platform/upstream/caffeonacl.git
9 years agohot fix for warning
qipeng [Tue, 26 Aug 2014 19:21:06 +0000 (12:21 -0700)]
hot fix for warning

9 years agolint
qipeng [Tue, 26 Aug 2014 19:02:20 +0000 (12:02 -0700)]
lint

9 years agoRe-added solver switch into the new caffe main excutable; fixed AdaGrad MNIST example
qipeng [Tue, 26 Aug 2014 19:01:26 +0000 (12:01 -0700)]
Re-added solver switch into the new caffe main excutable; fixed AdaGrad MNIST example

9 years agolint
qipeng [Tue, 26 Aug 2014 06:17:51 +0000 (23:17 -0700)]
lint

9 years agoAdded sanity check for AdaGradSolver; added MNIST examples for solvers
qipeng [Tue, 26 Aug 2014 06:02:56 +0000 (23:02 -0700)]
Added sanity check for AdaGradSolver; added MNIST examples for solvers

9 years agoMerge Test{SGD,AdaGrad,Nesterov}Solver; they become subclasses of
Jeff Donahue [Fri, 22 Aug 2014 18:51:16 +0000 (11:51 -0700)]
Merge Test{SGD,AdaGrad,Nesterov}Solver; they become subclasses of
TestGradientBasedSolver

9 years agocleanup caffe.proto
Jeff Donahue [Fri, 22 Aug 2014 18:51:29 +0000 (11:51 -0700)]
cleanup caffe.proto

9 years agoadded unit test for solvers and fixed solver bugs
qipeng [Thu, 21 Aug 2014 03:07:53 +0000 (20:07 -0700)]
added unit test for solvers and fixed solver bugs

9 years agoproto conflit, lint, and math_functions (compiler complaint)
qipeng [Wed, 20 Aug 2014 22:19:30 +0000 (15:19 -0700)]
proto conflit, lint, and math_functions (compiler complaint)

9 years agofixes after rebase
qipeng [Wed, 30 Jul 2014 02:56:36 +0000 (19:56 -0700)]
fixes after rebase

9 years agoAddressed Yangqing's comments
qipeng [Wed, 30 Jul 2014 02:46:37 +0000 (19:46 -0700)]
Addressed Yangqing's comments

9 years agofixed caffe.proto after a mistaken rebase
qipeng [Tue, 29 Jul 2014 17:06:32 +0000 (10:06 -0700)]
fixed caffe.proto after a mistaken rebase

9 years agoAdded L1 regularization support for the weights
qipeng [Thu, 24 Jul 2014 20:09:28 +0000 (13:09 -0700)]
Added L1 regularization support for the weights

9 years agobugfixes for AdaGrad
qipeng [Wed, 23 Jul 2014 17:25:44 +0000 (10:25 -0700)]
bugfixes for AdaGrad

9 years agoimproved numerical stability for AdaGrad
qipeng [Wed, 23 Jul 2014 04:17:19 +0000 (21:17 -0700)]
improved numerical stability for AdaGrad

9 years agofixed solver constructor in train_net.cpp
qipeng [Mon, 21 Jul 2014 17:47:55 +0000 (10:47 -0700)]
fixed solver constructor in train_net.cpp

9 years agoconverted pointers to shared_ptr
qipeng [Sun, 20 Jul 2014 15:57:33 +0000 (08:57 -0700)]
converted pointers to shared_ptr

9 years agorestored vituals in solver.hpp
qipeng [Sun, 20 Jul 2014 00:33:35 +0000 (17:33 -0700)]
restored vituals in solver.hpp

9 years agoSolver switching support & implementation of Nesterov's accelerated gradient and...
qipeng [Sat, 19 Jul 2014 21:14:21 +0000 (14:14 -0700)]
Solver switching support & implementation of Nesterov's accelerated gradient and AdaGrad

9 years agouse LMDB in mnist autoencoder examples
Jeff Donahue [Mon, 1 Sep 2014 18:32:12 +0000 (11:32 -0700)]
use LMDB in mnist autoencoder examples

9 years agomake no GPU error in CPU-only mode a little clearer
Evan Shelhamer [Mon, 1 Sep 2014 06:28:28 +0000 (23:28 -0700)]
make no GPU error in CPU-only mode a little clearer

