Merge pull request lisa-lab#150 from nouiz/master2

lamblin · lamblin · commit 27760353a4be · 2016-03-22T12:31:30.000-04:00
small stuff and use update interface to Theano 0.8
diff --git a/code/DBN.py b/code/DBN.py
@@ -174,7 +174,7 @@ def pretraining_functions(self, train_set_x, batch_size, k):
 
             # compile the theano function
             fn = theano.function(
-                inputs=[index, theano.Param(learning_rate, default=0.1)],
+                inputs=[index, theano.In(learning_rate, value=0.1)],
                 outputs=cost,
                 updates=updates,
                 givens={
diff --git a/code/convolutional_mlp.py b/code/convolutional_mlp.py
@@ -94,7 +94,7 @@ def __init__(self, rng, input, filter_shape, image_shape, poolsize=(2, 2)):
             input=input,
             filters=self.W,
             filter_shape=filter_shape,
-            image_shape=image_shape
+            input_shape=image_shape
         )
 
         # downsample each feature map individually, using maxpooling
diff --git a/code/rbm.py b/code/rbm.py
@@ -257,7 +257,8 @@ def get_cost_updates(self, lr=0.1, persistent=None, k=1):
             # chain_start is the initial state corresponding to the
             # 6th output
             outputs_info=[None, None, None, None, None, chain_start],
-            n_steps=k
+            n_steps=k,
+            name="gibbs_hvh"
         )
         # start-snippet-3
         # determine gradients on RBM parameters
@@ -496,7 +497,8 @@ def test_rbm(learning_rate=0.1, training_epochs=15,
     ) = theano.scan(
         rbm.gibbs_vhv,
         outputs_info=[None, None, None, None, None, persistent_vis_chain],
-        n_steps=plot_every
+        n_steps=plot_every,
+        name="gibbs_vhv"
     )
 
     # add to updates the shared variable that takes care of our persistent
diff --git a/data/download.sh b/data/download.sh
@@ -5,7 +5,7 @@ WGET=$?
 which curl >/dev/null 2>&1
 CURL=$?
 if [ "$WGET" -eq 0 ]; then
-    DL_CMD="wget -c"
+    DL_CMD="wget --no-verbose -c"
 elif [ "$CURL" -eq 0 ]; then
     DL_CMD="curl -C - -O"
 else
diff --git a/doc/lenet.txt b/doc/lenet.txt
@@ -196,7 +196,7 @@ one of Figure 1. The input consists of 3 features maps (an RGB color image) of s
 
         import theano
         from theano import tensor as T
-        from theano.tensor.nnet import conv
+        from theano.tensor.nnet import conv2d
 
         import numpy
 
@@ -226,7 +226,7 @@ one of Figure 1. The input consists of 3 features maps (an RGB color image) of s
                     dtype=input.dtype), name ='b')
 
         # build symbolic expression that computes the convolution of input with filters in w
-        conv_out = conv.conv2d(input, W)
+        conv_out = conv2d(input, W)
 
         # build symbolic expression to add bias and apply activation function, i.e. produce neural net layer output
         # A few words on ``dimshuffle`` :
@@ -404,7 +404,7 @@ to be compatible with our previous MLP implementation.
     Note that the term "convolution" could corresponds to different mathematical operations:
 
     1. `theano.tensor.nnet.conv2d
-       <http://deeplearning.net/software/theano/library/tensor/nnet/conv.html#theano.tensor.nnet.conv.conv2d>`_,
+       <http://deeplearning.net/software/theano/library/tensor/nnet/conv.html#theano.tensor.nnet.conv2d>`_,
        which is the most common one in almost all of the recent published
        convolutional models.
        In this operation, each output feature map is connected to each

Original file line number	Diff line number	Diff line change
`@@ -94,7 +94,7 @@ def __init__(self, rng, input, filter_shape, image_shape, poolsize=(2, 2)):`
`94`	`94`	`input=input,`
`95`	`95`	`filters=self.W,`
`96`	`96`	`filter_shape=filter_shape,`
`97`		`- image_shape=image_shape`
	`97`	`+ input_shape=image_shape`
`98`	`98`	`)`
`99`	`99`
`100`	`100`	`# downsample each feature map individually, using maxpooling`