Tue, 24 Aug 2010 19:24:54 -0400 |
James Bergstra |
mcRBM - post code-review #1 with Guillaume
|
Tue, 24 Aug 2010 17:01:09 -0400 |
James Bergstra |
added sgd_updates to gd/sgd.py. Modif mcRBM to use it.
|
Tue, 24 Aug 2010 16:51:53 -0400 |
James Bergstra |
added the image_patches dataset used in RanzatoHinton2010
|
Tue, 24 Aug 2010 16:00:08 -0400 |
James Bergstra |
mcRBM - removed container logic that was redundant with global methods
|
Tue, 24 Aug 2010 15:50:14 -0400 |
James Bergstra |
mcRBM - made sampler a global function
|
Tue, 24 Aug 2010 15:28:13 -0400 |
James Bergstra |
mcRBM - cleaned up new_from_dims
|
Tue, 24 Aug 2010 14:54:41 -0400 |
James Bergstra |
mcRBM - removed old train_mcRBM cut-and-paste
|
Tue, 24 Aug 2010 14:54:25 -0400 |
James Bergstra |
mcRBM - cleaned up hmc_sampler
|
Tue, 24 Aug 2010 14:52:09 -0400 |
James Bergstra |
mcRBM - it works and committing it is taking forever... lets try this approach
|
Tue, 24 Aug 2010 14:12:53 -0400 |
James Bergstra |
mcRBM - cleaned up expected hidden unit function
|
Tue, 24 Aug 2010 14:08:24 -0400 |
James Bergstra |
mcRBM - moved contrastive_gradient to global function
|
Tue, 24 Aug 2010 14:03:07 -0400 |
James Bergstra |
mcRBM - cleaned up the preactivation function
|
Tue, 24 Aug 2010 14:02:29 -0400 |
James Bergstra |
mcRBM - moved some things to the top of the file
|
Tue, 24 Aug 2010 13:54:29 -0400 |
James Bergstra |
mcRBM - comment to as_shared
|
Tue, 24 Aug 2010 13:53:29 -0400 |
James Bergstra |
mcRBM - removed numpy_project_onto_ball
|
Tue, 24 Aug 2010 13:51:26 -0400 |
James Bergstra |
mcRBM - removed references to dropped 'P' matrix
|
Tue, 24 Aug 2010 13:50:26 -0400 |
James Bergstra |
mcRBM - added notes to leading comment
|
Mon, 23 Aug 2010 16:06:23 -0400 |
James Bergstra |
image-tiling - better error reporting
|
Mon, 10 May 2010 00:27:02 -0400 |
James Bergstra |
merge
|
Mon, 10 May 2010 00:26:48 -0400 |
James Bergstra |
added help message to glviewer
|
Thu, 15 Apr 2010 10:08:57 -0400 |
James Bergstra |
added __str__ to early Stopper
|
Mon, 23 Aug 2010 16:05:31 -0400 |
James Bergstra |
mcRBM - hacks to driver
|
Mon, 23 Aug 2010 16:04:31 -0400 |
James Bergstra |
mcRBM - particles initialized w randn instead of rand()
|
Mon, 23 Aug 2010 16:04:10 -0400 |
James Bergstra |
mcRBM - changed init of U to match M'A.R's code
|
Mon, 23 Aug 2010 16:03:38 -0400 |
James Bergstra |
mcRBM - added extra free_energy param
|
Mon, 23 Aug 2010 16:02:34 -0400 |
James Bergstra |
mcRBM - put softplus in directly for num. stability
|
Mon, 23 Aug 2010 16:02:02 -0400 |
James Bergstra |
mcRBM - added L1 penalties and normal sign convention to contrastive grad
|
Mon, 23 Aug 2010 15:59:40 -0400 |
James Bergstra |
mcRBM - added numpy import
|
Mon, 23 Aug 2010 15:59:21 -0400 |
James Bergstra |
mcRBM - adding math and comments
|
Mon, 23 Aug 2010 15:54:54 -0400 |
James Bergstra |
image_patches - return centered data by default
|
Mon, 23 Aug 2010 15:54:19 -0400 |
James Bergstra |
typo in image_tiling
|
Fri, 20 Aug 2010 15:29:13 -0400 |
James Bergstra |
image_tiling - added dynamic_range parameter to avoid amplifying noise
|
Fri, 20 Aug 2010 15:28:28 -0400 |
James Bergstra |
image_patches - returning image stack in conventional order
|
Fri, 20 Aug 2010 13:58:56 -0400 |
James Bergstra |
adding algorithms/mcRBM, but it is not done yet
|
Fri, 20 Aug 2010 13:58:41 -0400 |
James Bergstra |
adding algorithms/sparse_coding
|
Fri, 20 Aug 2010 09:31:39 -0400 |
James Bergstra |
merge
|
Fri, 20 Aug 2010 09:31:24 -0400 |
James Bergstra |
fixed saving in image_tiling.py to work for greyscale and colour images
|
Fri, 20 Aug 2010 09:30:34 -0400 |
James Bergstra |
datasets - added olshausen_field data loaders, and an image_patches
|
Fri, 20 Aug 2010 09:29:30 -0400 |
James Bergstra |
work on pca file
|
Fri, 20 Aug 2010 09:28:54 -0400 |
James Bergstra |
Moved stopper to gd/ module
|
Fri, 20 Aug 2010 09:28:39 -0400 |
James Bergstra |
Removed algorithms/minimizer - old and useless code.