9 years agoinclude comment on CPU mode fine-tuning for Flickr example
Evan Shelhamer [Mon, 1 Sep 2014 06:25:31 +0000 (23:25 -0700)]
include comment on CPU mode fine-tuning for Flickr example

9 years agoMerge pull request #1014 from longjon/cleaner-pycaffe
Evan Shelhamer [Mon, 1 Sep 2014 06:22:21 +0000 (23:22 -0700)]
Merge pull request #1014 from longjon/cleaner-pycaffe

Clean up pycaffe core

9 years agoMerge pull request #1008 from mohomran/mnist_with_lmdb
Evan Shelhamer [Sun, 31 Aug 2014 18:02:45 +0000 (11:02 -0700)]
Merge pull request #1008 from mohomran/mnist_with_lmdb

convert MNIST demo to lmdb, fixes

9 years agoMerge pull request #1015 from mohomran/fixing_parse_log_script
Evan Shelhamer [Sun, 31 Aug 2014 18:00:48 +0000 (11:00 -0700)]
Merge pull request #1015 from mohomran/fixing_parse_log_script

fix for tools/extra/parse_log.sh

9 years agoparse_log.sh adapted to new training log format + fixed typos and updated description
Mohamed Omran [Sun, 31 Aug 2014 13:00:09 +0000 (15:00 +0200)]
parse_log.sh adapted to new training log format + fixed typos and updated description

9 years agoupdated lenet_train_test.prototxt + minor correction to create_mnist.sh
Mohamed Omran [Sun, 31 Aug 2014 12:27:33 +0000 (14:27 +0200)]
updated lenet_train_test.prototxt + minor correction to create_mnist.sh

9 years agominor changes to variable names and error messages + set default backed in convert_mn...
Mohamed Omran [Fri, 29 Aug 2014 15:16:41 +0000 (17:16 +0200)]
minor changes to variable names and error messages + set default backed in convert_mnist_data.cpp to lmdb

9 years agodata now written to backend in batches
Mohamed Omran [Fri, 29 Aug 2014 15:04:25 +0000 (17:04 +0200)]
data now written to backend in batches

9 years agomnist demo now works with lmdb and leveldb (set parameter in create_mnist.sh), switch...
Mohamed Omran [Fri, 29 Aug 2014 14:04:24 +0000 (16:04 +0200)]
mnist demo now works with lmdb and leveldb (set parameter in create_mnist.sh), switched around includes

9 years ago[pycaffe] expose Blob.Reshape
Jonathan L Long [Sun, 31 Aug 2014 08:14:58 +0000 (01:14 -0700)]
[pycaffe] expose Blob.Reshape

While this function should never be called in the current interface, it
will be necessary to have when embedding Python.

9 years ago[pycaffe] remove name property from PyBlob and PyLayer
Jonathan L Long [Sun, 31 Aug 2014 07:45:06 +0000 (00:45 -0700)]
[pycaffe] remove name property from PyBlob and PyLayer

Attaching names to pycaffe Blobs and Layers isn't necessary anymore,
since they are now accessed by name. This change brings PyBlob and
PyLayer closer to their C++ counterparts, and makes it easier to pass
around blobs without knowing their names, e.g., when embedding Python
in a layer.

9 years ago[pycaffe] declare the _caffe module init function
Jonathan L Long [Sun, 31 Aug 2014 07:36:26 +0000 (00:36 -0700)]
[pycaffe] declare the _caffe module init function

9 years ago[pycaffe] split _caffe into source and header files
Jonathan L Long [Sun, 31 Aug 2014 07:26:12 +0000 (00:26 -0700)]
[pycaffe] split _caffe into source and header files

This brings pycaffe more in line with usual C++ style, and will allow
the wrapper classes to be used within caffe when embedding Python.

9 years ago[pycaffe] make PyBlob a template over Dtype
Jonathan L Long [Sun, 31 Aug 2014 07:07:43 +0000 (00:07 -0700)]
[pycaffe] make PyBlob a template over Dtype

We will not actually support double nets in Python with this commit, but
this change will be necessary to create layers that embed Python.

