In syncedmem.cpp, functions shall return a value

8 years agoMerge pull request #3261 from kli-nlpr/patch-1
Evan Shelhamer [Mon, 16 Nov 2015 07:00:44 +0000 (23:00 -0800)]
Merge pull request #3261 from kli-nlpr/patch-1

fix plot_training_log.py.example usage text

8 years agoMerge pull request #3332 from alexlee-gk/master
Evan Shelhamer [Mon, 16 Nov 2015 06:51:51 +0000 (22:51 -0800)]
Merge pull request #3332 from alexlee-gk/master

[bug] fix outs and diffs being overwritten in forward_backward_all

8 years agoFix outs and diffs being overwritten in forward_backward_all.
Alex Lee [Sat, 14 Nov 2015 20:49:05 +0000 (12:49 -0800)]
Fix outs and diffs being overwritten in forward_backward_all.

8 years agoMerge pull request #3299 from kkhoot/fix_bn
Evan Shelhamer [Sat, 14 Nov 2015 19:53:26 +0000 (11:53 -0800)]
Merge pull request #3299 from kkhoot/fix_bn

Update BatchNormLayer: more numerical stability and backward with global stats

8 years agoMerge pull request #3323 from BonsaiAI/remove-unnecessary-includes
Evan Shelhamer [Sat, 14 Nov 2015 19:49:00 +0000 (11:49 -0800)]
Merge pull request #3323 from BonsaiAI/remove-unnecessary-includes

Remove un-necessary includes

8 years agoDisplay and store cuDNN version numbers during cmake.
Balint Cristian [Fri, 13 Nov 2015 11:58:49 +0000 (13:58 +0200)]
Display and store cuDNN version numbers during cmake.

8 years agoMerge pull request #3328 from w1res/patch-1
Evan Shelhamer [Fri, 13 Nov 2015 08:25:12 +0000 (00:25 -0800)]
Merge pull request #3328 from w1res/patch-1

[doc] Add parentheses to backward_{cpu,gpu} method.

8 years agoAdd parentheses to backward_{cpu,gpu} method.
Adam Siembida [Thu, 12 Nov 2015 21:03:41 +0000 (16:03 -0500)]
Add parentheses to backward_{cpu,gpu} method.

8 years agoMake backward pass work when global stats is active for BatchNormLayer
Kang Kim [Thu, 12 Nov 2015 17:20:02 +0000 (02:20 +0900)]
Make backward pass work when global stats is active for BatchNormLayer
including minor code cleaning

8 years agoUpdate computation of variance and global stats in BatchNormLayer
Kang Kim [Sat, 7 Nov 2015 03:49:15 +0000 (12:49 +0900)]
Update computation of variance and global stats in BatchNormLayer

8 years agoFix loss of last iteration when average_loss > 1
Benedikt Wilbertz [Wed, 30 Sep 2015 21:02:34 +0000 (23:02 +0200)]
Fix loss of last iteration when average_loss > 1

refactor duplicate code into separate update function for smoothed loss

fix naming convention