Merge pull request #12 from arnaud-roussel/init_tweaks

Better docstring. Image size and device found in forward.

Author: ndrplz

convlstm.py

@@ -1,18 +1,15 @@
 import torch.nn as nn
-from torch.autograd import Variable
 import torch
 
 
 class ConvLSTMCell(nn.Module):
 
-    def __init__(self, input_size, input_dim, hidden_dim, kernel_size, bias):
+    def __init__(self, input_dim, hidden_dim, kernel_size, bias):
         """
         Initialize ConvLSTM cell.
-        
+
         Parameters
         ----------
-        input_size: (int, int)
-            Height and width of input tensor as (height, width).
         input_dim: int
             Number of channels of input tensor.
         hidden_dim: int
@@ -25,60 +22,83 @@ def __init__(self, input_size, input_dim, hidden_dim, kernel_size, bias):
 
         super(ConvLSTMCell, self).__init__()
 
-        self.height, self.width = input_size
-        self.input_dim  = input_dim
+        self.input_dim = input_dim
         self.hidden_dim = hidden_dim
 
         self.kernel_size = kernel_size
-        self.padding     = kernel_size[0] // 2, kernel_size[1] // 2
-        self.bias        = bias
-        
+        self.padding = kernel_size[0] // 2, kernel_size[1] // 2
+        self.bias = bias
+
         self.conv = nn.Conv2d(in_channels=self.input_dim + self.hidden_dim,
                               out_channels=4 * self.hidden_dim,
                               kernel_size=self.kernel_size,
                               padding=self.padding,
                               bias=self.bias)
 
     def forward(self, input_tensor, cur_state):
-        
         h_cur, c_cur = cur_state
-        
+
         combined = torch.cat([input_tensor, h_cur], dim=1)  # concatenate along channel axis
-        
+
         combined_conv = self.conv(combined)
-        cc_i, cc_f, cc_o, cc_g = torch.split(combined_conv, self.hidden_dim, dim=1) 
+        cc_i, cc_f, cc_o, cc_g = torch.split(combined_conv, self.hidden_dim, dim=1)
         i = torch.sigmoid(cc_i)
         f = torch.sigmoid(cc_f)
         o = torch.sigmoid(cc_o)
         g = torch.tanh(cc_g)
 
         c_next = f * c_cur + i * g
         h_next = o * torch.tanh(c_next)
-        
+
         return h_next, c_next
 
-    def init_hidden(self, batch_size):
-        return (Variable(torch.zeros(batch_size, self.hidden_dim, self.height, self.width)).cuda(),
-                Variable(torch.zeros(batch_size, self.hidden_dim, self.height, self.width)).cuda())
+    def init_hidden(self, batch_size, image_size):
+        height, width = image_size
+        return (torch.zeros(batch_size, self.hidden_dim, height, width, device=self.conv.weight.device),
+                torch.zeros(batch_size, self.hidden_dim, height, width, device=self.conv.weight.device))
 
 
 class ConvLSTM(nn.Module):
 
-    def __init__(self, input_size, input_dim, hidden_dim, kernel_size, num_layers,
+    """
+
+    Parameters:
+        input_dim: Number of channels in input
+        hidden_dim: Number of hidden channels
+        kernel_size: Size of kernel in convolutions
+        num_layers: Number of LSTM layers stacked on each other
+        batch_first: Whether or not dimension 0 is the batch or not
+        bias: Bias or no bias in Convolution
+        return_all_layers: Return the list of computations for all layers
+        Note: Will do same padding.
+
+    Input:
+        A tensor of size B, T, C, H, W or T, B, C, H, W
+    Output:
+        A tuple of two lists of length num_layers (or length 1 if return_all_layers is False).
+            0 - layer_output_list is the list of lists of length T of each output
+            1 - last_state_list is the list of last states
+                    each element of the list is a tuple (h, c) for hidden state and memory
+    Example:
+        >> x = torch.rand((32, 10, 64, 128, 128))
+        >> convlstm = ConvLSTM(64, 16, 3, 1, True, True, False)
+        >> _, last_states = convlstm(x)
+        >> h = last_states[0][0]  # 0 for layer index, 0 for h index
+    """
+
+    def __init__(self, input_dim, hidden_dim, kernel_size, num_layers,
                  batch_first=False, bias=True, return_all_layers=False):
         super(ConvLSTM, self).__init__()
 
         self._check_kernel_size_consistency(kernel_size)
 
         # Make sure that both `kernel_size` and `hidden_dim` are lists having len == num_layers
         kernel_size = self._extend_for_multilayer(kernel_size, num_layers)
-        hidden_dim  = self._extend_for_multilayer(hidden_dim, num_layers)
+        hidden_dim = self._extend_for_multilayer(hidden_dim, num_layers)
         if not len(kernel_size) == len(hidden_dim) == num_layers:
             raise ValueError('Inconsistent list length.')
 