9 years ago[pycaffe] make PyNet a class, not a struct
Jonathan L Long [Sun, 31 Aug 2014 07:01:20 +0000 (00:01 -0700)]
[pycaffe] make PyNet a class, not a struct

9 years ago[pycaffe] use a namespace alias instead of using directives for boost::python
Jonathan L Long [Sun, 31 Aug 2014 06:55:57 +0000 (23:55 -0700)]
[pycaffe] use a namespace alias instead of using directives for boost::python

9 years ago[pycaffe] live in caffe namespace instead of opening it
Jonathan L Long [Sun, 31 Aug 2014 06:49:44 +0000 (23:49 -0700)]
[pycaffe] live in caffe namespace instead of opening it

9 years ago[pycaffe] use class names of the form Py* instead of Caffe*
Jonathan L Long [Sun, 31 Aug 2014 06:48:41 +0000 (23:48 -0700)]
[pycaffe] use class names of the form Py* instead of Caffe*

9 years agoremove residual pthread references, but restore in build for gtest
Evan Shelhamer [Sun, 31 Aug 2014 04:37:12 +0000 (21:37 -0700)]
remove residual pthread references, but restore in build for gtest

gtest depends on pthread in Ubuntu 14.04

9 years agoignore leveldb extension: ldb
Evan Shelhamer [Sun, 31 Aug 2014 04:32:21 +0000 (21:32 -0700)]
ignore leveldb extension: ldb

9 years agoMerge pull request #1004 from kloudkl/ignore
Evan Shelhamer [Sat, 30 Aug 2014 23:59:21 +0000 (16:59 -0700)]
Merge pull request #1004 from kloudkl/ignore

Ignore leveldb more thoroughly

9 years agofix up leveldb ignore
Evan Shelhamer [Sat, 30 Aug 2014 23:59:02 +0000 (16:59 -0700)]
fix up leveldb ignore

9 years agoMerge pull request #1010 from qipeng/boost-thread-with-nvcc
Evan Shelhamer [Sat, 30 Aug 2014 23:41:46 +0000 (16:41 -0700)]
Merge pull request #1010 from qipeng/boost-thread-with-nvcc

Fix boost-CUDA trouble in OS X build.

9 years agoMakefile: fix boost::thread linking, drop pthread, sort
Evan Shelhamer [Sat, 30 Aug 2014 23:41:01 +0000 (16:41 -0700)]
Makefile: fix boost::thread linking, drop pthread, sort

9 years agofix up renaming
Evan Shelhamer [Sat, 30 Aug 2014 23:37:51 +0000 (16:37 -0700)]
fix up renaming

9 years agorenaming && typo fix
qipeng [Sat, 30 Aug 2014 23:31:33 +0000 (16:31 -0700)]
renaming && typo fix

9 years agoMerge pull request #1012 from shelhamer/pr-policy
Evan Shelhamer [Sat, 30 Aug 2014 23:10:38 +0000 (16:10 -0700)]
Merge pull request #1012 from shelhamer/pr-policy

new PR policy: send master fixes + docs to master

9 years ago[docs] new PR policy: send master fixes + docs to master
Evan Shelhamer [Sat, 30 Aug 2014 23:06:17 +0000 (16:06 -0700)]
[docs] new PR policy: send master fixes + docs to master

To address fixes and documentation updates to master between official
releases from `dev` to `master` PRs will be accepted to `master`.

All new development must still branch off and PR to `dev`.

9 years agofix data_transformer param_name
Jeff Donahue [Fri, 29 Aug 2014 10:31:33 +0000 (03:31 -0700)]
fix data_transformer param_name

9 years agoMerge pull request #1011 from YS-L/dev
Evan Shelhamer [Sat, 30 Aug 2014 22:37:37 +0000 (15:37 -0700)]
Merge pull request #1011 from YS-L/dev

Fix norm_region param in cifar10 deployment net

9 years agoxcode compiler complaints (warnings)...
qipeng [Sat, 30 Aug 2014 17:38:33 +0000 (10:38 -0700)]
xcode compiler complaints (warnings)...

