Wed, 16 Mar 2011 00:05:39 -0400 |
Salah Rifai |
Script to generate resized version of mnist in two different ways: rescaling and zeropadding
|
Sun, 13 Mar 2011 18:25:25 -0400 |
Yoshua Bengio |
added jmlr_review1.txt and jmlr_review2.txt
|
Sun, 09 Jan 2011 22:00:39 -0500 |
Yoshua Bengio |
entered revisions for AMT and SVMs
|
Sun, 09 Jan 2011 21:47:28 -0500 |
Yoshua Bengio |
added aistats2011_revised.tex
|
Sun, 09 Jan 2011 21:33:55 -0500 |
Yoshua Bengio |
merge
|
Sun, 09 Jan 2011 14:35:03 -0500 |
Yoshua Bengio |
more changes to rebuttal
|
Sun, 09 Jan 2011 14:15:04 -0500 |
Yoshua Bengio |
review response
|
Sun, 09 Jan 2011 14:13:23 -0500 |
Yoshua Bengio |
review response changes
|
Sun, 09 Jan 2011 12:45:44 -0500 |
Salah Rifai |
Experiment to test the performance of shallower networks.
|
Sun, 09 Jan 2011 12:13:45 -0500 |
Yoshua Bengio |
added aistats_review_response.txt
|
Fri, 07 Jan 2011 11:44:23 -0500 |
Frederic Bastien |
repair the class/fct that allow to read pnist07 and others by allowing them to read gziped file.
|
Thu, 06 Jan 2011 14:23:41 -0500 |
Frederic Bastien |
reverted previous change as it did not work.
|
Thu, 06 Jan 2011 13:57:05 -0500 |
Frederic Bastien |
fix the reading of PNIST dataset following Dumi compression of the data.
|
Mon, 20 Dec 2010 11:54:35 -0500 |
Yoshua Bengio |
reviews AISTATS
|
Thu, 02 Dec 2010 17:46:29 -0500 |
fsavard |
Inverted two boxes in poster
|
Thu, 02 Dec 2010 17:10:06 -0500 |
fsavard |
Added poster for NIPS2010 (openoffice draw source file)
|
Wed, 01 Dec 2010 18:04:19 -0500 |
Yoshua Bengio |
spotlight pdf
|
Wed, 01 Dec 2010 18:03:11 -0500 |
Yoshua Bengio |
spotlight
|
Fri, 26 Nov 2010 17:41:43 -0500 |
Yoshua Bengio |
NIPS workshop spotlight
|
Mon, 22 Nov 2010 16:03:46 -0500 |
Frederic Bastien |
more fix to newpage command.
|
Mon, 22 Nov 2010 15:13:45 -0500 |
Frederic Bastien |
fix transformation description printing.
|
Mon, 22 Nov 2010 14:52:33 -0500 |
Yoshua Bengio |
draft of NIPS 2010 workshop camera-ready version
|
Sun, 31 Oct 2010 22:40:33 -0400 |
Yoshua Bengio |
aistats submission
|
Sun, 31 Oct 2010 22:27:30 -0400 |
Yoshua Bengio |
aistats submission looking good
|
Sun, 31 Oct 2010 09:12:06 -0400 |
Yoshua Bengio |
merge
|
Sun, 31 Oct 2010 09:11:47 -0400 |
Yoshua Bengio |
draft submission to AISTATS 2011
|
Fri, 15 Oct 2010 14:14:06 -0400 |
boulanni |
added authors and emails list
|
Thu, 14 Oct 2010 22:09:55 -0400 |
Yoshua Bengio |
minor changes to nips workshop submission
|
Thu, 14 Oct 2010 18:04:11 -0400 |
boulanni |
NIPS deep learning workshop submission new .tex file compressed to 8 pages
|
Thu, 14 Oct 2010 15:52:02 -0400 |
Yoshua Bengio |
nips2010_submission.pdf
|
Tue, 05 Oct 2010 15:07:33 -0400 |
Yoshua Bengio |
submitted JMLR pdf
|
Tue, 05 Oct 2010 15:07:12 -0400 |
Yoshua Bengio |
submitted JMLR paper
|
Fri, 01 Oct 2010 15:54:34 -0400 |
Yoshua Bengio |
jmlr_submission
|
Thu, 30 Sep 2010 18:00:37 -0400 |
fsavard |
added springer latex stuff
|
Thu, 30 Sep 2010 17:57:27 -0400 |
fsavard |
Correction de merge
|
Thu, 30 Sep 2010 17:54:56 -0400 |
fsavard |
Changé mlj_submission.tex pour merge
|
Thu, 30 Sep 2010 17:54:23 -0400 |
Yoshua Bengio |
merge
|
Thu, 30 Sep 2010 17:51:46 -0400 |
fsavard |
merge
|
Thu, 30 Sep 2010 17:51:02 -0400 |
fsavard |
Added mlj_submission to group every file needed for that.
|
Thu, 30 Sep 2010 17:43:48 -0400 |
Yoshua Bengio |
contributions.tex
|
Wed, 29 Sep 2010 21:06:47 -0400 |
Yoshua Bengio |
MLJ submission
|
Sat, 18 Sep 2010 18:25:11 -0400 |
Yoshua Bengio |
corrections to techreport.tex
|
Sat, 18 Sep 2010 16:44:46 -0400 |
Yoshua Bengio |
DIRO techreport, sent to arXiv
|
Sat, 18 Sep 2010 16:32:08 -0400 |
Yoshua Bengio |
techreport.tex
|
Thu, 16 Sep 2010 16:21:52 -0400 |
Yoshua Bengio |
ajoute les auteurs
|
Sun, 08 Aug 2010 13:41:46 -0400 |
Yoshua Bengio |
minor
|
Sun, 08 Aug 2010 13:38:55 -0400 |
Yoshua Bengio |
SVM performance is worse
|
Sun, 08 Aug 2010 08:16:21 -0400 |
Yoshua Bengio |
typo fixed, uploaded to CMT
|
Sat, 07 Aug 2010 22:56:46 -0400 |
Yoshua Bengio |
removed linebreaks
|
Sat, 07 Aug 2010 22:54:54 -0400 |
Yoshua Bengio |
now fits
|
Sat, 07 Aug 2010 22:46:12 -0400 |
Yoshua Bengio |
nips_rebuttal_clean
|
Sat, 07 Aug 2010 18:39:36 -0700 |
Dumitru Erhan |
cleaned-up version, fewer typos, shortened (but need 700 chars less)
|
Fri, 06 Aug 2010 15:26:58 -0400 |
Yoshua Bengio |
minor{
|
Fri, 06 Aug 2010 15:12:01 -0400 |
Yoshua Bengio |
gone through all
|
Fri, 06 Aug 2010 14:50:56 -0400 |
Yoshua Bengio |
rebuttal
|
Wed, 04 Aug 2010 15:04:46 -0400 |
Yoshua Bengio |
nips_reviews.txt added
|
Thu, 03 Jun 2010 19:05:08 -0400 |
Yoshua Bengio |
commentaires de Joseph Turian
|
Thu, 03 Jun 2010 13:19:16 -0400 |
Yoshua Bengio |
nouveaux graphes
|
Thu, 03 Jun 2010 13:16:53 -0400 |
Olivier Delalleau |
Merged
|
Thu, 03 Jun 2010 13:16:00 -0400 |
Olivier Delalleau |
Small fixes
|
Thu, 03 Jun 2010 13:15:14 -0400 |
Olivier Delalleau |
Using NIST instead of nist in bibliography
|
Thu, 03 Jun 2010 13:12:47 -0400 |
Frederic Bastien |
bold in graph.
|
Thu, 03 Jun 2010 13:05:38 -0400 |
Frederic Bastien |
graph a little big smaller.
|
Thu, 03 Jun 2010 12:59:11 -0400 |
Frederic Bastien |
new version of graph.
|
Thu, 03 Jun 2010 12:35:41 -0400 |
Frederic Bastien |
added file.