|
Fri, 20 Aug 2010 09:27:59 -0400 |
James Bergstra |
Moved sgd to gd/ module
|
Thu, 19 Aug 2010 14:44:15 -0400 |
James Bergstra |
Removed old examples/ submodule
|
Thu, 19 Aug 2010 12:01:30 -0400 |
James Bergstra |
added doc to dataset_ops protocol
|
Thu, 19 Aug 2010 12:00:35 -0400 |
James Bergstra |
Added root __init__ docstring
|
Thu, 19 Aug 2010 12:00:18 -0400 |
James Bergstra |
Adding sampling module
|
Thu, 19 Aug 2010 11:54:12 -0400 |
James Bergstra |
Removed lib module
|
Thu, 19 Aug 2010 11:53:52 -0400 |
James Bergstra |
Added 'preprocessing' module
|
Thu, 19 Aug 2010 11:53:19 -0400 |
James Bergstra |
Moving sgd to a new 'gd' pylearn module, where it should be joined by TONGA
|
Wed, 18 Aug 2010 13:00:19 -0400 |
James Bergstra |
added comments and image-rendering code to cifar-10
|
Wed, 18 Aug 2010 12:59:47 -0400 |
James Bergstra |
tweaks to image_tiling
|
Mon, 16 Aug 2010 10:39:36 -0400 |
gdesjardins |
merge
|
Mon, 16 Aug 2010 10:38:55 -0400 |
gdesjardins |
Toy dataset used in Desjardins et al. (AISTATS 2010).
|
Thu, 12 Aug 2010 09:45:14 -0400 |
Frederic Bastien |
small addition
|
Wed, 11 Aug 2010 21:32:31 -0400 |
Yoshua Bengio |
additions to formulas, data access, hyper-params, scripts
|
Wed, 11 Aug 2010 14:35:57 -0400 |
James Bergstra |
additions to v2_planning
|
Wed, 11 Aug 2010 13:16:51 -0400 |
James Bergstra |
merge
|
Wed, 11 Aug 2010 13:16:05 -0400 |
James Bergstra |
merge
|
Fri, 16 Jul 2010 14:20:48 -0400 |
James Bergstra |
shared/layers/sgd removed cruft
|
Wed, 11 Aug 2010 08:54:13 -0400 |
James Bergstra |
v2_planning.txt
|
Mon, 21 Jun 2010 15:47:30 -0400 |
Yoshua Bengio |
merge
|
Mon, 21 Jun 2010 15:47:12 -0400 |
Yoshua Bengio |
added sigmoid_output_SdA.py
|
Sat, 22 May 2010 17:48:03 -0400 |
gdesjardins |
asdf
|
Sat, 22 May 2010 17:44:26 -0400 |
gdesjardins |
misc. small binary datasets
|
Mon, 10 May 2010 00:25:42 -0400 |
James Bergstra |
added code comments and "all" attribute to datasets.cifar10
|
Sat, 17 Apr 2010 18:33:53 -0400 |
fsavard |
Fixed bug with ErrorSeries not supporting hdf5_group parameter. Added test to check for that bug.