9 years agopatch MacOS NVCC boost::thread issue
qipeng [Sat, 30 Aug 2014 17:28:15 +0000 (10:28 -0700)]
patch MacOS NVCC boost::thread issue

9 years agoFix norm_region param in cifar10 deployment net
Yung Siang Liau [Sat, 30 Aug 2014 17:22:04 +0000 (01:22 +0800)]
Fix norm_region param in cifar10 deployment net

9 years agoback-merge
Evan Shelhamer [Sat, 30 Aug 2014 16:38:13 +0000 (09:38 -0700)]
back-merge

  [docs] Several documentation fixes
  Explain how to fix homebrew to allow updates after editing formulae

9 years agoMerge pull request #997 from ozancaglayn/dev
Evan Shelhamer [Sat, 30 Aug 2014 16:37:49 +0000 (09:37 -0700)]
Merge pull request #997 from ozancaglayn/dev

  [docs] Several documentation fixes

9 years ago[docs] Several documentation fixes
Ozan Çağlayan [Thu, 28 Aug 2014 11:05:09 +0000 (14:05 +0300)]
[docs] Several documentation fixes

- Fix a typo and a path in MNIST tutorial
- Rework instructions about Fedora installation

9 years agoMerge pull request #1001 from drdan14/homebrew-git-repair
Evan Shelhamer [Sat, 30 Aug 2014 16:35:07 +0000 (09:35 -0700)]
Merge pull request #1001 from drdan14/homebrew-git-repair

  Explain how to fix homebrew to allow updates after editing formulae

9 years agoExplain how to fix homebrew to allow updates after editing formulae
Daniel Golden [Wed, 27 Aug 2014 16:25:10 +0000 (09:25 -0700)]
Explain how to fix homebrew to allow updates after editing formulae

9 years agoIgnore LevelDB files
Kai Li [Fri, 29 Aug 2014 06:14:32 +0000 (14:14 +0800)]
Ignore LevelDB files

9 years agoMerge pull request #970 from sergeyk/dev
Evan Shelhamer [Fri, 29 Aug 2014 05:40:59 +0000 (22:40 -0700)]
Merge pull request #970 from sergeyk/dev

Fine-tuning tutorial: adapt CaffeNet to Flickr Style data

9 years ago[example] edit fine-tuning and train on ~2000 images, 1557 / 382 split
Evan Shelhamer [Fri, 29 Aug 2014 04:45:46 +0000 (21:45 -0700)]
[example] edit fine-tuning and train on ~2000 images, 1557 / 382 split

- further detail merits of fine-tuning: less starving for itme and data
- set random seed for reproducing the tutorial
- 1557 train / 382 test split is more indicative of training quality
  than splits of 200 images

9 years ago[example] finetuning CaffeNet on Flickr Style data
Sergey Karayev [Sat, 23 Aug 2014 02:30:07 +0000 (19:30 -0700)]
[example] finetuning CaffeNet on Flickr Style data

9 years agoMerge pull request #1003 from shelhamer/examples-from-root
Evan Shelhamer [Fri, 29 Aug 2014 02:47:53 +0000 (19:47 -0700)]
Merge pull request #1003 from shelhamer/examples-from-root

Standardize examples

9 years agoset examples paths relative to root
Evan Shelhamer [Thu, 28 Aug 2014 23:34:57 +0000 (16:34 -0700)]
set examples paths relative to root

9 years agoignore caffe generated files and stop ignoring examples
Evan Shelhamer [Thu, 28 Aug 2014 23:12:01 +0000 (16:12 -0700)]
ignore caffe generated files and stop ignoring examples

- ignore models and snapshots, means, and input databases
- excluding exceptions to commits is easier than including exhaustively

9 years agoMerge pull request #1000 from kloudkl/boost-thread
Evan Shelhamer [Thu, 28 Aug 2014 22:07:11 +0000 (15:07 -0700)]
Merge pull request #1000 from kloudkl/boost-thread

Replace pthread with boost::thread

9 years agoBack-merge to dev for doc fixes + cherry-picks
Evan Shelhamer [Thu, 28 Aug 2014 17:55:33 +0000 (10:55 -0700)]
Back-merge to dev for doc fixes + cherry-picks

Back-merge these doc fixes:

  clarify project origin
  [docs] fix citation bibtex

All other changes are spurious commits from cherry-picking.