|
Thu, 03 Jun 2010 11:02:39 -0400 |
Olivier Delalleau |
Small fixes (typos / precisions)
|
Thu, 03 Jun 2010 09:18:02 -0400 |
Olivier Delalleau |
Merged
|
Thu, 03 Jun 2010 09:16:53 -0400 |
Olivier Delalleau |
Fixed compilation: .PNG -> .png
|
Thu, 03 Jun 2010 09:17:01 -0400 |
Yoshua Bengio |
encore du visuel de Myriam
|
Thu, 03 Jun 2010 08:14:08 -0400 |
Yoshua Bengio |
cleaner le supplementary material
|
Thu, 03 Jun 2010 08:09:35 -0400 |
Yoshua Bengio |
ameliorer l'aspect visuel, Myriam
|
Wed, 02 Jun 2010 18:17:52 -0400 |
Yoshua Bengio |
minor
|
Wed, 02 Jun 2010 17:40:43 -0400 |
Yoshua Bengio |
ca fitte de nouveau
|
Wed, 02 Jun 2010 17:28:43 -0400 |
Frederic Bastien |
added the real picture
|
Wed, 02 Jun 2010 17:00:11 -0400 |
Frederic Bastien |
separete the transmo image
|
Wed, 02 Jun 2010 15:44:46 -0400 |
Yoshua Bengio |
suggestions from Ian
|
Wed, 02 Jun 2010 13:56:01 -0400 |
Yoshua Bengio |
ca fitte
|
Wed, 02 Jun 2010 13:30:35 -0400 |
Yoshua Bengio |
suggestions de Myriam
|
Wed, 02 Jun 2010 13:09:27 -0400 |
Yoshua Bengio |
charts
|
Wed, 02 Jun 2010 11:45:17 -0400 |
Yoshua Bengio |
changements aux graphes
|
Wed, 02 Jun 2010 10:33:52 -0400 |
Yoshua Bengio |
merge
|
Wed, 02 Jun 2010 10:33:37 -0400 |
Yoshua Bengio |
ca fitte maintenant
|
Wed, 02 Jun 2010 10:29:38 -0400 |
Frederic Bastien |
add dependency to the make file.
|
Wed, 02 Jun 2010 10:29:25 -0400 |
Frederic Bastien |
modif transfo.png file
|
Wed, 02 Jun 2010 10:23:33 -0400 |
Yoshua Bengio |
reduction espace
|
Wed, 02 Jun 2010 01:34:49 -0400 |
boulanni |
Added demo source files
|
Tue, 01 Jun 2010 19:34:22 -0700 |
Dumitru Erhan |
merge
|
Tue, 01 Jun 2010 19:34:00 -0700 |
Dumitru Erhan |
typos and stuff
|
Tue, 01 Jun 2010 18:28:43 -0700 |
Dumitru Erhan |
merge
|
Tue, 01 Jun 2010 18:28:09 -0700 |
Dumitru Erhan |
One comma
|
Tue, 01 Jun 2010 22:25:35 -0400 |
Yoshua Bengio |
typo
|
Tue, 01 Jun 2010 22:12:13 -0400 |
Yoshua Bengio |
added all sizes
|
Tue, 01 Jun 2010 21:24:39 -0400 |
Yoshua Bengio |
resolved conflit
|
Tue, 01 Jun 2010 21:19:54 -0400 |
Yoshua Bengio |
changements aux charts.ods
|
Tue, 01 Jun 2010 18:19:40 -0700 |
Dumitru Erhan |
merging, praying that I did not delete stuff this time :)
|
Tue, 01 Jun 2010 18:18:01 -0700 |
Dumitru Erhan |
Clarifying the experimental setup, typos here and there
|
Tue, 01 Jun 2010 20:48:05 -0400 |
Yoshua Bengio |
longer conclusion
|
Tue, 01 Jun 2010 20:23:59 -0400 |
fsavard |
Added source file (openoffice) for charts
|
Tue, 01 Jun 2010 20:22:42 -0400 |
fsavard |
Small mistake on previous update to chart, corrected.
|
Tue, 01 Jun 2010 20:18:55 -0400 |
fsavard |
Updated chart for error rates (1.4% error on humans, and updated bibliographic references)
|
Tue, 01 Jun 2010 14:23:34 -0700 |
Dumitru Erhan |
correct bib for the supplementary
|
Tue, 01 Jun 2010 14:06:43 -0700 |
Dumitru Erhan |
added paragraph comparing "our" self-taught learning with "theirs"
|
Tue, 01 Jun 2010 16:06:32 -0400 |
Yoshua Bengio |
space adjustments
|
Tue, 01 Jun 2010 15:52:54 -0400 |
Yoshua Bengio |
remis les choses qui avaient disparu
|
Tue, 01 Jun 2010 15:48:46 -0400 |
Yoshua Bengio |
des choses ont disparu
|
Tue, 01 Jun 2010 11:58:14 -0700 |
Dumitru Erhan |
more after lunch :)
|
Tue, 01 Jun 2010 11:32:04 -0700 |
Dumitru Erhan |
section 3 quickpass
|
Tue, 01 Jun 2010 11:15:37 -0700 |
Dumitru Erhan |
merging is fun, merging is fun, merging is fun
|
Tue, 01 Jun 2010 11:14:48 -0700 |
Dumitru Erhan |
section 2, quick first pass
|
Tue, 01 Jun 2010 14:08:44 -0400 |
Frederic Bastien |
make the reference more compact.
|
Tue, 01 Jun 2010 14:08:14 -0400 |
Frederic Bastien |
created a makefile to genereate the pdf
|
Tue, 01 Jun 2010 14:05:21 -0400 |
Yoshua Bengio |
merge
|
Tue, 01 Jun 2010 14:05:02 -0400 |
Yoshua Bengio |
resolved merge conflict
|
Tue, 01 Jun 2010 14:02:04 -0400 |
Yoshua Bengio |
todo done
|
Tue, 01 Jun 2010 11:04:09 -0700 |
Dumitru Erhan |
a few changes in the first section
|
Tue, 01 Jun 2010 10:59:47 -0700 |
Dumitru Erhan |
another try:
|
Tue, 01 Jun 2010 10:58:15 -0700 |
Dumitru Erhan |
argh, sorry about that
|
Tue, 01 Jun 2010 10:56:40 -0700 |
Dumitru Erhan |
rolling back
|
Tue, 01 Jun 2010 13:57:16 -0400 |
Yoshua Bengio |
merge
|
Tue, 01 Jun 2010 13:56:56 -0400 |
Yoshua Bengio |
desiderata
|
Tue, 01 Jun 2010 10:55:08 -0700 |
Dumitru Erhan |
merge
|
Tue, 01 Jun 2010 10:53:07 -0700 |
dumitru |
commit early, commit often: a couple of changes to kick-start things
|
Tue, 01 Jun 2010 13:53:05 -0400 |
Yoshua Bengio |
-
|
Tue, 01 Jun 2010 13:37:40 -0400 |
Yoshua Bengio |
changements de Myriam
|
Tue, 01 Jun 2010 12:28:05 -0400 |
Yoshua Bengio |
-
|
Tue, 01 Jun 2010 12:13:10 -0400 |
Yoshua Bengio |
merge
|
Tue, 01 Jun 2010 12:12:52 -0400 |
Yoshua Bengio |
changements de Myriam
|
Tue, 01 Jun 2010 11:36:41 -0400 |
Frederic Bastien |
add missing file for bibtex and make it smaller.
|
Tue, 01 Jun 2010 11:36:07 -0400 |
Frederic Bastien |
more file to ignore.
|
Tue, 01 Jun 2010 11:07:25 -0400 |
Frederic Bastien |
make the reference shorter.
|
Tue, 01 Jun 2010 11:02:10 -0400 |
Yoshua Bengio |
typos
|
Tue, 01 Jun 2010 07:56:00 -0400 |
Yoshua Bengio |
merge
|
Tue, 01 Jun 2010 07:55:38 -0400 |
Yoshua Bengio |
difference stat. sign.
