-
Notifications
You must be signed in to change notification settings - Fork 27
/
train.py
165 lines (144 loc) · 8.01 KB
/
train.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
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
164
165
"""
Code to train IPNet.
Author: Bharat
Cite: Combining Implicit Function Learning and Parametric Models for 3D Human Reconstruction, ECCV 2020.
"""
import models.local_model_body_full as model
from data_loader.data_loader import DataLoaderFullBodyParts, DataLoaderFullBodyPartsSV, DataLoader
from models.trainer import TrainerIPNet, TrainerIPNetMano, Trainer
import argparse
import torch
import os
# os.environ["CUDA_DEVICE_ORDER"]="PCI_BUS_ID"
# os.environ["CUDA_VISIBLE_DEVICES"]="1"
parser = argparse.ArgumentParser(
description='Run Model'
)
# number of points in input in case of pointcloud input
parser.add_argument('-pc_samples', default=3000, type=int)
# number of points to predict as output
parser.add_argument('-num_sample_points', default=25000, type=int)
# distribution of samples used constructed via different standard devations
parser.add_argument('-dist', '--sample_distribution', default=[1], nargs='+', type=float)
# the standard deviations from the surface used to compute inside/outside samples
parser.add_argument('-std_dev', '--sample_sigmas', default=[0.005], nargs='+', type=float)
# defines how much input data is unsed as a batch.
parser.add_argument('-batch_size', default=30, type=int)
# the resolution of the input
parser.add_argument('-res', default=32, type=int)
# keep this fixed
parser.add_argument('-h_dim', '--decoder_hidden_dim', default=256, type=int)
# which model to use, e.g. "-m IPNet"
parser.add_argument('-m', '--model', default='IPNetSingleSurface', type=str)
# keep this fixed
parser.add_argument('-o', '--optimizer', default='Adam', type=str)
# data suffix
parser.add_argument('-suffix', '--suffix', default='', type=str)
# ext for data suffix
parser.add_argument('-ext', '--ext', default='', type=str)
# experiment id for folder suffix
parser.add_argument('-exp_id', '--exp_id', default='', type=str)
# Select singleView mode
parser.add_argument('-SV', dest='SV', action='store_true', default=False)
# Epochs
parser.add_argument('-epochs', default=150, type=int)
args = parser.parse_args()
if args.model == 'IPNet':
net = model.IPNet(hidden_dim=args.decoder_hidden_dim, num_parts=14)
elif args.model == 'IPNetMano':
net = model.IPNetMano(hidden_dim=args.decoder_hidden_dim, num_parts=7)
elif args.model == 'IPNetSingleSurface':
net = model.IPNetSingleSurface(hidden_dim=args.decoder_hidden_dim, num_parts=14)
else:
print('Wow watch where u goin\' with that model')
exit()
if args.model == 'IPNetMano':
args.split_file = '/BS/bharat-3/work/IPNet/DO_NOT_RELEASE/test_data/mano_data_split_01.pkl'
elif args.model == 'IPNetSingleSurface':
args.split_file = '/BS/bharat-3/work/IPNet/DO_NOT_RELEASE/test_data/data_split_single_surface.pkl'
else:
args.split_file = '/BS/bharat-3/work/IPNet/DO_NOT_RELEASE/test_data/data_split_double_surface.pkl'
if args.SV:
train_dataset = DataLoaderFullBodyPartsSV('train', pointcloud_samples=args.pc_samples, res=args.res,
sample_distribution=args.sample_distribution,
sample_sigmas=args.sample_sigmas,
num_sample_points=args.num_sample_points,
batch_size=args.batch_size, num_workers=30,
suffix=args.suffix, ext=args.ext,
split_file=args.split_file)
val_dataset = DataLoaderFullBodyPartsSV('val', pointcloud_samples=args.pc_samples,