|
Thu, 15 Apr 2010 10:52:02 -0400 |
James Bergstra |
merge
|
Thu, 15 Apr 2010 10:50:10 -0400 |
James Bergstra |
forcing int32 label dtype in shapeset1
|
Thu, 15 Apr 2010 10:49:46 -0400 |
James Bergstra |
removed shape argument from tensor constructor
|
Thu, 15 Apr 2010 10:50:10 -0400 |
James Bergstra |
forcing int32 label dtype in shapeset1
|
Thu, 15 Apr 2010 10:49:46 -0400 |
James Bergstra |
removed shape argument from tensor constructor
|
Thu, 15 Apr 2010 09:11:14 -0400 |
fsavard |
Added other targets (printing to stdout, notably) to seriestables, and corresponding doc
|
Fri, 09 Apr 2010 11:14:24 -0400 |
Pierre-Antoine Manzagol |
Add nist_digits dataset.
|
Thu, 01 Apr 2010 11:53:09 -0400 |
Frederic Bastien |
make test faster in debug mode.
|
Wed, 31 Mar 2010 15:18:09 -0400 |
Frederic Bastien |
make test faster in debug mode
|
Wed, 31 Mar 2010 15:17:40 -0400 |
Frederic Bastien |
use new import path
|
Tue, 30 Mar 2010 13:46:08 -0400 |
Frederic Bastien |
make tests faster in debug mode
|
Tue, 30 Mar 2010 12:09:11 -0400 |
Frederic Bastien |
make the code run faster in debug mode.
|
Fri, 26 Mar 2010 14:57:35 -0400 |
Arnaud Bergeron |
NoteImplementedError -> NotImplementedError
|
Mon, 22 Mar 2010 09:55:12 -0400 |
Frederic Bastien |
removed old parameter that don't exist anymore.
|
Sat, 20 Mar 2010 15:19:43 -0400 |
James Bergstra |
removed bad tracks from MajorMiner dataset
|
Sat, 20 Mar 2010 15:18:54 -0400 |
James Bergstra |
using madplay in audio.AudioRead
|
Fri, 19 Mar 2010 23:32:23 -0400 |
James Bergstra |
majorminer op - make it only work on one track at a time
|
Fri, 19 Mar 2010 23:31:38 -0400 |
James Bergstra |
bugfix in majorminer - must trim newline chars from each line
|
Fri, 19 Mar 2010 23:30:50 -0400 |
James Bergstra |
added new AudioRead op to io.audio
|
Fri, 19 Mar 2010 17:55:23 -0400 |
James Bergstra |
adding majorminer dataset
|
Fri, 19 Mar 2010 17:54:45 -0400 |
James Bergstra |
logreg - added name parameter and autonaming of logreg shared variables
|
Thu, 18 Mar 2010 15:39:32 -0400 |
fsavard |
Changed epy.doc slightly; commented a part concerning a wiki which does not seem to exist anymore; changed THEANO_API_ROOT to PYLEARN_API_ROOT
|
Thu, 18 Mar 2010 15:31:18 -0400 |
fsavard |
Added ext.py from Theano/doc, with some changes
|
Thu, 18 Mar 2010 12:29:36 -0400 |
fsavard |
Had forgotten to hg add SeriesTable .txt doc
|
Thu, 18 Mar 2010 12:23:14 -0400 |
fsavard |
Added SeriesTables module (which I had placed originally in the IFT6266 repository) and its doc. Added a logo for the doc and changed conf.py to reflect correct title.
|
Thu, 18 Mar 2010 11:34:29 -0400 |
fsavard |
Merge
|
Thu, 18 Mar 2010 11:33:49 -0400 |
fsavard |
Added basics for documentation with sphinx and epydoc, by copying files from the Theano/doc directory
|
Thu, 18 Mar 2010 11:18:20 -0400 |
Frederic Bastien |
fix import and fix method name change. This uncover other change not reflected into the test.
|
Thu, 18 Mar 2010 10:49:09 -0400 |
Frederic Bastien |
fix some import.
|
Thu, 18 Mar 2010 10:45:29 -0400 |
Frederic Bastien |
fix some import error.