|
Mon, 31 May 2010 19:19:21 -0700 |
dumitru |
adding a section title
|
Mon, 31 May 2010 22:15:44 -0400 |
Yoshua Bengio |
merge
|
Mon, 31 May 2010 22:14:32 -0400 |
Yoshua Bengio |
caption + consequent
|
Mon, 31 May 2010 19:07:59 -0700 |
dumitru |
merge
|
Mon, 31 May 2010 19:07:35 -0700 |
dumitru |
Moved the tables into a separate supplementary material file
|
Mon, 31 May 2010 22:04:44 -0400 |
Yoshua Bengio |
-
|
Mon, 31 May 2010 22:03:35 -0400 |
Yoshua Bengio |
section resultats et appendice
|
Mon, 31 May 2010 21:50:00 -0400 |
Yoshua Bengio |
les tables enlevées
|
Mon, 31 May 2010 20:42:22 -0400 |
Yoshua Bengio |
reduction de taille
|
Mon, 31 May 2010 17:57:45 -0400 |
Yoshua Bengio |
pointeur a la demo
|
Mon, 31 May 2010 13:59:11 -0400 |
Yoshua Bengio |
changement au titre et reecriture abstract
|
Sun, 30 May 2010 19:43:13 -0400 |
Dumitru Erhan |
background.png and Background.PNG produce a case collision on Mac and Windows, renamed one of them
|
Sun, 30 May 2010 19:05:22 -0400 |
Xavier Glorot |
added number of train test and valid for NIST
|
Sun, 30 May 2010 18:54:36 -0400 |
Xavier Glorot |
Added transformation figure
|
Sun, 30 May 2010 17:49:14 -0400 |
Xavier Glorot |
added new transformation images
|
Sun, 30 May 2010 17:24:26 -0400 |
Xavier Glorot |
small desription of the font added
|
Sun, 30 May 2010 12:06:45 -0400 |
fsavard |
Merge
|
Sun, 30 May 2010 12:04:05 -0400 |
fsavard |
Added charts to nips2010_submission.tex
|
Sun, 30 May 2010 11:18:11 -0400 |
Yoshua Bengio |
fits now, but still now graphics
|
Sun, 30 May 2010 10:44:40 -0400 |
Yoshua Bengio |
merge
|
Sun, 30 May 2010 10:44:20 -0400 |
Yoshua Bengio |
conclusion
|
Sat, 29 May 2010 15:38:13 -0700 |
Dumitru Erhan |
PNIST & NIST results on digits-only data
|
Sat, 29 May 2010 15:25:59 -0700 |
Dumitru Erhan |
Digits-only results on P07 data
|
Sat, 29 May 2010 18:03:37 -0400 |
Yoshua Bengio |
redone bib style
|
Sat, 29 May 2010 16:56:49 -0400 |
Yoshua Bengio |
merged ml.bib with svn/articles/bib
|
Sat, 29 May 2010 16:50:03 -0400 |
Yoshua Bengio |
refs
|
Fri, 28 May 2010 17:39:22 -0600 |
Yoshua Bengio |
asking the questions
|
Fri, 28 May 2010 17:33:15 -0600 |
Yoshua Bengio |
ajouter nips10submit_e.sty
|
Fri, 28 May 2010 17:21:21 -0600 |
Yoshua Bengio |
nips2010_submission.tex
|
Fri, 28 May 2010 19:07:14 -0400 |
Xavier Glorot |
added small information on NISTP
|
Fri, 28 May 2010 08:49:36 -0600 |
Yoshua Bengio |
cleaner le techreport
|
Fri, 28 May 2010 07:48:14 -0600 |
Yoshua Bengio |
lit. results
|
Thu, 27 May 2010 21:53:24 -0600 |
Yoshua Bengio |
ajoute dernier tableau de resultats
|
Thu, 27 May 2010 08:29:26 -0600 |
Yoshua Bengio |
merge
|
Thu, 27 May 2010 08:29:04 -0600 |
Yoshua Bengio |
added many results
|
Wed, 26 May 2010 20:25:39 -0400 |
SylvainPL |
Rajout de la ligne qui compte le nombre de mini-batch (enleve precedemment)
|
Wed, 26 May 2010 20:24:16 -0400 |
SylvainPL |
Ajout d'option pour choisir l'ensemble de donnees (train/valid/test) ainsi que mettre NIST,P07 et PNIST pour tous les modeles testes
|
Wed, 26 May 2010 20:23:02 -0400 |
SylvainPL |
Ajout de trois lignes de code pour le calcul de l'erreur standart
|
Thu, 13 May 2010 12:15:16 -0400 |
fsavard |
Autre test
|
Thu, 13 May 2010 12:03:57 -0400 |
fsavard |
Test
|
Mon, 10 May 2010 13:52:57 -0400 |
Arnaud Bergeron |
Add a small paragraph mentionning the distribution differences and a figure illustrating the difference.
|
Mon, 10 May 2010 13:44:11 -0400 |
Arnaud Bergeron |
Add a graph for the NIST training set and normalize the values.
|
Mon, 10 May 2010 11:41:02 -0400 |
Arnaud Bergeron |
New script to create histograms for datasets (for the report)
|
Sun, 09 May 2010 15:01:31 -0400 |
humel |
Added support to calculate the human consensus error
|
Fri, 07 May 2010 17:24:21 -0400 |
SylvainPL |
Correction d'un petit bug d'indice. Le script est maintenant plus juste
|
Fri, 07 May 2010 16:59:03 -0400 |
humel |
Tables of classes frequencies
|
Tue, 04 May 2010 19:59:14 -0400 |
SylvainPL |
Ajout de fonctionnalite pour calculer les erreurs de validation avec PNIST et P07
|
Tue, 04 May 2010 11:17:27 -0400 |
xaviermuller |
added jobman all test + val error and sigmoid output
|
Mon, 03 May 2010 18:43:24 -0400 |
Xavier Glorot |
small syntax fix
|
Mon, 03 May 2010 18:38:58 -0400 |
Xavier Glorot |
merge
|
Mon, 03 May 2010 18:38:27 -0400 |
Xavier Glorot |
added permut pixel
|
Mon, 03 May 2010 13:55:03 -0400 |
Arnaud Bergeron |
merge
|
Mon, 03 May 2010 12:18:03 -0400 |
Guillaume Sicard |
updating values in Training with More Classes than Necessary
|
Mon, 03 May 2010 09:42:36 -0400 |
Guillaume Sicard |
Bug fix in per-class a priori error : should be lower than before for both upper and lower
|
Mon, 03 May 2010 07:46:18 -0400 |
Yoshua Bengio |
added AMT
|
Mon, 03 May 2010 06:17:54 -0400 |
Guillaume Sicard |
added Training with More Classes than Necessary
|
Mon, 03 May 2010 06:14:05 -0400 |
Guillaume Sicard |
Added : single class version of SDA
|
Mon, 03 May 2010 06:11:30 -0400 |
Guillaume Sicard |
bug fix in output
|
Mon, 03 May 2010 03:08:34 -0400 |
goldfinger |
added description of nist19 and captcha data sources
|
Mon, 03 May 2010 02:44:11 -0400 |
goldfinger |
added entry referenced in the description of nist data source
|
Mon, 03 May 2010 02:43:08 -0400 |
goldfinger |
added description of nist19 and captcha data sources
|
Mon, 03 May 2010 01:07:21 -0400 |
Xavier Glorot |
correction in the transformation descripition
|
Sun, 02 May 2010 22:26:19 -0400 |
Dumitru Erhan |
fixed error computation for 36 classes
|
Sun, 02 May 2010 21:32:09 -0400 |
Dumitru Erhan |
AMT, full results
|
Sat, 01 May 2010 14:27:56 -0400 |
fsavard |
Added text for ratures filter
|
Fri, 30 Apr 2010 16:29:43 -0400 |
Xavier Glorot |
merge
|
Fri, 30 Apr 2010 16:29:17 -0400 |
Xavier Glorot |
change order of transformations
|
Fri, 30 Apr 2010 16:24:35 -0400 |
Xavier Glorot |
small syntax error
|
Fri, 30 Apr 2010 16:28:13 -0400 |
Arnaud Bergeron |
Fix latex syntax in 2.12 (would not compile otherwise).
|
Fri, 30 Apr 2010 16:25:37 -0400 |
Arnaud Bergeron |
merge
|
Fri, 30 Apr 2010 16:24:30 -0400 |
Arnaud Bergeron |
Basic text for section 3, add a bit more detail to section 4.2.2
|
Fri, 30 Apr 2010 16:23:45 -0400 |
Xavier Glorot |
merge
|
Fri, 30 Apr 2010 16:23:15 -0400 |
Xavier Glorot |
added transformations description and began dataset descriptions
|
Fri, 30 Apr 2010 14:48:08 -0400 |
SylvainPL |
Correction d'une petite erreur dans le nom des fichiers de parametres de pretrain
|
Fri, 30 Apr 2010 14:47:25 -0400 |
SylvainPL |
Sert a calculer l'erreur sur les differentes classes de NIST.
