-
Notifications
You must be signed in to change notification settings - Fork 11
/
Copy patheval.py
178 lines (143 loc) · 6.91 KB
/
eval.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
166
167
168
169
170
171
172
173
174
175
176
177
178
import os
import argparse
import numpy as np
import torch
from torch.utils.data import DataLoader
from pytorch_lightning import Trainer
from pytorch_lightning.loggers import WandbLogger
import MinkowskiEngine as ME
import utils.models as models
from utils.datasets.initialization import get_dataset
from configs import get_config
from utils.collation import CollateFN
from utils.pipelines import PLTTester
parser = argparse.ArgumentParser()
parser.add_argument("--config_file",
default="configs/source/synlidar_semantickitti.yaml",
type=str,
help="Path to config file")
parser.add_argument("--eval_source",
action='store_true',
default=False)
parser.add_argument("--eval_target",
action='store_true',
default=False)
parser.add_argument("--resume_path",
type=str,
default=None)
parser.add_argument("--is_student",
default=False,
action='store_true')
parser.add_argument("--save_predictions",
default=False,
action='store_true')
def load_model(checkpoint_path, model):
# reloads model
def clean_state_dict(state):
# clean state dict from names of PL
for k in list(ckpt.keys()):
if "model" in k:
ckpt[k.replace("model.", "")] = ckpt[k]
del ckpt[k]
return state
ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu'))["state_dict"]
ckpt = clean_state_dict(ckpt)
model.load_state_dict(ckpt, strict=True)
return model
def load_student_model(checkpoint_path, model):
# reloads model
def clean_state_dict(state):
# clean state dict from names of PL
for k in list(ckpt.keys()):
if "model" in k:
ckpt[k.replace("student_model.", "")] = ckpt[k]
del ckpt[k]
return state
ckpt = torch.load(checkpoint_path, map_location=torch.device('cpu'))["state_dict"]
ckpt = clean_state_dict(ckpt)
model.load_state_dict(ckpt, strict=True)
return model
def test(config, resume_checkpoint):
def get_dataloader(dataset, shuffle=False, pin_memory=True):
return DataLoader(dataset,
batch_size=config.pipeline.dataloader.batch_size*4,
collate_fn=CollateFN(),
shuffle=shuffle,
num_workers=config.pipeline.dataloader.num_workers,
pin_memory=pin_memory)
try:
mapping_path = config.dataset.mapping_path
except AttributeError('--> Setting default class mapping path!'):
mapping_path = None
_, validation_dataset, target_dataset = get_dataset(dataset_name=config.dataset.name,
dataset_path=config.dataset.dataset_path,
target_name=config.dataset.target,
voxel_size=config.dataset.voxel_size,
augment_data=config.dataset.augment_data,
version=config.dataset.version,
sub_num=config.dataset.num_pts,
num_classes=config.model.out_classes,
ignore_label=config.dataset.ignore_label,
mapping_path=mapping_path)
validation_dataloader = get_dataloader(validation_dataset, shuffle=False)
target_dataloader = get_dataloader(target_dataset, shuffle=False)
validation_dataloader = [validation_dataloader]
target_dataloader = [target_dataloader]
Model = getattr(models, config.model.name)
model = Model(config.model.in_feat_size, config.model.out_classes)
model = ME.MinkowskiSyncBatchNorm.convert_sync_batchnorm(model)
if not args.is_student:
model = load_model(resume_checkpoint, model)
print(f'--> LOADED MODEL FROM {resume_checkpoint}')
else:
model = load_student_model(resume_checkpoint, model)
print(f'--> LOADED STUDENT MODEL FROM {resume_checkpoint}')
dataset = validation_dataset if args.eval_source else target_dataset
main_dir, _ = os.path.split(resume_checkpoint)
save_dir = os.path.join(main_dir, 'evaluation')
save_preds_dir = os.path.join(main_dir, 'predictions')
plt_model = PLTTester(model,
criterion=config.pipeline.loss,
dataset=dataset,
num_classes=config.model.out_classes,
checkpoint_path=resume_checkpoint,
save_predictions=args.save_predictions,
save_folder=save_preds_dir)
wandb_logger = WandbLogger(project=config.pipeline.wandb.project_name,
entity=config.pipeline.wandb.entity_name,
name=save_dir,
offline=True)
os.makedirs(save_dir, exist_ok=True)
loggers = [wandb_logger]
tester = Trainer(max_epochs=config.pipeline.epochs,
gpus=[0],
logger=loggers,
default_root_dir=save_dir,
weights_save_path=save_dir,
val_check_interval=1.0,
num_sanity_val_steps=0)
if args.eval_source:
tester.test(plt_model, dataloaders=validation_dataloader, verbose=False)
elif args.eval_target:
tester.test(plt_model, dataloaders=target_dataloader, verbose=False)
else:
print('Not evaluating!')
def multiple_test(config, path):
list_checkpoint = os.listdir(os.path.join(path, 'checkpoints'))
list_checkpoint = [c for c in list_checkpoint if c.endswith('.ckpt')]
for checkpoint in list_checkpoint:
if not os.path.isfile(os.path.join(path, 'checkpoints', 'evaluation', 'results', checkpoint[:-5]+'_test.csv')):
print(f'############### EVALUATING {checkpoint} ####################')
test(config, os.path.join(path, 'checkpoints', checkpoint))
else:
print(f'############### EVALUATION {checkpoint} SKIPPED - ALREADY PRESENT ####################')
if __name__ == '__main__':
args = parser.parse_args()
config = get_config(args.config_file)
# fix random seed
os.environ['PYTHONHASHSEED'] = str(config.pipeline.seed)
np.random.seed(config.pipeline.seed)
torch.manual_seed(config.pipeline.seed)
torch.cuda.manual_seed(config.pipeline.seed)
torch.backends.cudnn.benchmark = True
multiple_test(config, args.resume_path)