1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
|
local LinearTHNN, parent = torch.class('nn.LinearTHNN', 'nn.Module')
function LinearTHNN:__init(inputSize, outputSize, bias)
parent.__init(self)
local bias = ((bias == nil) and true) or bias
self.weight = torch.Tensor(outputSize, inputSize)
self.gradWeight = torch.Tensor(outputSize, inputSize)
if bias then
self.bias = torch.Tensor(outputSize)
self.gradBias = torch.Tensor(outputSize)
end
self.addBuffer = torch.Tensor(outputSize)
self:reset()
end
function LinearTHNN:noBias()
self.bias = nil
self.gradBias = nil
return self
end
function LinearTHNN:reset(stdv)
if stdv then
stdv = stdv * math.sqrt(3)
else
stdv = 1./math.sqrt(self.weight:size(2))
end
if nn.oldSeed then
for i=1,self.weight:size(1) do
self.weight:select(1, i):apply(function()
return torch.uniform(-stdv, stdv)
end)
end
if self.bias then
for i=1,self.bias:nElement() do
self.bias[i] = torch.uniform(-stdv, stdv)
end
end
else
self.weight:uniform(-stdv, stdv)
if self.bias then self.bias:uniform(-stdv, stdv) end
end
return self
end
function LinearTHNN:updateOutput(input)
input.THNN.Linear_updateOutput(
input:cdata(),
self.output:cdata(),
self.weight:cdata(),
self.bias and self.bias:cdata(),
self.addBuffer:cdata()
)
return self.output
end
function LinearTHNN:updateGradInput(input, gradOutput)
input.THNN.Linear_updateGradInput(
input:cdata(),
gradOutput:cdata(),
self.gradInput:cdata(),
self.weight:cdata()
)
return self.gradInput
end
function LinearTHNN:accGradParameters(input, gradOutput, scale)
input.THNN.Linear_accGradParameters(
input:cdata(),
gradOutput:cdata(),
self.gradInput:cdata(),
self.weight:cdata(),
self.bias and self.bias:cdata(),
self.gradWeight:cdata(),
self.bias and self.gradBias:cdata(),
self.addBuffer:cdata(),
scale or 1
)
return self.gradWeight
end
function LinearTHNN:sharedAccUpdateGradParameters(input, gradOutput, lr)
-- we do not need to accumulate parameters when sharing:
self:defaultAccUpdateGradParameters(input, gradOutput, lr)
end
function LinearTHNN:clearState()
if self.addBuffer then self.addBuffer:set() end
return parent.clearState(self)
end
function LinearTHNN:__tostring__()
return torch.type(self) ..
string.format('(%d -> %d)', self.weight:size(2), self.weight:size(1)) ..
(self.bias == nil and ' without bias' or '')
end
|
