forked from HarisIqbal88/PlotNeuralNet
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathblocks.py
75 lines (64 loc) · 2.88 KB
/
blocks.py
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
from .tikzeng import *
#define new block
def block_2ConvPool( name, botton, top, s_filer=256, n_filer=64, offset="(1,0,0)", size=(32,32,3.5), opacity=0.5 ):
return [
to_ConvConvRelu(
name="ccr_{}".format( name ),
s_filer=str(s_filer),
n_filer=(n_filer,n_filer),
offset=offset,
to="({}-east)".format( botton ),
width=(size[2],size[2]),
height=size[0],
depth=size[1],
),
to_Pool(
name="{}".format( top ),
offset="(0,0,0)",
to="(ccr_{}-east)".format( name ),
width=1,
height=size[0] - int(size[0]/4),
depth=size[1] - int(size[0]/4),
opacity=opacity, ),
to_connection(
"{}".format( botton ),
"ccr_{}".format( name )
)
]
def block_Unconv( name, botton, top, s_filer=256, n_filer=64, offset="(1,0,0)", size=(32,32,3.5), opacity=0.5 ):
return [
to_UnPool( name='unpool_{}'.format(name), offset=offset, to="({}-east)".format(botton), width=1, height=size[0], depth=size[1], opacity=opacity ),
to_ConvRes( name='ccr_res_{}'.format(name), offset="(0,0,0)", to="(unpool_{}-east)".format(name), s_filer=str(s_filer), n_filer=str(n_filer), width=size[2], height=size[0], depth=size[1], opacity=opacity ),
to_Conv( name='ccr_{}'.format(name), offset="(0,0,0)", to="(ccr_res_{}-east)".format(name), s_filer=str(s_filer), n_filer=str(n_filer), width=size[2], height=size[0], depth=size[1] ),
to_ConvRes( name='ccr_res_c_{}'.format(name), offset="(0,0,0)", to="(ccr_{}-east)".format(name), s_filer=str(s_filer), n_filer=str(n_filer), width=size[2], height=size[0], depth=size[1], opacity=opacity ),
to_Conv( name='{}'.format(top), offset="(0,0,0)", to="(ccr_res_c_{}-east)".format(name), s_filer=str(s_filer), n_filer=str(n_filer), width=size[2], height=size[0], depth=size[1] ),
to_connection(
"{}".format( botton ),
"unpool_{}".format( name )
)
]
def block_Res( num, name, botton, top, s_filer=256, n_filer=64, offset="(0,0,0)", size=(32,32,3.5), opacity=0.5 ):
lys = []
layers = [ *[ '{}_{}'.format(name,i) for i in range(num-1) ], top]
for name in layers:
ly = [ to_Conv(
name='{}'.format(name),
offset=offset,
to="({}-east)".format( botton ),
s_filer=str(s_filer),
n_filer=str(n_filer),
width=size[2],
height=size[0],
depth=size[1]
),
to_connection(
"{}".format( botton ),
"{}".format( name )
)
]
botton = name
lys+=ly
lys += [
to_skip( of=layers[1], to=layers[-2], pos=1.25),
]
return lys