9 years agoThe return value of WaitForInternalThreadToExit has reversed
Kai Li [Thu, 28 Aug 2014 16:16:13 +0000 (00:16 +0800)]
The return value of WaitForInternalThreadToExit has reversed

9 years agoAdd boost thread in the travis install script
Kai Li [Thu, 28 Aug 2014 16:05:49 +0000 (00:05 +0800)]
Add boost thread in the travis install script

9 years agoReplace pthread with boost::thread
Kai Li [Thu, 28 Aug 2014 11:17:07 +0000 (19:17 +0800)]
Replace pthread with boost::thread

9 years ago[docs] Update installation docs to include Fedora
Ozan Çağlayan [Mon, 25 Aug 2014 12:16:34 +0000 (15:16 +0300)]
[docs] Update installation docs to include Fedora

- All the instructions mentioned for CentOS/RHEL are also valid for Fedora Linux so add it.
- The package name for atlas is atlas-devel not libatlas-devel for CentOS/RHEL/Fedora so fix it.
- Add 'sudo' prefix to missing places to be consistent with the rest of the document.
- Add instructions for installing python headers in Fedora into the python support section.
- Update Linux installation section to reflect Fedora, refactor a bit.

9 years agoclarify project origin
Evan Shelhamer [Tue, 26 Aug 2014 16:51:19 +0000 (09:51 -0700)]
clarify project origin

9 years agoMerge pull request #857 from netheril96/gflags
Evan Shelhamer [Thu, 28 Aug 2014 04:43:53 +0000 (21:43 -0700)]
Merge pull request #857 from netheril96/gflags

Use gflags to parse command line arguments for convert_imageset

9 years agocreate_imagenet.sh updated to new syntax
netheril96 [Thu, 28 Aug 2014 03:39:28 +0000 (11:39 +0800)]
create_imagenet.sh updated to new syntax

9 years agoMerge pull request #993 from Yangqing/sweep
Jeff Donahue [Wed, 27 Aug 2014 23:01:23 +0000 (16:01 -0700)]
Merge pull request #993 from Yangqing/sweep

fix layer_factory.cpp bug: there should be no ifdefs

9 years agofix layer_factory.cpp bug: there should be no ifdefs
Yangqing Jia [Wed, 27 Aug 2014 22:40:14 +0000 (15:40 -0700)]
fix layer_factory.cpp bug: there should be no ifdefs

9 years agoMerge pull request #977 from ozancaglayan/patch-1
Evan Shelhamer [Wed, 27 Aug 2014 07:18:09 +0000 (00:18 -0700)]
Merge pull request #977 from ozancaglayan/patch-1

[docs] Update installation docs to include Fedora

9 years agoMerge pull request #984 from shelhamer/drop-curand-reset
Evan Shelhamer [Tue, 26 Aug 2014 23:26:15 +0000 (16:26 -0700)]
Merge pull request #984 from shelhamer/drop-curand-reset

Drop obsolete CURAND reset for CUDA 6.5 compatibility

9 years agodefault ilsvrc solving to GPU
Evan Shelhamer [Tue, 26 Aug 2014 21:32:14 +0000 (14:32 -0700)]
default ilsvrc solving to GPU

9 years agodefault ilsvrc solving to GPU
Evan Shelhamer [Tue, 26 Aug 2014 21:32:14 +0000 (14:32 -0700)]
default ilsvrc solving to GPU

9 years agoFIX drop obsolete CURAND reset for CUDA 6.5 compatibility
Jonathan L Long [Tue, 26 Aug 2014 21:01:20 +0000 (14:01 -0700)]
FIX drop obsolete CURAND reset for CUDA 6.5 compatibility

Drop the legacy CURAND initialization steps; these are unnecessary and
cause dramatic slowdowns for CUDA 6.5. This does no harm for K20 usage
counter to the note at least for CUDA 6.5 and 5.0.

9 years agoclarify project origin
Evan Shelhamer [Tue, 26 Aug 2014 16:51:19 +0000 (09:51 -0700)]
clarify project origin

9 years agoFIX web_demo upload was not processing grayscale correctly
Sergey Karayev [Tue, 26 Aug 2014 07:44:56 +0000 (00:44 -0700)]
FIX web_demo upload was not processing grayscale correctly

9 years agoFIX web_demo upload was not processing grayscale correctly
Sergey Karayev [Tue, 26 Aug 2014 07:44:56 +0000 (00:44 -0700)]
FIX web_demo upload was not processing grayscale correctly

9 years agoMerge pull request #981 from jeffdonahue/fix-eltwise-product
Jeff Donahue [Tue, 26 Aug 2014 00:14:10 +0000 (17:14 -0700)]
Merge pull request #981 from jeffdonahue/fix-eltwise-product

