Mercurial > ift6266
changeset 235:ecb69e17950b
correction de bugs
| author | SylvainPL <sylvain.pannetier.lebeuf@umontreal.ca> |
|---|---|
| date | Sun, 14 Mar 2010 20:25:12 -0400 |
| parents | c452e3a0a3b1 |
| children | 7be1f086a89e |
| files | deep/stacked_dae/v_sylvain/nist_sda.py deep/stacked_dae/v_sylvain/sgd_optimization.py deep/stacked_dae/v_sylvain/stacked_dae.py |
| diffstat | 3 files changed, 10 insertions(+), 3 deletions(-) [+] |
line wrap: on
line diff
--- a/deep/stacked_dae/v_sylvain/nist_sda.py Sun Mar 14 15:17:04 2010 -0400 +++ b/deep/stacked_dae/v_sylvain/nist_sda.py Sun Mar 14 20:25:12 2010 -0400 @@ -120,10 +120,10 @@ n_ins=n_ins, n_outs=n_outs,\ series=series) - optimizer.pretrain() + optimizer.pretrain(datasets.nist_all) channel.save() - optimizer.finetune() + optimizer.finetune(datasets.nist_all) channel.save() return channel.COMPLETE
--- a/deep/stacked_dae/v_sylvain/sgd_optimization.py Sun Mar 14 15:17:04 2010 -0400 +++ b/deep/stacked_dae/v_sylvain/sgd_optimization.py Sun Mar 14 20:25:12 2010 -0400 @@ -118,9 +118,11 @@ # go through pretraining epochs for epoch in xrange(self.hp.pretraining_epochs_per_layer): # go through the training set + batch_index=int(0) for x,y in dataset.train(self.hp.minibatch_size): c = self.classifier.pretrain_functions[i](x) - + batch_index+=1 + self.series["reconstruction_error"].append((epoch, batch_index), c) print 'Pre-training layer %i, epoch %d, cost '%(i,epoch),c @@ -140,6 +142,8 @@ #index = T.lscalar() # index to a [mini]batch minibatch_size = self.hp.minibatch_size + ensemble_x = T.matrix('ensemble_x') + ensemble_y = T.ivector('ensemble_y') # create a function to compute the mistakes that are made by the model # on the validation set, or testing set
--- a/deep/stacked_dae/v_sylvain/stacked_dae.py Sun Mar 14 15:17:04 2010 -0400 +++ b/deep/stacked_dae/v_sylvain/stacked_dae.py Sun Mar 14 20:25:12 2010 -0400 @@ -204,6 +204,9 @@ self.x = T.matrix('x') # the data is presented as rasterized images self.y = T.ivector('y') # the labels are presented as 1D vector of # [int] labels + ensemble = T.matrix('ensemble') + ensemble_x = T.matrix('ensemble_x') + ensemble_y = T.ivector('ensemble_y') for i in xrange( self.n_layers ): # construct the sigmoidal layer