|
Fri, 30 Apr 2010 09:25:20 -0400 |
fsavard |
Completed the pinch transformation text
|
Thu, 29 Apr 2010 18:26:30 -0400 |
fsavard |
Correction d'une erreur pour pinch et ajout d'une ref bibliographique
|
Thu, 29 Apr 2010 17:21:48 -0400 |
fsavard |
Added the parts concerning the transformations I'd announced I'd do: Local elastic deformations; occlusions; gimp transformations; salt and pepper noise; background images
|
Thu, 29 Apr 2010 17:04:12 -0400 |
xaviermuller |
saving test score from best validation score in db now
|
Thu, 29 Apr 2010 16:51:03 -0400 |
youssouf |
initial commit of mlp with options for detection and 36 classes
|
Thu, 29 Apr 2010 14:28:52 -0400 |
humel |
Added support for calculating the test error over different set of classes (lower,upper,digits,all,36)
|
Thu, 29 Apr 2010 13:18:15 -0400 |
Arnaud Bergeron |
Better description of the model parameters.
|
Thu, 29 Apr 2010 12:55:57 -0400 |
Arnaud Bergeron |
Description brève de MLP et SdA
|
Wed, 28 Apr 2010 16:39:10 -0400 |
Xavier Glorot |
added some images to illustrate the transformation
|
Wed, 28 Apr 2010 14:35:30 -0400 |
Xavier Glorot |
merge
|
Wed, 28 Apr 2010 14:35:01 -0400 |
Xavier Glorot |
description des transformations en cours ajout d un fichier special.bib pour des references specifiques
|
Wed, 28 Apr 2010 14:28:32 -0400 |
Arnaud Bergeron |
Add bibliography style files.
|
Wed, 28 Apr 2010 12:59:29 -0400 |
xaviermuller |
corrected bug where we inverted min and maj
|
Wed, 28 Apr 2010 11:45:14 -0400 |
xaviermuller |
added digit/char error
|
Wed, 28 Apr 2010 11:30:37 -0400 |
humel |
Undoing some unwanted changes
|
Wed, 28 Apr 2010 11:28:28 -0400 |
humel |
Added more stats printing. Now you dont need to parameters which dataset you are testing, it will detect it automatically
|
Wed, 28 Apr 2010 01:25:52 -0400 |
humel |
Fixed class number bug
|
Wed, 28 Apr 2010 00:49:41 -0400 |
humel |
Adding comments and fixing out of bounds index
|
Wed, 28 Apr 2010 00:38:31 -0400 |
humel |
Initial commit for calculating the test error of the AMT classifier
|
Tue, 27 Apr 2010 20:54:28 -0400 |
SylvainPL |
Sert pour calculer les erreurs de test sur les differentes parties de NIST pour un modele donne
|
Tue, 27 Apr 2010 20:50:56 -0400 |
SylvainPL |
Ajout de fonctionnalite pour pouvoir facilement calculer une erreur sur differentes parties de NSIT
|
Tue, 27 Apr 2010 13:47:33 -0400 |
goldfinger |
utils for the amazon MT code
|
Tue, 27 Apr 2010 13:45:32 -0400 |
goldfinger |
code for amazon MT
|
Tue, 27 Apr 2010 13:19:00 -0400 |
humel |
Generation script for AMT data
|
Tue, 27 Apr 2010 09:59:40 -0400 |
Yoshua Bengio |
added examples of figure and table of results
|
Tue, 27 Apr 2010 09:56:18 -0400 |
Yoshua Bengio |
possible image for illustrating perturbations
|
Tue, 27 Apr 2010 08:55:30 -0400 |
SylvainPL |
Correction d'un bug mineurpour le nom de la serie h5
|
Tue, 27 Apr 2010 08:55:16 -0400 |
SylvainPL |
Correction d'un bug mineurpour le nom de la serie h5
|
Tue, 27 Apr 2010 08:43:04 -0400 |
SylvainPL |
Ajout d'une option finetune amelioree pour PNIST07
|
Tue, 27 Apr 2010 08:42:43 -0400 |
SylvainPL |
Changement pour prendre en compte l'option finetune amelioree pour PNIST07
|
Tue, 27 Apr 2010 08:41:58 -0400 |
SylvainPL |
Ajout d'une option finetune amelioree pour PNIST07
|
Tue, 27 Apr 2010 08:41:44 -0400 |
SylvainPL |
Ajout d'une option finetune amelioree pour PNIST07
|
Tue, 27 Apr 2010 08:20:59 -0400 |
SylvainPL |
Ajout de la faculte de prendre PNIST07
|
Tue, 27 Apr 2010 08:20:37 -0400 |
SylvainPL |
Ajout de la faculte de prendre PNIST07
|
Tue, 27 Apr 2010 08:20:12 -0400 |
SylvainPL |
Ajout de la faculte de prendre PNIST07
|
Tue, 27 Apr 2010 08:18:10 -0400 |
SylvainPL |
Ajout de la faculte de prendre PNIST07
|
Mon, 26 Apr 2010 22:36:07 -0400 |
Yoshua Bengio |
authors
|
Mon, 26 Apr 2010 14:56:34 -0400 |
Yoshua Bengio |
merge
|
Mon, 26 Apr 2010 14:56:06 -0400 |
Yoshua Bengio |
added writeup skeleton
|
Mon, 26 Apr 2010 14:39:03 -0400 |
xaviermuller |
added initial model for weights in jobman
|
Sun, 25 Apr 2010 17:12:03 -0400 |
Arnaud Bergeron |
branch merge
|
Sun, 25 Apr 2010 17:10:09 -0400 |
Arnaud Bergeron |
Fix filter sizes to have meaningful networks with more that 2 layers.
|
Sun, 25 Apr 2010 14:53:10 -0400 |
Xavier Glorot |
Changes to cast NIST data to floatX for rbm code
|
Sun, 25 Apr 2010 13:55:07 -0400 |
Xavier Glorot |
merge2
|
Sun, 25 Apr 2010 13:54:57 -0400 |
Xavier Glorot |
merge1
|
Sun, 25 Apr 2010 13:28:45 -0400 |
goldfinger |
correcting some bugs
|
Sun, 25 Apr 2010 12:31:22 -0400 |
youssouf |
initial commit of sda with 36 classes
|
Sat, 24 Apr 2010 15:34:07 -0400 |
Xavier Glorot |
changes in generation script for the new data
|
Sat, 24 Apr 2010 11:32:26 -0400 |
goldfinger |
modified to run experiments with PNIST
|
Fri, 23 Apr 2010 12:12:03 -0400 |
SylvainPL |
Correction d'un bug avec la fonction de log-likelihood pour utilisation de non-linearite de sortie sigmoides
|
Fri, 23 Apr 2010 11:39:55 -0400 |
SylvainPL |
correction d'un petit bug dans la fonction traduire()
|
Thu, 22 Apr 2010 20:06:11 -0400 |
humel |
undoing unwanted changes to setup_batches.py
|
Thu, 22 Apr 2010 19:57:05 -0400 |
humel |
Fixing a wrong commit and committing more files.
|
Thu, 22 Apr 2010 19:50:21 -0400 |
humel |
Fixing a wrong commit and committing more files.