|
Fri, 12 Mar 2010 01:50:52 -0500 |
James Bergstra |
added comment to sgd file
|
Fri, 12 Mar 2010 01:50:40 -0500 |
James Bergstra |
extension to kouh2008 layer have linear and softplus inputs
|
Mon, 08 Mar 2010 17:56:51 -0500 |
James Bergstra |
added comment to record_versions
|
Wed, 24 Feb 2010 13:55:33 -0500 |
Frederic Bastien |
copy the fct record_version from LeDeepNet
|
Wed, 24 Feb 2010 13:36:01 -0500 |
Frederic Bastien |
backport to python 2.4
|
Thu, 11 Feb 2010 09:44:34 -0500 |
Pierre-Antoine Manzagol |
NistSD dataset: Add range argument for input [-1,1] or [0,1].
|
Tue, 09 Feb 2010 22:11:52 -0500 |
Pierre-Antoine Manzagol |
Nist SD preproc: ensuring upcast to float during centering.
|
Tue, 09 Feb 2010 22:02:34 -0500 |
Pierre-Antoine Manzagol |
Nist SD dataset: modified preproc so input is between -1 an 1.
|
Tue, 09 Feb 2010 20:14:46 -0500 |
Pierre-Antoine Manzagol |
Dataset: added preprocess attribute.
|
Fri, 05 Feb 2010 11:53:48 -0500 |
Pierre-Antoine Manzagol |
Provides a ``Dataset`` for the nist reshuffled digits dataset.
|
Thu, 28 Jan 2010 15:17:57 -0500 |
Arnaud Bergeron |
Add option to use subtensor start in pylearn.io.filetensor
|
Wed, 27 Jan 2010 09:56:37 -0500 |
Frederic Bastien |
fix import for the move of shared var outside of the sandbox.
|
Thu, 21 Jan 2010 13:17:09 -0500 |
Frederic Bastien |
backport to python 2.4
|
Tue, 05 Jan 2010 10:48:59 -0500 |
Frederic Bastien |
some backport to python 2.4
|
Tue, 17 Nov 2009 15:20:40 -0500 |
James Bergstra |
merge
|
Tue, 17 Nov 2009 15:20:16 -0500 |
James Bergstra |
test
|
Tue, 17 Nov 2009 13:33:04 -0500 |
James Bergstra |
mrege
|
Mon, 16 Nov 2009 13:43:53 -0500 |
James Bergstra |
fixes to dataset_ops.shapeset1
|
Mon, 16 Nov 2009 13:37:01 -0500 |
James Bergstra |
Removed hard-coded log level in glviewer
|
Mon, 16 Nov 2009 13:20:51 -0500 |
James Bergstra |
added dataset_ops.shapeset1
|
Tue, 10 Nov 2009 17:59:54 -0500 |
James Bergstra |
merge
|
Tue, 10 Nov 2009 17:49:10 -0500 |
James Bergstra |
improved the initialization range for weights in convolutional version of Rust2005
|
Tue, 10 Nov 2009 17:44:36 -0500 |
James Bergstra |
added cache member to memo-ized functions
|
Tue, 03 Nov 2009 15:26:01 -0500 |
James Bergstra |
added Rust2005Conv
|
Thu, 12 Nov 2009 12:24:05 -0500 |
Pascal Lamblin |
export 'cache' variable
|
Tue, 17 Nov 2009 13:03:04 -0500 |
James Bergstra |
fixed typo in dataset_ops/shapeset1
|
Mon, 16 Nov 2009 13:43:53 -0500 |
James Bergstra |
fixes to dataset_ops.shapeset1
|
Mon, 16 Nov 2009 13:37:52 -0500 |
James Bergstra |
Added support for still images in float storage to glviewer
|
Mon, 16 Nov 2009 13:37:01 -0500 |
James Bergstra |
Removed hard-coded log level in glviewer
|
Mon, 16 Nov 2009 13:21:12 -0500 |
James Bergstra |
added todo note in rust2005
|
Mon, 16 Nov 2009 13:20:51 -0500 |
James Bergstra |
added dataset_ops.shapeset1
|
Mon, 16 Nov 2009 13:20:28 -0500 |
James Bergstra |
cut commented code
|
Mon, 16 Nov 2009 13:19:43 -0500 |
James Bergstra |
added comment to TensorFnDataset
|
Wed, 11 Nov 2009 10:47:15 -0500 |
James Bergstra |
merge
|
Tue, 10 Nov 2009 17:59:54 -0500 |
James Bergstra |
merge
|
Mon, 09 Nov 2009 14:55:08 -0500 |
James Bergstra |
merge
|
Tue, 10 Nov 2009 17:49:10 -0500 |
James Bergstra |
improved the initialization range for weights in convolutional version of Rust2005
|
