-
-
Notifications
You must be signed in to change notification settings - Fork 95
/
softmax.nim
58 lines (49 loc) · 1.54 KB
/
softmax.nim
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
# Copyright 2017 the Arraymancer contributors
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
import ../../autograd,
../../tensor,
../../nn_primitives
type SoftmaxActivation* [TT] = ref object of Gate[TT]
cache: TT
proc softmax_backward_ag[TT](self: Gate[TT], payload: Payload[TT]): SmallDiffs[TT] =
let self = SoftmaxActivation[TT](Gate)
let gradient = payload.variable.grad
result = newDiffs[TT](1)
result[0] = gradient.softmax_backward(self.cache)
proc softmax_cache[TT](result: Variable[TT], a: Variable[TT]) =
# Gate
var gate: SoftmaxActivation[TT]
new gate
gate.cache = result.value
# Result setup
result.grad = zeros_like(result.value)
result.requires_grad = true
# Add to graph
register_node(
"Softmax",
gate,
softmax_backward_ag[TT],
result,
a
)
proc softmax*[TT](a: Variable[TT]): Variable[TT] =
## Input:
## - A variable
# Resulting var
new result
result.context = a.context
result.value = softmax a.value
# Caching for backprop
if a.is_grad_needed:
result.softmax_cache(a)