|
Thu, 22 Apr 2010 13:43:04 -0400 |
SylvainPL |
Correction d'un bug dans le pre-train du SDA cause par tanh
|
Thu, 22 Apr 2010 13:18:12 -0400 |
SylvainPL |
programme pour voir les images ainsi que la prediction faite par le modele
|
Thu, 22 Apr 2010 13:17:19 -0400 |
SylvainPL |
Ajout d'une fonctionnalite utile avec le programme voir_erreurs.py afin de voir les exemples ainsi que la prediction du modele donne dans le fichier config.py
|
Thu, 22 Apr 2010 10:34:26 -0400 |
fsavard |
Problèmes de révisions hg... tentative de merger
|
Thu, 22 Apr 2010 10:19:07 -0400 |
fsavard |
Fichier config.py.example supprimé je ne sais pas pourquoi ?! Enfin je le réajoute
|
Thu, 22 Apr 2010 00:49:42 -0400 |
humel |
Initial commit for the experimental setup of the denoising convolutional network
|
Thu, 22 Apr 2010 00:00:09 -0400 |
Guillaume Sicard |
version mlp modifiée pour la selection du ratio de la classe principale
|
Wed, 21 Apr 2010 23:47:50 -0400 |
Guillaume Sicard |
Extension du choix de la classe principale pour les batches d'entrainement
|
Wed, 21 Apr 2010 15:07:09 -0400 |
xaviermuller |
added support for pnist in iterator. corrected a print bug in mlp
|
Wed, 21 Apr 2010 14:54:54 -0400 |
SylvainPL |
Ajout d'une fonctionnalite pour pouvoir avoir un taux d'apprentissage decroissant dans le pretrain
|
Wed, 21 Apr 2010 14:51:14 -0400 |
SylvainPL |
Ajout d'une fonctionnalite pour pouvoir avoir un taux d'apprentissage decroissant dans le pretrain
|
Wed, 21 Apr 2010 14:50:59 -0400 |
SylvainPL |
Ajout d'une fonctionnalite pour pouvoir avoir un taux d'apprentissage decroissant dans le pretrain
|
Wed, 21 Apr 2010 14:07:53 -0400 |
SylvainPL |
Ajout de la capacite d'utiliser des tanh a la place des sigmoides et sigmoide en sortie plutot que softmax
|
Tue, 20 Apr 2010 16:38:13 -0400 |
SylvainPL |
Amelioration de l'efficacite de la sauvegarde des parametres
|
Mon, 19 Apr 2010 10:12:17 -0400 |
xaviermuller |
added pnist support, will check in code for data set iterator later
|
Mon, 19 Apr 2010 08:17:45 -0400 |
goldfinger |
training an rbm using cd or pcd
|
Mon, 19 Apr 2010 08:16:56 -0400 |
goldfinger |
base class for an rbm
|
Mon, 19 Apr 2010 07:09:44 -0400 |
Guillaume Sicard |
Ajouté dans set_batches : choix de la classe principale
|
Sun, 18 Apr 2010 13:22:24 -0400 |
fsavard |
example config file needs EXPERIMENT_PATH outside isolate_experiment
|
Sun, 18 Apr 2010 13:21:21 -0400 |
fsavard |
Correcting bugs after trying to launch on cluster. Image dir must be created after workdir is changed.
|
Sun, 18 Apr 2010 11:54:57 -0400 |
fsavard |
Last few bugfixes before launching
|
Sun, 18 Apr 2010 11:47:35 -0400 |
fsavard |
Slight bug
|
Sun, 18 Apr 2010 11:41:27 -0400 |
fsavard |
Added init.py to crbm
|
Sun, 18 Apr 2010 11:39:24 -0400 |
fsavard |
Corrected a few bugs, no new features. Supposedly ready to run on cluster.
|
Sat, 17 Apr 2010 20:29:18 -0400 |
fsavard |
Aded code to run experiment on cluster, separate configuration from other machinery. Not tested yet.
|
Sat, 17 Apr 2010 12:42:48 -0400 |
xaviermuller |
added async save, restart from old model and independant error calculation based on Arnaud's iterator
|
Fri, 16 Apr 2010 16:05:55 -0400 |
fsavard |
Added convolutional RBM (ala Lee09) code, imported from my working dir elsewhere. Seems to work for one layer. No subsampling yet.
|
Thu, 15 Apr 2010 14:41:00 -0400 |
sylvainpl |
Arrange pour avoir un taux d'apprentissage decroissant decent pour NIST
|
Thu, 15 Apr 2010 12:53:03 -0400 |
Arnaud Bergeron |
noisyness -> noise
|
Wed, 14 Apr 2010 16:06:43 -0400 |
Arnaud Bergeron |
Name the important layers to be able to change them easily later.
|
Wed, 14 Apr 2010 16:06:04 -0400 |
Arnaud Bergeron |
Adapt parameters (especially the number of epochs) to P07
|
Wed, 14 Apr 2010 11:51:18 -0400 |
Guillaume Sicard |
Correction de bug numpy array et ajout d'une deuxième classe auxiliaire
|
Wed, 14 Apr 2010 10:17:33 -0400 |
SylvainPL |
Ajout d'une ligne faisant rapetisser le finetune_lr pour NIST lorsqu'on a P07+NIST
|
Sun, 11 Apr 2010 19:52:52 -0400 |
SylvainPL |
Rajout de la capacite de faire decroitre le taux d'apprentissage si demande
|
Sun, 11 Apr 2010 19:52:44 -0400 |
SylvainPL |
Rajout de la capacite de faire decroitre le taux d'apprentissage si demande
|
Sun, 11 Apr 2010 19:52:35 -0400 |
SylvainPL |
Rajout de la capacite de faire decroitre le taux d'apprentissage si demande
|
Sun, 11 Apr 2010 19:52:28 -0400 |
SylvainPL |
Rajout de la capacite de faire decroitre le taux d'apprentissage si demande
|
Fri, 09 Apr 2010 15:50:11 -0400 |
SylvainPL |
Calcule l'erreur d'entrainement
|
Fri, 09 Apr 2010 15:49:42 -0400 |
SylvainPL |
Ajout d'une fonction pour calculer l'erreur effectuee par le modele sur un ensemble pre-determine
|
Thu, 08 Apr 2010 11:22:11 -0400 |
xaviermuller |
fixed bug in jobman interface
|
Thu, 08 Apr 2010 11:01:55 -0400 |
xaviermuller |
finished code clean up and testing
|
Tue, 06 Apr 2010 16:00:52 -0400 |
xaviermuller |
code clean up in progress
|
Fri, 02 Apr 2010 14:54:05 -0400 |
SylvainPL |
finetune NIST+P07 change pour P07+NIST, les experiences n'ont pas ete concluentes
|
Fri, 02 Apr 2010 14:53:46 -0400 |
SylvainPL |
finetune NIST+P07 change pour P07+NIST, les experiences n'ont pas ete concluentes
|
Fri, 02 Apr 2010 14:53:32 -0400 |
SylvainPL |
finetune NIST+P07 change pour P07+NIST, les experiences n'ont pas ete concluentes
|
Fri, 02 Apr 2010 09:12:40 -0400 |
SylvainPL |
Ajoute plus de fonctionnalite
|
Thu, 01 Apr 2010 20:09:14 -0400 |
SylvainPL |
Ajout de noms differents pour les series produites pour differents choix de pretrain
|
Thu, 01 Apr 2010 17:47:53 -0400 |
SylvainPL |
Correct a little bug about maximum number of minibatches
|
Thu, 01 Apr 2010 14:29:48 -0400 |
Arnaud Bergeron |
branch merge
|
Thu, 01 Apr 2010 14:28:50 -0400 |
Arnaud Bergeron |
Print the error sometimes in the pretrain loop.
|
Thu, 01 Apr 2010 14:25:55 -0400 |
SylvainPL |
Avoir exactement le meme jeu de donnees pour pre-train et finetune
|
Thu, 01 Apr 2010 14:25:40 -0400 |
SylvainPL |
Avoir exactement le meme jeu de donnees pour pre-train et finetune
|
Thu, 01 Apr 2010 13:44:30 -0400 |
SylvainPL |
Petits changements pour pouvoir utiliser le GPU
|
Thu, 01 Apr 2010 13:44:04 -0400 |
SylvainPL |
Petits changements pour pouvoir utiliser le GPU
|
Thu, 01 Apr 2010 13:43:43 -0400 |
SylvainPL |
Petits changements pour pouvoir utiliser le GPU
|
Thu, 01 Apr 2010 09:12:48 -0400 |
SylvainPL |
Ajout d'une ligne d'impression importante
|
Wed, 31 Mar 2010 21:03:41 -0400 |
SylvainPL |
Ajout d'une protection supplementaire (un peu de paranoia) ainsi que commentaires
|
Wed, 31 Mar 2010 21:02:27 -0400 |
SylvainPL |
Meilleure souplesse d'execution, un parametre hard-coade est maintenant plus propre
|
Wed, 31 Mar 2010 21:00:59 -0400 |
SylvainPL |
Ajout de commentaires ainsi que correction de bug pour sauvegarder les parametres
|
Wed, 31 Mar 2010 16:06:55 -0400 |
xaviermuller |
added type 2 adaptive learning configurable learning weight + versionning
|
Wed, 31 Mar 2010 15:54:47 -0400 |
Arnaud Bergeron |
Use sigmoids with cross-entropy cost in the ConvAutoencoders.
