Make order and name of init parameters consistent.

Author: lucasb-eyer

beacon8/layers/Linear.py

@@ -8,16 +8,16 @@
 
 class Linear(Module):
 
-    def __init__(self, nin, nout, init=xavier(), with_bias=True, init_b=zero):
+    def __init__(self, nin, nout, with_bias=True, initW=xavier(), initB=zero):
         Module.__init__(self)
 
         self.nin = nin
         self.nout = nout
         self.with_bias = with_bias
 
-        self.weight, self.grad_weight = create_param_and_grad((nin, nout), init, fan=(nin, nout), name='Wlin_{}x{}'.format(nin, nout))
+        self.weight, self.grad_weight = create_param_and_grad((nin, nout), initW, fan=(nin, nout), name='Wlin_{}x{}'.format(nin, nout))
         if self.with_bias:
-            self.bias, self.grad_bias = create_param_and_grad(nout, init_b, name='blin_{}'.format(nout))
+            self.bias, self.grad_bias = create_param_and_grad(nout, initB, name='blin_{}'.format(nout))
 
     def symb_forward(self, symb_input):
         out = _th.tensor.dot(symb_input, self.weight)

beacon8/layers/SpatialConvolution.py

@@ -7,7 +7,7 @@
 
 
 class SpatialConvolution(Module):
-    def __init__(self, n_input_plane, n_output_plane, k_w, k_h, d_w=1, d_h=1, with_bias=True, border_mode='valid', imshape=None, init=xavier(), init_b=zero):
+    def __init__(self, n_input_plane, n_output_plane, k_w, k_h, d_w=1, d_h=1, with_bias=True, initW=xavier(), initB=zero, border_mode='valid', imshape=None):
         Module.__init__(self)
         self.n_input_plane = n_input_plane
         self.n_output_plane = n_output_plane
@@ -22,9 +22,9 @@ def __init__(self, n_input_plane, n_output_plane, k_w, k_h, d_w=1, d_h=1, with_b
         self.w_shape = (n_output_plane, n_input_plane, k_h, k_w)
         w_fan = (n_input_plane*k_w*k_h, n_output_plane*k_w*k_h)
 
-        self.weight, self.grad_weight = create_param_and_grad(self.w_shape, init, fan=w_fan, name='Wconv_{},{}@{}x{}'.format(n_input_plane, n_output_plane, k_w, k_h))
+        self.weight, self.grad_weight = create_param_and_grad(self.w_shape, initW, fan=w_fan, name='Wconv_{},{}@{}x{}'.format(n_input_plane, n_output_plane, k_w, k_h))
         if self.with_bias:
-            self.bias, self.grad_bias = create_param_and_grad(n_output_plane, init_b, name='bconv_{}'.format(n_output_plane))
+            self.bias, self.grad_bias = create_param_and_grad(n_output_plane, initB, name='bconv_{}'.format(n_output_plane))
 
     def symb_forward(self, symb_input):
         conv_output = _th.tensor.nnet.conv.conv2d(symb_input, self.weight,

beacon8/layers/SpatialConvolutionCUDNN.py

@@ -8,7 +8,7 @@
 
 
 class SpatialConvolutionCUDNN(Module):
-    def __init__(self, n_input_plane, n_output_plane, k_w, k_h, d_w=1, d_h=1, pad_w=0, pad_h=0, with_bias=True, init=xavier(), init_b=zero):
+    def __init__(self, n_input_plane, n_output_plane, k_w, k_h, d_w=1, d_h=1, pad_w=0, pad_h=0, with_bias=True, initW=xavier(), initB=zero):
         Module.__init__(self)
         self.n_input_plane = n_input_plane
         self.n_output_plane = n_output_plane
@@ -23,9 +23,9 @@ def __init__(self, n_input_plane, n_output_plane, k_w, k_h, d_w=1, d_h=1, pad_w=
         w_shape = (n_output_plane, n_input_plane, k_h, k_w)
         w_fan = (n_input_plane*k_w*k_h, n_output_plane*k_w*k_h)
 
-        self.weight, self.grad_weight = create_param_and_grad(w_shape, init, fan=w_fan, name='Wconv_{},{}@{}x{}'.format(n_input_plane, n_output_plane, k_w, k_h))
+        self.weight, self.grad_weight = create_param_and_grad(w_shape, initW, fan=w_fan, name='Wconv_{},{}@{}x{}'.format(n_input_plane, n_output_plane, k_w, k_h))
         if self.with_bias:
-            self.bias, self.grad_bias = create_param_and_grad(n_output_plane, init_b, name='bconv_{}'.format(n_output_plane))
+            self.bias, self.grad_bias = create_param_and_grad(n_output_plane, initB, name='bconv_{}'.format(n_output_plane))
 
     def symb_forward(self, symb_input):
         conv_output = _dnn.dnn_conv(img=symb_input,