Fix doc

theano/sandbox/cuda/blas.py

@@ -708,8 +708,10 @@ class GpuConv(GpuOp):
                         block size dimensions 0 (blockDim.x) used by the
                         GPU function.
         :param fft_opt: desactivate fft_opt optimization at the op level when
-                        set to False.
-
+                        set to False. Note that by default fft optimization
+                        aren't enabled. See
+                        :ref:convolution documentation <libdoc_tensor_nnet_conv>
+                        to enable them.
         """
         self.border_mode = border_mode
         self.subsample = subsample

theano/tensor/nnet/conv.py

@@ -350,7 +350,7 @@ class ConvOp(OpenMPOp):
         :param verbose: passed to GpuConv
         :type version: int or str
         :param version: passed to GpuConv, if version='no_fft', fft
-            optimization will be desactivated the at the op level.
+            optimization will be desactivated at the op level.
 
         The 3 following parameters are used internally when we generate
         the gradient when dx!=1 or dy!=1.