|
Tue, 30 Mar 2010 14:40:54 -0400 |
Arnaud Bergeron |
Fix concatenation of 1-dim datasets (such as int target vectors).
|
Mon, 29 Mar 2010 18:14:30 -0400 |
Arnaud Bergeron |
Fix last-minute bugs in the code.
|
Mon, 29 Mar 2010 17:54:01 -0400 |
Arnaud Bergeron |
Adjust sgd_opt parameters and use nist_all.
|
Mon, 29 Mar 2010 17:42:44 -0400 |
Arnaud Bergeron |
branch merge
|
Mon, 29 Mar 2010 17:36:22 -0400 |
Arnaud Bergeron |
- Adapt to scdae to input_shape change in pynnet
|
Wed, 10 Feb 2010 11:15:04 -0500 |
Arnaud Bergeron |
&& -> and
|
Wed, 10 Feb 2010 10:58:44 -0500 |
Arnaud Bergeron |
Fix testmod so that modules with no parameters are accepted.
|
Mon, 29 Mar 2010 09:18:54 -0400 |
Guillaume Sicard |
modifié setup_batches.py pour compatibilité avec mlp_nist.py
|
Sat, 27 Mar 2010 13:39:48 -0400 |
Arnaud Bergeron |
Save best valid score and test score in the db.
|
Fri, 26 Mar 2010 19:18:03 -0400 |
Arnaud Bergeron |
Add some prints to see the current step.
|
Fri, 26 Mar 2010 18:49:27 -0400 |
Arnaud Bergeron |
Fix stuff (imports, ...) so that it can run under jobman properly.
|
Fri, 26 Mar 2010 18:35:23 -0400 |
Arnaud Bergeron |
Split out the run_exp method.
|
Fri, 26 Mar 2010 17:35:23 -0400 |
Arnaud Bergeron |
Add commas, since that way it's not a syntax error anymore.
|
Fri, 26 Mar 2010 17:25:06 -0400 |
Arnaud Bergeron |
branch merge
|
Fri, 26 Mar 2010 17:24:17 -0400 |
Arnaud Bergeron |
Add net saving (only the best model) and error saving using SeriesTable
|
Fri, 26 Mar 2010 16:17:56 -0400 |
SylvainPL |
Programme pour reprendre une partie des experiences seulement. Utile seulement pour un usage tres specifique
|
Thu, 25 Mar 2010 12:20:27 -0400 |
SylvainPL |
correction d'un bug pour affichage des resultats de pre-train avec P07
|
Wed, 24 Mar 2010 15:14:24 -0400 |
fsavard |
Merge
|
Wed, 24 Mar 2010 15:13:48 -0400 |
fsavard |
Enlevé mécanique pour limiter le nombre d'exemples utilisés (remplacé par paramètre dans l'appel au code de dataset), et ajouté option pour sauvegarde des poids à la fin de l'entraînement
|
Wed, 24 Mar 2010 14:58:58 -0400 |
SylvainPL |
correction d'un bug sur l'indice des mini-batches
|
Wed, 24 Mar 2010 14:45:02 -0400 |
SylvainPL |
rajout de methode reliant toutes les couches cachees a la logistic et changeant seulement les parametres de la logistic durant finetune
|
Wed, 24 Mar 2010 14:44:41 -0400 |
SylvainPL |
rajout de methode reliant toutes les couches cachees a la logistic et changeant seulement les parametres de la logistic durant finetune
|
Wed, 24 Mar 2010 14:44:24 -0400 |
SylvainPL |
rajout de methode reliant toutes les couches cachees a la logistic et changeant seulement les parametres de la logistic durant finetune
|
Wed, 24 Mar 2010 14:36:55 -0400 |
fsavard |
Merge
|
Wed, 24 Mar 2010 14:35:11 -0400 |
fsavard |
Ajouté fonctionnalité pour réinsérer de nouvelles jobs aux paramètres choisis manuellement, en liste, pour repartir des jobs spécifiques qui ont crashé
|
Mon, 22 Mar 2010 16:37:34 -0400 |
Arnaud Bergeron |
Use softmax for the output layer and rework the dset iterator stuff.
|
Mon, 22 Mar 2010 13:33:29 -0400 |
Arnaud Bergeron |
Add reworked code for convolutional auto-encoder.
|
Mon, 22 Mar 2010 10:20:10 -0400 |
Xavier Glorot |
merge
|
Mon, 22 Mar 2010 10:19:45 -0400 |
Xavier Glorot |
small change in generating script (bug fix)
|
Sun, 21 Mar 2010 16:45:48 -0400 |
Xavier Glorot |
changes in ttf2jpg and pycaptcha to load a file list with cPickle (instead of doing a listdir()) in order to have always the same list order from different machine (to reproduce generation)
|
Mon, 22 Mar 2010 08:34:48 -0400 |
Guillaume Sicard |
ajouté : module de préparation de batches en fonction d'un ratio de classes
|
Sun, 21 Mar 2010 19:30:05 -0400 |
Yoshua Bengio |
fixed a problem with maxsize when not provided
|
Sat, 20 Mar 2010 15:49:55 -0400 |
Jeremy Eustache |
Reseau a convolution avec le bon dataset
|
Sat, 20 Mar 2010 10:19:11 -0400 |
SylvainPL |
Ajout d'une fonctionnalite pour pouvoir definir un range de fichiers d'entrainement de P07 utilises. Utile pour pre-entrainer et fine-tuner avec des donnees differentes
|
Fri, 19 Mar 2010 11:31:57 -0400 |
fsavard |
Légères corrections à config.py.example et ajusté nist_sda pour utiliser la nouvelle 'interface' des datasets (nist_all()).
|
Fri, 19 Mar 2010 11:12:40 -0400 |
fsavard |
Modifs à nist_sda.py et config.py.example pour corrections viz le mécanisme pour isolation d'expérience
|
Fri, 19 Mar 2010 10:56:16 -0400 |
fsavard |
Merge. Ah, et dans le dernier commit, j'avais oublié de mentionner que j'ai ajouté du code pour gérer l'isolation de différents clones pour rouler des expériences et modifier le code en même temps.
|
Fri, 19 Mar 2010 10:54:39 -0400 |
fsavard |
Déplacer le nouveau code de stacked_dae de v2 vers le répertoire de base 'stacked_dae', et bougé le vieux code vers le répertoire 'old'
|
Thu, 18 Mar 2010 10:53:02 -0400 |
SylvainPL |
ajout de fonctionnalitees pour deux testsets. Changements afin de pouvoir utiliser la nouvelle version de datasets
|
Thu, 18 Mar 2010 10:52:22 -0400 |
SylvainPL |
ajout de fonctionnalitees pour deux testsets
|
Wed, 17 Mar 2010 16:41:51 -0400 |
Xavier Glorot |
merge
|
Wed, 17 Mar 2010 16:41:16 -0400 |
Xavier Glorot |
important bug fix in pipeline corresponding to an iterator always giving 0 complexity, need to redo 10% of the P07 dataset
|
Wed, 17 Mar 2010 15:31:21 -0400 |
Arnaud Bergeron |
branch merge.
|
Wed, 17 Mar 2010 15:24:25 -0400 |
Arnaud Bergeron |
Make img_size a parameter, and remove the passing of the image size to the ConvOp. This will have to get back in later somehow.
|
Tue, 16 Mar 2010 19:05:59 -0400 |
Arnaud Bergeron |
maxsize -> self.maxsize
|
Tue, 16 Mar 2010 18:51:27 -0400 |
Arnaud Bergeron |
Make the datasets lazy-loading and add a maxsize parameter.