Tue, 10 Nov 2009 17:48:44 -0500 |
James Bergstra |
improved the initialization range for weights in lecun1998
|
Tue, 10 Nov 2009 17:44:36 -0500 |
James Bergstra |
added cache member to memo-ized functions
|
Tue, 10 Nov 2009 17:43:53 -0500 |
James Bergstra |
added code comment for vim shortcut
|
Tue, 03 Nov 2009 15:26:25 -0500 |
James Bergstra |
merge
|
Tue, 03 Nov 2009 15:26:07 -0500 |
James Bergstra |
added Rust2005Conv
|
Tue, 03 Nov 2009 15:26:01 -0500 |
James Bergstra |
added Rust2005Conv
|
Tue, 03 Nov 2009 15:25:08 -0500 |
James Bergstra |
reparametrization of kouh2008
|
Tue, 03 Nov 2009 15:24:27 -0500 |
James Bergstra |
removed useless print
|
Thu, 12 Nov 2009 12:24:36 -0500 |
Pascal Lamblin |
merge
|
Thu, 12 Nov 2009 12:24:05 -0500 |
Pascal Lamblin |
export 'cache' variable
|
Mon, 09 Nov 2009 16:12:09 -0500 |
Xavier Glorot |
sum change to mean for logistic regression cost over mini-batches for LogReg2
|
Mon, 09 Nov 2009 15:57:00 -0500 |
Xavier Glorot |
sum change to mean for logistic regression cost over mini-batches
|
Mon, 09 Nov 2009 14:41:00 -0500 |
desjagui |
adding simple cifar10 dataset (bypassing dataset_ops), AT&T face dataset and
|
Mon, 09 Nov 2009 14:15:19 -0500 |
desjagui |
* added option for loading norb_small in mono instead of stereo (old mod)
|
Tue, 03 Nov 2009 10:32:51 -0500 |
Frederic Bastien |
fix backport.
|
Tue, 03 Nov 2009 09:50:54 -0500 |
Frederic Bastien |
some fix to the c code of FillMissing. It was not compiling.
|
Mon, 26 Oct 2009 16:33:16 -0400 |
Pascal Lamblin |
merge.
|
Mon, 26 Oct 2009 16:33:03 -0400 |
Pascal Lamblin |
Small type fix
|
Mon, 26 Oct 2009 15:35:38 -0400 |
James Bergstra |
added function documentation to lecun layer
|
Sun, 25 Oct 2009 17:15:31 -0400 |
Pascal Lamblin |
Fix missing "return"
|
Thu, 22 Oct 2009 19:10:15 -0400 |
James Bergstra |
fixed typo in mnist glviewer
|
Thu, 22 Oct 2009 19:10:00 -0400 |
James Bergstra |
added sandbox init
|
Thu, 22 Oct 2009 19:09:45 -0400 |
James Bergstra |
added glviewer to cifar10
|
Thu, 22 Oct 2009 19:09:27 -0400 |
James Bergstra |
added forget() to tearDown of cifar unittests
|
Thu, 22 Oct 2009 18:53:50 -0400 |
James Bergstra |
many changes to kouh2008
|
Thu, 22 Oct 2009 18:53:16 -0400 |
James Bergstra |
moved COIL100 code to sandbox
|
Thu, 22 Oct 2009 18:52:47 -0400 |
James Bergstra |
added the ability to pass fn_args to TensorFnDataset
|
Thu, 22 Oct 2009 18:52:12 -0400 |
James Bergstra |
added forget function to memo in dataset_ops.memo
|
Thu, 22 Oct 2009 18:51:45 -0400 |
James Bergstra |
fixed typo in MNIST
|
Thu, 22 Oct 2009 18:51:20 -0400 |
James Bergstra |
adding cifar10 dataset
|
Fri, 16 Oct 2009 12:20:57 -0400 |
James Bergstra |
merge
|
Fri, 16 Oct 2009 12:19:12 -0400 |
James Bergstra |
removed scan (has been moved to theano)
|
Fri, 16 Oct 2009 12:17:23 -0400 |
James Bergstra |
removed old_dataset
|
Fri, 16 Oct 2009 12:14:43 -0400 |
James Bergstra |
added shared.layers
|
Fri, 16 Oct 2009 12:14:11 -0400 |
James Bergstra |
dataset.config uses logging for warnings
|
Fri, 16 Oct 2009 12:04:05 -0400 |
James Bergstra |
added dataset_ops
|
Wed, 14 Oct 2009 10:19:37 -0400 |
Xavier Glorot |
merge
|
Wed, 14 Oct 2009 10:19:14 -0400 |
Xavier Glorot |
better initialisation and tanh DAA input groups
|
Wed, 07 Oct 2009 11:17:49 -0400 |
Pascal Lamblin |
More descriptive error message when fpconst is missing.
