Merge pull request #5635 from nouiz/comp_opt

doc/library/config.txt

@@ -693,6 +693,13 @@ import theano and print the config variable, as in:
 
     If ``'False'``, do not use cuDNN or check if it is available.
 
+.. attribute:: config.conv.assert_shape
+
+    If False, AbstractConv* ops won't add assert that verify that
+    the user provided shapes are also the one at run time.
+
+    This can speed up compilation time and/or execution time.
+
 .. attribute:: config.dnn.conv.workmem
 
     Deprecated, use :attr:`config.dnn.conv.algo_fwd`.

theano/configdefaults.py

@@ -126,6 +126,13 @@ AddConfigVar(
     BoolParam(False, allow_override=False),
     in_c_key=False)
 
+AddConfigVar(
+    'conv.assert_shape',
+    "If False, AbstractConv* ops won't add assert that verify that"
+    " the user provided shapes are also the one at run time",
+    BoolParam(True),
+    in_c_key=False)
+
 AddConfigVar(
     'print_global_stats',
     "Print some global statistics (time spent) at the end",

theano/gof/opt.py

@@ -607,11 +607,20 @@ class MergeFeature(object):
         # properly.
         # The clients should at least contain `node` itself!
         if node.inputs:
-            assert len(node.inputs[0].clients) > 0
-            assert (node, 0) in node.inputs[0].clients
+            # Take the smallest clients list. Some ops like elemwise
+            # have optimization that put constant as the first inputs.
+            # As constant have in general more clients than other type of nodes
+            # using always inputs[0] make us look at more nodes.
+            # Always pick the smallest clints list between inputs 0
+            # and -1 speed up optimization.
 
-            merge_candidates = [c for (c, i) in node.inputs[0].clients
-                                if c in self.nodes_seen]
+            if len(node.inputs[0].clients) < len(node.inputs[-1].clients):
+                clients = node.inputs[0].clients
+            else:
+                clients = node.inputs[-1].clients
+            assert len(clients) > 0
+
+            merge_candidates = [c for c, i in clients if c in self.nodes_seen]
 
             # Put all clients of Assert inputs (if exist) into merge_candidates
             # TODO: Deactivated for now as this cause cycle in the graph.

theano/gpuarray/basic_ops.py

@@ -270,12 +270,13 @@ class Kernel(object):
 def get_ctype(dtype):
     if dtype is gpuarray.GpuArray:
         return "gpudata *"
+    elif isinstance(dtype, np.dtype):
+        return 'npy_' + dtype.name
     elif dtype == gpuarray.SIZE:
         return "size_t"
     elif dtype == gpuarray.SSIZE:
         return "ssize_t"
     else:
-        if not isinstance(dtype, np.dtype):
         dtype = np.dtype(dtype)
         return 'npy_' + dtype.name
 

theano/scan_module/scan_utils.py

@@ -1353,4 +1353,6 @@ def forced_replace(out, x, y):
         elif graph.owner:
             q.extendleft(graph.owner.inputs)
 
+    if len(to_replace) == 0:
+        return out
     return clone(out, replace=to_replace)

theano/tensor/nnet/abstract_conv.py

@@ -493,7 +493,7 @@ def assert_shape(x, expected_shape, msg='Unexpected shape.'):
         will return `x` directly.
 
     """
-    if expected_shape is None:
+    if expected_shape is None or not theano.config.conv.assert_shape:
         return x
     shape = x.shape
     tests = []
@@ -1680,19 +1680,20 @@ class AbstractConv2d(AbstractConv):
     def grad(self, inp, grads):
         bottom, weights = inp
         top, = grads
+        # Don't add the assert again, as it was already added in the forward.
         d_bottom = AbstractConv2d_gradInputs(self.imshp, self.kshp,
                                              self.border_mode,
                                              self.subsample,
                                              self.filter_flip,
                                              self.filter_dilation)(
-            weights, top, bottom.shape[-2:])
+            weights, top, bottom.shape[-2:], add_assert_shape=False)
         d_weights = AbstractConv2d_gradWeights(self.imshp, self.kshp,
                                                self.border_mode,
                                                self.subsample,
                                                self.filter_flip,
                                                self.filter_dilation)(
 
-            bottom, top, weights.shape[-2:])
+            bottom, top, weights.shape[-2:], add_assert_shape=False)
 
         # Make sure that the broadcastable pattern of the inputs is used
         # for the gradients, even if the grad opts are not able to infer
@@ -1781,7 +1782,7 @@ class AbstractConv_gradWeights(BaseAbstractConv):
                                                        filter_dilation=filter_dilation)
 
     # Update shape/height_width
-    def make_node(self, img, topgrad, shape):
+    def make_node(self, img, topgrad, shape, add_assert_shape=True):
         # Make sure both inputs are Variables with the same Type
         if not isinstance(img, theano.Variable):
             img = as_tensor_variable(img)
@@ -1795,7 +1796,7 @@ class AbstractConv_gradWeights(BaseAbstractConv):
             raise TypeError('img must be %dD tensor' % (2 + self.convdim))
         if topgrad.type.ndim != 2 + self.convdim:
             raise TypeError('topgrad must be %dD tensor' % (2 + self.convdim))
-
+        if add_assert_shape:
             img = assert_shape(img, self.imshp,
                                'AbstractConv_gradWeights shape mismatch: shape of '
                                'image does not match given imshp.')
@@ -2020,7 +2021,7 @@ class AbstractConv_gradInputs(BaseAbstractConv):
                                                       filter_dilation=filter_dilation)
 
     # Update shape/height_width
-    def make_node(self, kern, topgrad, shape):
+    def make_node(self, kern, topgrad, shape, add_assert_shape=True):
         # Make sure both inputs are Variables with the same Type
         if not isinstance(kern, theano.Variable):
             kern = as_tensor_variable(kern)
@@ -2035,6 +2036,7 @@ class AbstractConv_gradInputs(BaseAbstractConv):
         if topgrad.type.ndim != 2 + self.convdim:
             raise TypeError('topgrad must be %dD tensor' % (2 + self.convdim))
 
+        if add_assert_shape:
             kern = assert_shape(kern, self.kshp,
                                 'AbstractConv_gradInputs shape mismatch: shape of '
                                 'filters does not match given kshp.')
@@ -2158,7 +2160,8 @@ class AbstractConv2d_gradInputs(AbstractConv_gradInputs):
                                                self.border_mode,
                                                self.subsample,
                                                self.filter_flip,
-                                               self.filter_dilation)(bottom, top,
+                                               self.filter_dilation)(
+                                                   bottom, top,
                                                    weights.shape[-2:])
         d_top = AbstractConv2d(self.imshp, self.kshp,
                                self.border_mode,

theano/tensor/opt.py

@@ -2418,6 +2418,10 @@ compile.optdb['specialize'].register('local_remove_all_assert',
                                      local_remove_all_assert,
                                      'unsafe',
                                      use_db_name_as_tag=False)
+compile.optdb['useless'].register('local_remove_all_assert',
+                                  local_remove_all_assert,
+                                  'unsafe',
+                                  use_db_name_as_tag=False)
 
 #######################
 # Constant Canonicalization