-        self.height, self.width = input_size
-
-        self.input_dim  = input_dim
+        self.input_dim = input_dim
         self.hidden_dim = hidden_dim
         self.kernel_size = kernel_size
         self.num_layers = num_layers
@@ -88,10 +108,9 @@ def __init__(self, input_size, input_dim, hidden_dim, kernel_size, num_layers,
 
         cell_list = []
         for i in range(0, self.num_layers):
-            cur_input_dim = self.input_dim if i == 0 else self.hidden_dim[i-1]
+            cur_input_dim = self.input_dim if i == 0 else self.hidden_dim[i - 1]
 
-            cell_list.append(ConvLSTMCell(input_size=(self.height, self.width),
-                                          input_dim=cur_input_dim,
+            cell_list.append(ConvLSTMCell(input_dim=cur_input_dim,
                                           hidden_dim=self.hidden_dim[i],
                                           kernel_size=self.kernel_size[i],
                                           bias=self.bias))
@@ -100,14 +119,14 @@ def __init__(self, input_size, input_dim, hidden_dim, kernel_size, num_layers,
 
     def forward(self, input_tensor, hidden_state=None):
         """
-        
+
         Parameters
         ----------
-        input_tensor: todo 
+        input_tensor: todo
             5-D Tensor either of shape (t, b, c, h, w) or (b, t, c, h, w)
         hidden_state: todo
             None. todo implement stateful
-            
+
         Returns
         -------
         last_state_list, layer_output
@@ -116,14 +135,18 @@ def forward(self, input_tensor, hidden_state=None):
             # (t, b, c, h, w) -> (b, t, c, h, w)
             input_tensor = input_tensor.permute(1, 0, 2, 3, 4)
 
+        b, _, _, h, w = input_tensor.size()
+
         # Implement stateful ConvLSTM
         if hidden_state is not None:
             raise NotImplementedError()
         else:
-            hidden_state = self._init_hidden(batch_size=input_tensor.size(0))
+            # Since the init is done in forward. Can send image size here
+            hidden_state = self._init_hidden(batch_size=b,
+                                             image_size=(h, w))
 
         layer_output_list = []
-        last_state_list   = []
+        last_state_list = []
 
         seq_len = input_tensor.size(1)
         cur_layer_input = input_tensor
@@ -133,7 +156,6 @@ def forward(self, input_tensor, hidden_state=None):
             h, c = hidden_state[layer_idx]
             output_inner = []
             for t in range(seq_len):
-
                 h, c = self.cell_list[layer_idx](input_tensor=cur_layer_input[:, t, :, :, :],
                                                  cur_state=[h, c])
                 output_inner.append(h)
@@ -146,20 +168,20 @@ def forward(self, input_tensor, hidden_state=None):
 
         if not self.return_all_layers:
             layer_output_list = layer_output_list[-1:]
-            last_state_list   = last_state_list[-1:]
+            last_state_list = last_state_list[-1:]
 
         return layer_output_list, last_state_list
 
-    def _init_hidden(self, batch_size):
+    def _init_hidden(self, batch_size, image_size):
         init_states = []
         for i in range(self.num_layers):
-            init_states.append(self.cell_list[i].init_hidden(batch_size))
+            init_states.append(self.cell_list[i].init_hidden(batch_size, image_size))
         return init_states
 
     @staticmethod
     def _check_kernel_size_consistency(kernel_size):
         if not (isinstance(kernel_size, tuple) or
-                    (isinstance(kernel_size, list) and all([isinstance(elem, tuple) for elem in kernel_size]))):
+                (isinstance(kernel_size, list) and all([isinstance(elem, tuple) for elem in kernel_size]))):
             raise ValueError('`kernel_size` must be tuple or list of tuples')
 
     @staticmethod