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
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
|
import path from 'path'
import fs from 'fs'
const name = 'pix2pix'
const cwd = process.env.PIX2PIX_CWD || path.join(process.env.HOME, 'code/' + name + '/')
/*
what are all the tasks that pix2pix has to do?
- fetch url
- fetch youtube
- ffmpeg movie into frames
- unzip zip file into sequence
- list sequences
*/
const fetch = {
type: 'perl',
script: 'get.pl',
params: (task) => {
console.log(task)
return [ task.module, task.opt.url ]
},
listen: (task, line, i) => {
// relay the new dataset name from youtube-dl or w/e
if ( line.match(/^created dataset: /) ) {
let filename = line.split(': ')[1].trim()
task.dataset = filename.split('.')[0]
task.opt.filename = filename
return { type: 'progress', action: 'resolve_dataset', task, }
}
return null
},
after: 'make_folds',
}
const make_folds = {
type: 'perl',
script: 'scripts/builders/pair-dataset-link.pl',
params: (task) => {
return [
'--sequence_dir', path.join(cwd, 'sequences', task.module, task.dataset),
'--ab_dir', path.join(cwd, 'a_b', task.module, task.dataset),
'--dt', task.opt.dt || 1,
]
},
after: 'combine_folds',
}
const combine_folds = {
type: 'pytorch',
script: 'datasets/combine_A_and_B.py',
params: (task) => {
return [
'--fold_A', path.join(cwd, 'a_b', task.module, task.dataset, 'A'),
'--fold_B', path.join(cwd, 'a_b', task.module, task.dataset, 'B'),
'--fold_AB', path.join(cwd, 'datasets', task.module, task.dataset),
]
}
}
const train = {
type: 'pytorch',
script: 'train.py',
params: (task) => {
let epoch = 0
const datasets_path = path.join(cwd, 'datasets', task.module, task.dataset)
const checkpoints_path = path.join(cwd, 'checkpoints', task.module, task.dataset)
if (fs.existsSync(checkpoints_path)) {
try {
const checkpoints = fs.readdirSync(checkpoints_path)
checkpoints.forEach(c => {
epoch = Math.max(parseInt(c.name) || 0, epoch)
})
console.log(task.module, task.dataset, epoch, task.epochs)
} catch (e) { }
}
let args = [
'--dataroot', datasets_path,
'--module-name', task.module,
'--name', task.dataset,
'--model', 'pix2pix',
'--loadSize', task.opt.load_size || 264,
'--fineSize', 256,
'--which_model_netG', 'unet_256',
'--which_direction', 'AtoB',
'--lambda_B', 100,
'--dataset_mode', 'aligned',
'--epoch_count', task.epoch + task.epochs + 1,
'--niter', task.epochs,
'--niter_decay', 0,
'--no_lsgan',
'--norm', 'batch',
'--pool_size', '0',
]
if (epoch) {
args = args.concat([
'--which_epoch', 'latest',
'--continue_train',
])
}
return args
},
}
const generate = {
type: 'pytorch',
script: 'test.py',
params: (task) => {
return [
'--dataroot', '/sequences/' + task.module + '/' + task.dataset,
'--module-name', task.module,
'--name', task.dataset,
'--start_img', '/sequences/' + task.module + '/' + task.dataset + '/frame_00001.png',
'--how_many', 1000,
'--model', 'test',
'--aspect_ratio', 1.777777,
'--which_model_netG', 'unet_256',
'--which_direction', 'AtoB',
'--dataset_mode', 'test',
'--loadSize', 256,
'--fineSize', 256,
'--norm', 'batch'
]
},
}
const live = {
type: 'pytorch',
script: 'live-mogrify.py',
params: (task) => {
console.log(task)
const opt = task.opt || {}
return [
'--dataroot', path.join(cwd, 'sequences', task.module, task.dataset),
'--start_img', path.join(cwd, 'sequences', task.module, task.dataset, 'frame_00001.png'),
'--checkpoint-name', task.checkpoint,
'--experiment', task.checkpoint,
'--name', task.checkpoint,
'--module-name', task.module,
'--sequence-name', task.dataset,
'--recursive', '--recursive-frac', 0.1,
'--sequence', '--sequence-frac', 0.3,
'--process-frac', 0.5,
'--nThreads', 0,
'--transition-min', 0.05,
'--how_many', 1000000, '--transition-period', 1000,
'--loadSize', 256, '--fineSize', 256,
'--just-copy', '--poll_delay', opt.poll_delay || 0.09,
'--model', 'test',
'--which_model_netG', 'unet_256',
'--which_direction', 'AtoB',
'--dataset_mode', 'recursive',
'--which_epoch', 'latest',
'--norm', 'batch',
]
},
}
export default {
name, cwd,
activities: {
fetch,
make_folds, combine_folds,
train,
generate,
live,
}
}
|