|
Wed, 30 Sep 2009 17:15:01 -0400 |
Xavier Glorot |
merge
|
Wed, 30 Sep 2009 17:14:22 -0400 |
Xavier Glorot |
Added a arsinh activation function DAA input groups
|
Wed, 16 Sep 2009 19:18:29 -0400 |
James Bergstra |
added io/ubyte file for reading MNIST official files
|
Fri, 11 Sep 2009 14:28:50 -0400 |
James Bergstra |
added __init__
|
Thu, 10 Sep 2009 10:30:50 -0400 |
James Bergstra |
merge
|
Thu, 10 Sep 2009 10:30:35 -0400 |
James Bergstra |
added daa example
|
Thu, 10 Sep 2009 10:30:21 -0400 |
James Bergstra |
deleting mnist_factory
|
Fri, 04 Sep 2009 17:37:35 -0400 |
Xavier Glorot |
add a tanh2 activation function DAA_input_groups
|
Wed, 02 Sep 2009 18:28:37 -0400 |
Xavier Glorot |
bug fixes DAA input groups
|
Wed, 02 Sep 2009 14:23:50 -0700 |
dumitru |
...
|
Wed, 02 Sep 2009 14:22:02 -0700 |
dumitru |
backporting to 2.4
|
Wed, 02 Sep 2009 13:36:41 -0400 |
Xavier Glorot |
new parameters and Op for DAA inputs groups
|
Mon, 17 Aug 2009 19:54:35 -0400 |
Xavier Glorot |
save daa input groups
|
Mon, 17 Aug 2009 15:40:57 -0400 |
Xavier Glorot |
model file writing error for DAA input groups
|
Fri, 14 Aug 2009 19:19:30 -0400 |
Xavier Glorot |
independant Tie weigths auxiliary bug fix DAA inputs groups
|
Thu, 13 Aug 2009 12:11:51 -0400 |
Xavier Glorot |
again bug fixes for rescalwsaturation DAA inputs groups
|
Thu, 13 Aug 2009 11:42:53 -0400 |
Xavier Glorot |
fix bug for rescalwsaturation instance method for DAA inputs groups
|
Thu, 13 Aug 2009 11:39:27 -0400 |
Xavier Glorot |
Added a rescalwsaturation instance method for DAA inputs groups
|
Wed, 12 Aug 2009 19:19:24 -0400 |
Xavier Glorot |
Changes in global updates to add local updates (avoid to lose information in the representation) DAA inputs groups
|
Wed, 12 Aug 2009 18:29:18 -0400 |
Xavier Glorot |
Changes in cost function, sum over quadratic and KL instead of cross entropy for global update for DAA inputs groups
|
Tue, 11 Aug 2009 18:37:22 -0400 |
Xavier Glorot |
small changes for quadratic and global update DAA input groups
|
Mon, 10 Aug 2009 19:12:11 -0400 |
Xavier Glorot |
bug fix in unsupupdate instance method DAA input groups
|
Mon, 10 Aug 2009 14:58:40 -0400 |
Xavier Glorot |
independent auxiliary tie weigths behavior DAA inputs groups
|
Thu, 30 Jul 2009 18:09:17 -0400 |
Xavier Glorot |
added a Tie initialisation of the weights option for DAA_inputs_groups
|
Tue, 28 Jul 2009 11:50:07 -0400 |
Xavier Glorot |
load mecanism no Tie model from Tie model for DAA_input_groups
|
Tue, 28 Jul 2009 10:07:01 -0400 |
Xavier Glorot |
small change (daa input groups)
|
Mon, 27 Jul 2009 19:07:22 -0400 |
Xavier Glorot |
added a supupdate method for DAA inputs groups
|
Thu, 23 Jul 2009 12:25:47 -0400 |
Xavier Glorot |
redifinition of saturation monitoring instance methods DAA_inputs_groups
|
Tue, 21 Jul 2009 17:22:20 -0400 |
Xavier Glorot |