|
Wed, 17 Mar 2010 14:04:12 -0400 |
Xavier Glorot |
merge
|
Wed, 17 Mar 2010 14:01:23 -0400 |
Xavier Glorot |
added PNIST generation script
|
Wed, 17 Mar 2010 13:57:15 -0400 |
Xavier Glorot |
added option to pipeline and gimp_script to produce NIST-friendly data
|
Wed, 17 Mar 2010 10:21:57 -0400 |
Jeremy Eustache |
Derniere version du reseau a convolution
|
Tue, 16 Mar 2010 21:24:41 -0400 |
SylvainPL |
ajout de fonctionnalite pour different finetune dataset
|
Tue, 16 Mar 2010 21:24:30 -0400 |
SylvainPL |
ajout de fonctionnalite pour different finetune dataset
|
Tue, 16 Mar 2010 21:24:09 -0400 |
SylvainPL |
ajout de fonctionnalite pour different finetune dataset
|
Tue, 16 Mar 2010 20:19:13 -0400 |
humel |
Fixed a bug with image_shape.
|
Tue, 16 Mar 2010 14:46:25 -0400 |
humel |
Optimized the call of ConvOp by specifying additional parameters. Specified image shape of the da_conv layer.
|
Tue, 16 Mar 2010 13:16:28 -0400 |
humel |
Fixed dataset import. Removed unuseful code from da_conv. Keys parameters are now passed as arguments.
|
Tue, 16 Mar 2010 12:14:10 -0400 |
Xavier Glorot |
merge
|
Tue, 16 Mar 2010 12:13:49 -0400 |
Xavier Glorot |
change in testtransformation to test PNIST pipeline
|
Tue, 16 Mar 2010 12:12:20 -0400 |
Xavier Glorot |
small change in gimp_script to avoid blurring in PNIST
|
Tue, 16 Mar 2010 12:01:31 -0400 |
fsavard |
Merge
|
Tue, 16 Mar 2010 12:00:51 -0400 |
fsavard |
Enlevé printout de débuggage
|
Tue, 16 Mar 2010 11:21:55 -0400 |
Myriam Cote |
fix call to log_reg
|
Tue, 16 Mar 2010 10:52:04 -0400 |
fsavard |
Corrigé bugs dans stacked_dae/v2 en rapport avec les modifs d'hier
|
Mon, 15 Mar 2010 18:30:21 -0400 |
fsavard |
Mergé (manuellement) les changements de Sylvain pour utiliser le code de dataset d'Arnaud, à cette différence près que je n'utilse pas les givens. J'ai probablement une approche différente pour limiter la taille du dataset dans mon débuggage, aussi.
|
Mon, 15 Mar 2010 13:22:20 -0400 |
SylvainPL |
Possibilite de restreindre la taille des ensemble d'entrainement, valid et test afin de pouvoir tester le code rapidement
|
Mon, 15 Mar 2010 10:09:50 -0400 |
xaviermuller |
corrected a bug in jobman interface
|
Mon, 15 Mar 2010 09:22:52 -0400 |
Myriam Cote |
added __init__.py to allow module loading of baseline
|
Sun, 14 Mar 2010 20:25:12 -0400 |
SylvainPL |
correction de bugs
|
Sun, 14 Mar 2010 15:17:04 -0400 |
SylvainPL |
Changement de la base de donnees qui sera utilisee
|
Sun, 14 Mar 2010 15:07:17 -0400 |
SylvainPL |
version pour utilisation du module dataset
|
Sat, 13 Mar 2010 15:45:43 -0500 |
Dumitru Erhan |
merge
|
Sat, 13 Mar 2010 15:44:50 -0500 |
Dumitru Erhan |
Added the possibility to get the paths from an env. variable + cleaned up the way we build the paths
|
Fri, 12 Mar 2010 16:47:10 -0500 |
SylvainPL |
Repertoire pour faire les tests avec les differents ensembles pour le finetuning
|
Fri, 12 Mar 2010 11:03:51 -0500 |
fsavard |
Corrigé bug avec nouvelle formule de coût et printout pour cas sur CPU
|
Fri, 12 Mar 2010 10:47:36 -0500 |
fsavard |
Corrigé une erreur dans la formule de coût modifiée dans stacked_dae, et enlevé des timers dans sgd_optimization
|
Fri, 12 Mar 2010 10:31:10 -0500 |
fsavard |
Ajouté une modification sur stacked_dae qui utilise les nouvelles SeriesTables. Je le met dans le repository pour que mes expériences en cours continuent sans perturbation, et pour que Sylvain puisse récupérer la version actuelle; je fusionnerai à moment donné.
|
Fri, 12 Mar 2010 09:02:40 -0500 |
fsavard |
Test pour cas sans index, après question de James
|
Thu, 11 Mar 2010 16:29:39 -0500 |
fsavard |
Merge
|
Thu, 11 Mar 2010 11:52:43 -0500 |
fsavard |
Corrigé un bug avec store_timestamp/cpuclock, et tests pour éviter ce cas
|
Thu, 11 Mar 2010 14:42:54 -0500 |
Arnaud Bergeron |
Branch merge
|
Thu, 11 Mar 2010 14:41:31 -0500 |
Arnaud Bergeron |
Add mnist to datasets (and supporting code).
|
Thu, 11 Mar 2010 11:08:42 -0500 |
fsavard |
Encore quelques typechecks à SeriesTables.
|
Thu, 11 Mar 2010 10:48:54 -0500 |
fsavard |
Ajouté un paquet de type/value checks à SeriesTables, et finalisé les docstrings. Ajouté 3-4 tests. Légers refactorings ici et là sans conséquences externes.
|
Wed, 10 Mar 2010 20:14:20 -0500 |
fsavard |
Merge
|
Wed, 10 Mar 2010 20:13:45 -0500 |
fsavard |
Modifications pour stocker des timestamps/cpuclock avec chaque rangée créée, selon suggestion de Yoshua ce matin
|
Wed, 10 Mar 2010 17:08:50 -0500 |
Dumitru Erhan |
merge, second try
|
Wed, 10 Mar 2010 17:08:27 -0500 |
Dumitru Erhan |
merge
|
Wed, 10 Mar 2010 13:48:16 -0500 |
Dumitru Erhan |
Changes that enable using this code when floatX=float32
|
Wed, 10 Mar 2010 17:01:02 -0500 |
Dumitru Erhan |
The in/outscale parameters were not passed to FTData
|
Wed, 10 Mar 2010 16:52:22 -0500 |
fsavard |
Renommé mon module de séries
|
Wed, 10 Mar 2010 16:17:59 -0500 |
xaviermuller |
added classic lr time decay and py code to calculate the error based on a saved model
|
Wed, 10 Mar 2010 09:56:02 -0500 |
Arnaud Bergeron |
Add nist_P07 datasets to the predefs.
|
Tue, 09 Mar 2010 10:15:19 -0500 |
savardf |
Commented table_series code, changed ParamsStatisticsArray to take shared params instead, create DummySeries to use when we don't want to save a named series
|
Fri, 05 Mar 2010 18:08:34 -0500 |
fsavard |
Merge
|
Fri, 05 Mar 2010 18:07:20 -0500 |
fsavard |
Completely rewrote my series module, now based on HDF5 and PyTables (in a separate directory called 'tables_series' for retrocompatibility of running code). Minor (inconsequential) changes to stacked_dae.
|
Thu, 04 Mar 2010 20:43:21 -0500 |
Arnaud Bergeron |
Merge branches from main repo.
|
Thu, 04 Mar 2010 09:43:23 -0500 |
youssouf |
removed one bug: the complexity value was not sent to the dA class
|
Thu, 04 Mar 2010 08:21:43 -0500 |
fsavard |
Merge
|
Thu, 04 Mar 2010 08:18:42 -0500 |
fsavard |
Changé le coût de reconstruction pour stabilité numérique, en ajoutant une petite constante dans le log.
|
Thu, 04 Mar 2010 00:54:12 -0500 |
boulanni |
Actual script used to launch images generation on 4 machines
|
Wed, 03 Mar 2010 16:52:10 -0500 |
Arnaud Bergeron |
Branch merge.
|
Wed, 03 Mar 2010 16:46:16 -0500 |
Arnaud Bergeron |
Branch merge
|
Tue, 02 Mar 2010 20:16:30 -0500 |
Arnaud Bergeron |
Adapt the sdae code for ift6266.datasets input.
|
Tue, 02 Mar 2010 18:43:54 -0500 |
Arnaud Bergeron |
Add MSE cost to log_reg.py
|
Tue, 02 Mar 2010 18:16:49 -0500 |
Arnaud Bergeron |
Modify the log_reg.py tutorial code to use the datasets module.