Make eltwise product gradient stabler (fixes issues with WITHIN_CHANNEL LRN in cifar_full example)

9 years agoremove warning about LRN layers in CPU mode
Jeff Donahue [Mon, 25 Aug 2014 23:12:33 +0000 (16:12 -0700)]
remove warning about LRN layers in CPU mode

9 years agoAdd "stable_prod_grad" option (on by default) to ELTWISE layer to
Jeff Donahue [Mon, 25 Aug 2014 23:10:22 +0000 (16:10 -0700)]
Add "stable_prod_grad" option (on by default) to ELTWISE layer to
compute the eltwise product gradient using a slower but stabler formula.

9 years agoMerge pull request #980 from jeffdonahue/fix-memory-used
Jeff Donahue [Mon, 25 Aug 2014 20:16:37 +0000 (13:16 -0700)]
Merge pull request #980 from jeffdonahue/fix-memory-used

fix memory_used_ by computing after SetUp

9 years agofix memory_used_ by computing after SetUp
Jeff Donahue [Mon, 25 Aug 2014 19:47:50 +0000 (12:47 -0700)]
fix memory_used_ by computing after SetUp

9 years agoMerge pull request #979 from jeffdonahue/caffe-test-output
Jeff Donahue [Mon, 25 Aug 2014 19:42:24 +0000 (12:42 -0700)]
Merge pull request #979 from jeffdonahue/caffe-test-output

'caffe test' computes and prints all scores and their names

9 years ago'caffe test' prints all scores and their names
Jeff Donahue [Mon, 25 Aug 2014 18:59:20 +0000 (11:59 -0700)]
'caffe test' prints all scores and their names

9 years agofix bug for resizing images.
alfredtofu [Mon, 25 Aug 2014 08:22:52 +0000 (16:22 +0800)]
fix bug for resizing images.

9 years agoMerge pull request #976 from alfredtofu/dev
Evan Shelhamer [Mon, 25 Aug 2014 17:25:53 +0000 (10:25 -0700)]
Merge pull request #976 from alfredtofu/dev

fix image resizing dimensions for IO

9 years ago[docs] Update installation docs to include Fedora
Ozan Çağlayan [Mon, 25 Aug 2014 12:16:34 +0000 (15:16 +0300)]
[docs] Update installation docs to include Fedora

- All the instructions mentioned for CentOS/RHEL are also valid for Fedora Linux so add it.
- The package name for atlas is atlas-devel not libatlas-devel for CentOS/RHEL/Fedora so fix it.
- Add 'sudo' prefix to missing places to be consistent with the rest of the document.
- Add instructions for installing python headers in Fedora into the python support section.
- Update Linux installation section to reflect Fedora, refactor a bit.

9 years agofix bug for resizing images.
alfredtofu [Mon, 25 Aug 2014 08:22:52 +0000 (16:22 +0800)]
fix bug for resizing images.

9 years ago[example] add fully-convolutional efficiency note + confidence map
Evan Shelhamer [Sun, 24 Aug 2014 22:36:52 +0000 (15:36 -0700)]
[example] add fully-convolutional efficiency note + confidence map

- spell out fully-convolutional efficiency
- add confidence map
- fix input size: 451 x 451 is correct for an 8 x 8 output map by the
  equation input size = 227 + 32(d-1) for output map dimension of d

9 years ago[example] add fully-convolutional efficiency note + confidence map
Evan Shelhamer [Sun, 24 Aug 2014 22:36:52 +0000 (15:36 -0700)]
[example] add fully-convolutional efficiency note + confidence map

- spell out fully-convolutional efficiency
- add confidence map
- fix input size: 451 x 451 is correct for an 8 x 8 output map by the
  equation input size = 227 + 32(d-1) for output map dimension of d

9 years agofix internal thread interface confusion
Evan Shelhamer [Sun, 24 Aug 2014 05:41:30 +0000 (22:41 -0700)]
fix internal thread interface confusion

thanks @ronghanghu for pointing this out in #964

9 years agomove {InnerProduct,Eltwise}Layer to common instead of vision
Evan Shelhamer [Sun, 24 Aug 2014 02:25:56 +0000 (19:25 -0700)]
move {InnerProduct,Eltwise}Layer to common instead of vision