fixed cost scaling bug for DAA_inputs_groups
|
Tue, 21 Jul 2009 13:31:43 -0400 |
Xavier Glorot |
DAA_inputs_groups readability
|
Tue, 21 Jul 2009 12:19:30 -0400 |
Xavier Glorot |
variable scale_cost for DAA_inputs_groups
|
Mon, 20 Jul 2009 13:07:51 -0400 |
Xavier Glorot |
fixed a bug in unsupupdate instance method for DAA_inputs_groups
|
Mon, 20 Jul 2009 13:04:18 -0400 |
Xavier Glorot |
added a separated auxiliary and normal inputs gradient computation behavior
|
Wed, 15 Jul 2009 13:19:56 -0400 |
Frederic Bastien |
put in comment a fct call all that seam to don't be used and conflict with numpy and python name.
|
Wed, 15 Jul 2009 13:18:55 -0400 |
Frederic Bastien |
get the mnist data from the pmat file and not the amat file
|
Mon, 13 Jul 2009 16:56:31 -0400 |
Xavier Glorot |
check saturation and noiseseed methods for DAA input groups
|
Fri, 10 Jul 2009 18:14:36 -0400 |
Xavier Glorot |
added a save and load mecanism and fixed bug to DAA_inputs_groups
|
Fri, 10 Jul 2009 16:56:52 -0400 |
Xavier Glorot |
changed behavior of DAAig and StackedDAAig, added a cost scaling parameter
|
Fri, 10 Jul 2009 12:16:20 -0400 |
Frederic Bastien |
fix FillMissing debug mode error. The error was caused by the fact that debug_mode generate an error by default for NaN value. We must disable this check for this op as we deal with NaN value.
|
Fri, 10 Jul 2009 11:31:32 -0400 |
Frederic Bastien |
gix on gcc4.1 FillMissing, seam broken on gcc 4.3
|
Tue, 30 Jun 2009 12:02:11 -0400 |
phamel |
Reverted previous commit
|
Tue, 30 Jun 2009 11:57:20 -0400 |
phamel |
Test commit - Will be reverted
|
Mon, 29 Jun 2009 16:08:00 -0400 |
Frederic Bastien |
dummy change to force the compilation with the good paramter of FillMissing.c_code()
|
Mon, 29 Jun 2009 11:55:29 -0400 |
desjagui |
* fixed test_stacker.py
|
Mon, 29 Jun 2009 09:49:28 -0400 |
Frederic Bastien |
removed manual call to Member and Variable as this is deprecated in theano.
|
Fri, 26 Jun 2009 16:44:55 -0400 |
Olivier Delalleau |
Merged
|
Fri, 26 Jun 2009 09:55:32 -0400 |
Olivier Delalleau |
Fixed test_float
|
Fri, 26 Jun 2009 10:46:45 -0400 |
Frederic Bastien |
implemented FillMissing.c_code. It use the new c_no_compile_args to remove -ffast-math and -ffinite-math-only as they broke isnan.
|
Wed, 24 Jun 2009 10:53:55 -0400 |
Olivier Delalleau |
Fixed some SGD tests
|
Sun, 21 Jun 2009 20:16:23 -0400 |
Xavier Glorot |
Sigmoid before logistic regression to avoid small gradient with tanh in StackedDAAig
|
Sat, 13 Jun 2009 22:02:13 -0400 |
Foo Bar |
Merge
|
Sat, 13 Jun 2009 10:11:43 -0400 |
Xavier Glorot |
change seed initialisation for random object and weigth init for logistic regression
|
Thu, 11 Jun 2009 15:44:32 -0400 |
Frederic Bastien |
small optimisation
|
Thu, 11 Jun 2009 13:35:31 -0400 |
Frederic Bastien |
test more case in FillMissing and print the time of the op.