res=args.res, sample_distribution=args.sample_distribution,
sample_sigmas=args.sample_sigmas, num_sample_points=args.num_sample_points,
batch_size=args.batch_size, num_workers=30,
suffix=args.suffix, ext=args.ext,
split_file=args.split_file)
elif args.model == 'IPNet':
train_dataset = DataLoaderFullBodyParts('train', pointcloud_samples=args.pc_samples, res=args.res,
sample_distribution=args.sample_distribution,
sample_sigmas=args.sample_sigmas, num_sample_points=args.num_sample_points,
batch_size=args.batch_size, num_workers=30,
suffix=args.suffix, ext=args.ext,
split_file=args.split_file)
val_dataset = DataLoaderFullBodyParts('val', pointcloud_samples=args.pc_samples,
res=args.res, sample_distribution=args.sample_distribution,
sample_sigmas=args.sample_sigmas, num_sample_points=args.num_sample_points,
batch_size=args.batch_size, num_workers=30,
suffix=args.suffix, ext=args.ext, split_file=args.split_file)
elif args.model == 'IPNetSingleSurface':
train_dataset = DataLoader('train', pointcloud_samples=args.pc_samples, res=args.res,
sample_distribution=args.sample_distribution,
sample_sigmas=args.sample_sigmas, num_sample_points=args.num_sample_points,
batch_size=args.batch_size, num_workers=30,
suffix=args.suffix, ext=args.ext,
split_file=args.split_file)
val_dataset = DataLoader('val', pointcloud_samples=args.pc_samples,
res=args.res, sample_distribution=args.sample_distribution,
sample_sigmas=args.sample_sigmas, num_sample_points=args.num_sample_points,
batch_size=args.batch_size, num_workers=30,
suffix=args.suffix, ext=args.ext, split_file=args.split_file)
if args.SV:
exp_name = '{}{}_{}_exp_id{}'.format(
args.model + '_SV',
'_p{}'.format(args.pc_samples),
args.ext,
args.exp_id
)
else:
exp_name = '{}{}_{}_exp_id{}'.format(
args.model,
'_p{}'.format(args.pc_samples),
args.ext,
args.exp_id
)
# Load pre-trained model. This model was trained for single layered predictions.
# We use this pre-training because inside surface is not available for all the scans.
# Skip if this model is not available.
if args.model != 'IPNetMano' and args.model != 'IPNetSingleSurface':
pre_path = 'IPNetSingleSurface_p5000_01s_exp_id02'
pre_trained = model.IPNetSingleSurface(hidden_dim=args.decoder_hidden_dim, num_parts=14)
pre_trainer = TrainerIPNet(pre_trained, torch.device("cuda"), None, None, pre_path,
optimizer=args.optimizer)
pre_trainer.load_checkpoint()
print('Loaded pretrained model from: ', pre_path)
# Copy weights for initial layers
import copy
for i, (src, tgt) in enumerate(zip(pre_trained.children(), net.children())):
if i > 9:
break
tgt.weight.data = copy.deepcopy(src.weight.data)
tgt.bias.data = copy.deepcopy(src.bias.data)
if args.model == 'IPNetMano':
exp_name += '_mano'
trainer = TrainerIPNetMano(net, torch.device("cuda"), train_dataset, val_dataset, exp_name,
optimizer=args.optimizer)
elif args.model == 'IPNet':
trainer = TrainerIPNet(net, torch.device("cuda"), train_dataset, val_dataset, exp_name,
optimizer=args.optimizer)
else: # single surface model no parts
trainer = Trainer(net, torch.device("cuda"), train_dataset, val_dataset, exp_name,
optimizer=args.optimizer)
trainer.train_model(args.epochs)
"""
python train.py -dist 0.5 0.5 -std_dev 0.15 0.015 -batch_size 4 -res 128 -m IPNetSingleSurface -ext 01s -suffix 01 -pc_samples 5000 -num_sample_points 20000 -exp_id 01
python train.py -dist 0.5 0.5 -std_dev 0.15 0.015 -batch_size 4 -res 128 -m IPNet -ext 01 -suffix 01 -pc_samples 5000 -num_sample_points 20000 -exp_id 01
"""