|
Tue, 02 Mar 2010 18:03:42 -0500 |
Arnaud Bergeron |
Add __init__.py files to the remaining directories that miss them.
|
Tue, 02 Mar 2010 18:03:09 -0500 |
Arnaud Bergeron |
Fix import errors (conv and downsample are no longer in sandbox).
|
Tue, 02 Mar 2010 18:01:22 -0500 |
Arnaud Bergeron |
Add option for test.py to test modules specified on the command-line.
|
Tue, 02 Mar 2010 17:28:14 -0500 |
Arnaud Bergeron |
Fix import of scalar_series.
|
Tue, 02 Mar 2010 17:00:59 -0500 |
Arnaud Bergeron |
Add option to test a module given on the command line.
|
Wed, 03 Mar 2010 12:51:40 -0500 |
fsavard |
Commented a few things, renamed the produit_croise_jobs function, replaced the cost function (NOT TESTED YET).
|
Tue, 02 Mar 2010 14:47:18 -0500 |
fsavard |
Ajouts mineurs à stacked_dae, juste printé l'heure je crois.
|
Tue, 02 Mar 2010 09:52:27 -0500 |
youssouf |
initial commit for autoencoder training
|
Mon, 01 Mar 2010 17:06:49 -0500 |
Arnaud Bergeron |
Fix bug where there would be a bunch of 0-length batches at the end under certain circumstances.
|
Mon, 01 Mar 2010 11:46:39 -0500 |
fsavard |
Ajouté __init__.py pour utils
|
Mon, 01 Mar 2010 11:45:59 -0500 |
fsavard |
Merge
|
Mon, 01 Mar 2010 11:45:25 -0500 |
fsavard |
Ajout du code de base pour scalar_series. Modifications à stacked_dae: réglé un problème avec les input_divider (empêchait une optimisation), et ajouté utilisation des séries. Si j'avais pas déjà commité, aussi, j'ai enlevé l'histoire de réutilisation du pretraining: c'était compliqué (error prone) et ça créait des jobs beaucoup trop longues.
|
Fri, 26 Feb 2010 17:45:52 -0500 |
fsavard |
Enlevé ce qui concernait la réutilisation de résultats de préentraînement (trop compliqué pour peu de bénéfice: c'est le finetuning qui est vraiment long
|
Sat, 27 Feb 2010 19:26:26 -0500 |
boulanni |
Transient exception handling in captchas (ie. lorsque le NFS est temporairement inaccessible)
|
Sat, 27 Feb 2010 19:21:57 -0500 |
boulanni |
Adapted pipeline to new directory structure
|
Sat, 27 Feb 2010 18:28:48 -0500 |
boulanni |
J'ai reséparé NIST/OCR purs pour avoir des ensembles de test et de validation de 80000 plutôt que 20000, comme on a discuté au cours
|
Sat, 27 Feb 2010 17:10:37 -0500 |
Arnaud Bergeron |
Remove a stray cast in the FTDataSet code and export the ocr dataset.
|
Sat, 27 Feb 2010 16:50:16 -0500 |
Arnaud Bergeron |
Add dtype conversion and rescaling to the read path.
|
Sat, 27 Feb 2010 16:07:09 -0500 |
Arnaud Bergeron |
Do not yield theano shared variables. They can only be used by theano.function().
|
Sat, 27 Feb 2010 15:09:02 -0500 |
Arnaud Bergeron |
Make the datasets iterators return theano shared slices with the appropriate types.
|
Sat, 27 Feb 2010 14:15:11 -0500 |
Arnaud Bergeron |
Use izip(), not zip() to return a lazy iterator. (datasets)
|
Sat, 27 Feb 2010 14:10:14 -0500 |
Arnaud Bergeron |
Update comments in the dataset definition (you can't pass 0 as minibatch size).
|
Sat, 27 Feb 2010 13:56:14 -0500 |
Arnaud Bergeron |
Define the ocr dataset and use the existing split for nist.
|
Sat, 27 Feb 2010 12:18:26 -0500 |
Arnaud Bergeron |
Make data_generation.transformations importable and fixup test.py to not try some of the modules.
|
Sat, 27 Feb 2010 12:01:08 -0500 |
Arnaud Bergeron |
Take the validation set at the end of the training set files rather than at the beginning.
|
Fri, 26 Feb 2010 15:25:44 -0500 |
Dumitru Erhan |
ignoring certain files when doing hg status
|
Fri, 26 Feb 2010 15:25:01 -0500 |
Dumitru Erhan |
making the data_generation folder (and subdirs) an importable module
|
Fri, 26 Feb 2010 14:28:48 -0500 |
Dumitru Erhan |
moved deepmlp code into baseline/deep_mlp
|
Fri, 26 Feb 2010 14:24:11 -0500 |
Dumitru Erhan |
directory name change
|
Fri, 26 Feb 2010 14:23:47 -0500 |
Dumitru Erhan |
Pipeline code shuffle
|
Fri, 26 Feb 2010 14:15:38 -0500 |
Dumitru Erhan |
More moves - transformations into data_generation, added "deep" folder
|
Fri, 26 Feb 2010 14:03:24 -0500 |
Dumitru Erhan |
Moving the convolutional MLP code into baseline
|
Fri, 26 Feb 2010 13:55:27 -0500 |
Dumitru Erhan |
Updating the tutorial code to the latest revisions.
|
Thu, 25 Feb 2010 18:53:02 -0500 |
Arnaud Bergeron |
Add nist_lower, nist_upper, nist_all
|
Thu, 25 Feb 2010 18:40:01 -0500 |
Arnaud Bergeron |
Add inital implementation of datasets.
|
Thu, 25 Feb 2010 18:12:18 -0500 |
Arnaud Bergeron |
Do not commit syntax errors. (in test.py)
|
Thu, 25 Feb 2010 18:11:25 -0500 |
Arnaud Bergeron |
Make test not test itself.
|
Thu, 25 Feb 2010 17:27:49 -0500 |
Arnaud Bergeron |
Simple script to doctest all modules beneath ift6266.
|
Thu, 25 Feb 2010 09:05:48 -0500 |
Myriam Cote |
merge
|
Thu, 25 Feb 2010 09:04:40 -0500 |
Myriam Cote |
nouveau répertoire régression logistique
|
Wed, 24 Feb 2010 19:27:38 -0500 |
boulanni |
Ajouté seed 100 à 207 aux ensembles de données générés
|
Wed, 24 Feb 2010 19:12:01 -0500 |
boulanni |
On peut maintenant launcher le pipeline avec un seed donné, résultats déterministes
|
Wed, 24 Feb 2010 18:27:09 -0500 |
boulanni |
Enlevé les re-seed constants, interfère avec les autres modules
|
Wed, 24 Feb 2010 16:59:04 -0500 |
boulanni |
Checked all modules to work with only numpy.random and random and to be deterministic after numpy.random.seed() and random.seed()
|
Wed, 24 Feb 2010 13:51:18 -0500 |
SylvainPL |
Script utilise pour creer le jeu de donnees utilise pour le module Occlusion.py. De plus, le fichier /data/lisa/data/ift6266h10/echantillon_occlusion.py a change afin d'etre cree avec seed random (jeu reproductible)
|
Wed, 24 Feb 2010 13:16:14 -0500 |
SylvainPL |
Rajout d'un seed random et d'une fonction get_seed
|
Wed, 24 Feb 2010 13:15:52 -0500 |
SylvainPL |
Rajout d'un seed random et d'une fonction get_seed
|
Wed, 24 Feb 2010 13:15:28 -0500 |
SylvainPL |
Rajout d'un seed random et d'une fonction get_seed
|
Wed, 24 Feb 2010 13:14:55 -0500 |
SylvainPL |
Rajout d'un seed random et d'une fonction get_seed.Aussi, import random enleve car non necessaire
|
Wed, 24 Feb 2010 13:14:02 -0500 |
SylvainPL |
Rajout d'un seed random et d'une fonction get_seed
|
Wed, 24 Feb 2010 13:13:33 -0500 |
SylvainPL |
Rajout d'un seed random et d'une fonction get_seed
|
Wed, 24 Feb 2010 12:44:39 -0500 |
Jeremy Eustache |
Reseau a convolution
|