|
Thu, 11 Jun 2009 11:37:24 -0400 |
Xavier Glorot |
fix a bug for the weight decay
|
Wed, 10 Jun 2009 13:45:52 -0400 |
Frederic Bastien |
remove something that should not have been commited.
|
Wed, 10 Jun 2009 13:43:27 -0400 |
Frederic Bastien |
added more test for the FillMissing op.
|
Wed, 10 Jun 2009 13:42:56 -0400 |
Frederic Bastien |
opt of the FillMissing op. Now 80-90% faster python implementation.
|
Tue, 09 Jun 2009 21:39:50 -0400 |
Frederic Bastien |
scall optimisation to FillMissing.
|
Tue, 09 Jun 2009 21:37:58 -0400 |
Frederic Bastien |
added test for FillMissing op.
|
Mon, 08 Jun 2009 13:14:09 -0400 |
Xavier Glorot |
small change to Checks of scan_inputs_groups ops
|
Mon, 08 Jun 2009 12:14:14 -0400 |
Frederic Bastien |
very small opt.
|
Fri, 05 Jun 2009 15:27:21 -0400 |
Xavier Glorot |
Better declaration of totallocal_update method in StackedDAAig
|
Fri, 05 Jun 2009 15:11:53 -0400 |
Xavier Glorot |
added a totallocal_update method to StackedDAAig
|
Wed, 03 Jun 2009 14:25:56 -0400 |
Xavier Glorot |
Normalized tanh to be equivalent to sigmoid in DAAig
|
Wed, 03 Jun 2009 13:54:31 -0400 |
Frederic Bastien |
fix typo in comment.
|
Wed, 03 Jun 2009 04:07:30 -0400 |
desjagui |
BUG FIXES !!!!!!!
|
Tue, 02 Jun 2009 22:29:49 -0400 |
Pascal Lamblin |
merge
|
Tue, 02 Jun 2009 22:29:21 -0400 |
Pascal Lamblin |
Finalize new validation split: indices are now randomized again inside
|
Tue, 02 Jun 2009 22:26:29 -0400 |
Olivier Delalleau |
Merged
|
Tue, 02 Jun 2009 22:26:04 -0400 |
Olivier Delalleau |
Fix to the 'by_pair' corruption pattern
|
Tue, 02 Jun 2009 14:40:43 -0400 |
Olivier Delalleau |
Fixed bug introduced in previous change meant to improve numerical stability
|
Tue, 02 Jun 2009 21:15:41 -0400 |
Pascal Lamblin |
merge
|
Tue, 02 Jun 2009 21:15:21 -0400 |
Pascal Lamblin |
Prepare the way for the new train/valid split, based on object instances.
|
Tue, 02 Jun 2009 20:21:35 -0400 |
James Bergstra |
merge
|
Tue, 02 Jun 2009 20:21:21 -0400 |
James Bergstra |
better numeric stability in tile_image
|
Tue, 02 Jun 2009 20:20:48 -0400 |
James Bergstra |
fixed typos in wrap_libsvm
|
Tue, 02 Jun 2009 20:20:03 -0400 |
James Bergstra |
added methods kwarg to sgd module
|
Tue, 02 Jun 2009 20:19:30 -0400 |
James Bergstra |
added doc
|
Tue, 02 Jun 2009 13:51:22 -0400 |
Olivier Delalleau |
Merged
|
Tue, 02 Jun 2009 13:50:43 -0400 |
Olivier Delalleau |
New option 'corruption_pattern' to alter the way inputs are corrupted
|
Tue, 02 Jun 2009 11:38:23 -0400 |
Xavier Glorot |
fixed numerical instability for cross entropy cost calculation of DAAig
|
Tue, 02 Jun 2009 11:31:50 -0400 |
Olivier Delalleau |
Added option to allow one to try and reconstruct originally missing inputs
|
Tue, 02 Jun 2009 10:14:00 -0400 |
Olivier Delalleau |
FillMissing can now impute missing values by an array instead of a single constant
|