-
Notifications
You must be signed in to change notification settings - Fork 31
/
engine.py
193 lines (144 loc) · 7.49 KB
/
engine.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
179
180
181
182
183
184
185
186
187
188
189
190
191
192
# ------------------------------------------------------------------------
# Train and eval functions used in main.py
# ------------------------------------------------------------------------
# Modified from Deformable DETR (https://github.com/fundamentalvision/Deformable-DETR)
# Copyright (c) 2020 SenseTime. All Rights Reserved.
# ------------------------------------------------------------------------
import math
import os
import sys
from typing import Iterable
import cv2
import numpy as np
import json
import copy
import torch
import util.misc as utils
from util.misc import NestedTensor
from datasets.coco_eval import CocoEvaluator
from datasets.panoptic_eval import PanopticEvaluator
from datasets.data_prefetcher import data_prefetcher
from PIL import Image, ImageDraw
from scipy.optimize import linear_sum_assignment
cv2.setNumThreads(0)
cv2.ocl.setUseOpenCL(False)
def train_one_epoch(model: torch.nn.Module, criterion: torch.nn.Module,
data_loader: Iterable, optimizer: torch.optim.Optimizer,
device: torch.device, epoch: int, max_norm: float = 0):
model.train()
criterion.train()
metric_logger = utils.MetricLogger(delimiter=" ")
metric_logger.add_meter('lr', utils.SmoothedValue(window_size=1, fmt='{value:.6f}'))
metric_logger.add_meter('class_error', utils.SmoothedValue(window_size=1, fmt='{value:.2f}'))
metric_logger.add_meter('grad_norm', utils.SmoothedValue(window_size=1, fmt='{value:.2f}'))
header = 'Epoch: [{}]'.format(epoch)
print_freq = 4000
prefetcher = data_prefetcher(data_loader, device, prefetch=True)
samples, targets = prefetcher.next()
for _ in metric_logger.log_every(range(len(data_loader)), print_freq, header):
outputs, loss_dict = model(samples, targets, criterion, train=True)
weight_dict = criterion.weight_dict
losses = sum(loss_dict[k] * weight_dict[k] for k in loss_dict.keys() if k in weight_dict)
# reduce losses over all GPUs for logging purposes
loss_dict_reduced = utils.reduce_dict(loss_dict)
loss_dict_reduced_unscaled = {f'{k}_unscaled': v
for k, v in loss_dict_reduced.items()}
loss_dict_reduced_scaled = {k: v * weight_dict[k]
for k, v in loss_dict_reduced.items() if k in weight_dict}
losses_reduced_scaled = sum(loss_dict_reduced_scaled.values())
loss_value = losses_reduced_scaled.item()
if not math.isfinite(loss_value):
print("Loss is {}, stopping training".format(loss_value))
print(loss_dict_reduced)
sys.exit(1)
optimizer.zero_grad()
losses.backward()
if max_norm > 0:
grad_total_norm = torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm)
else:
grad_total_norm = utils.get_total_grad_norm(model.parameters(), max_norm)
optimizer.step()
metric_logger.update(loss=loss_value, **loss_dict_reduced_scaled, **loss_dict_reduced_unscaled)
metric_logger.update(class_error=loss_dict_reduced['class_error'])
metric_logger.update(lr=optimizer.param_groups[0]["lr"])
metric_logger.update(grad_norm=grad_total_norm)
samples, targets = prefetcher.next()
torch.cuda.empty_cache()
# gather the stats from all processes
metric_logger.synchronize_between_processes()
print("Averaged stats:", metric_logger)
return {k: meter.global_avg for k, meter in metric_logger.meters.items()}
@torch.no_grad()
def evaluate(model, criterion, postprocessors, data_loader, base_ds, device, args):
num_frames = args.num_frames
eval_types = args.eval_types
model.eval()
criterion.eval()
metric_logger = utils.MetricLogger(delimiter=" ")
metric_logger.add_meter('class_error', utils.SmoothedValue(window_size=1, fmt='{value:.2f}'))
header = 'Test:'
coco_iou_types = [k for k in ['bbox', 'segm'] if k in postprocessors.keys()]
coco_evaluator = None
if 'coco' in eval_types:
coco_evaluator = CocoEvaluator(base_ds['coco'], coco_iou_types)
# coco_evaluator.coco_eval[iou_types[0]].params.iouThrs = [0, 0.1, 0.5, 0.75]
for samples, targets in metric_logger.log_every(data_loader, 1000, header):
samples = samples.to(device)
all_outputs, loss_dict = model(samples, targets, criterion, train=False)
#### reduce losses over all GPUs for logging purposes ####
targets = [{k: v.to(device) for k, v in t.items()} for t in targets]
weight_dict = criterion.weight_dict
loss_dict_reduced = utils.reduce_dict(loss_dict)
loss_dict_reduced_scaled = {k: v * weight_dict[k]
for k, v in loss_dict_reduced.items() if k in weight_dict}
loss_dict_reduced_unscaled = {f'{k}_unscaled': v
for k, v in loss_dict_reduced.items()}
metric_logger.update(loss=sum(loss_dict_reduced_scaled.values()),
**loss_dict_reduced_scaled,
**loss_dict_reduced_unscaled)
metric_logger.update(class_error=loss_dict_reduced['class_error'])
#### reduce losses over all GPUs for logging purposes ####
##### single clip input ######
if all_outputs['pred_boxes'].dim() == 3:
all_outputs['pred_boxes'] = all_outputs['pred_boxes'].unsqueeze(2)
orig_target_sizes = torch.stack([t["orig_size"] for t in targets], dim=0)
results = [{} for i in range(len(targets))]
if 'bbox' in postprocessors.keys():
results = postprocessors['bbox'](all_outputs, orig_target_sizes, num_frames=num_frames)
# scores: [num_ins]
# labels: [num_ins]
# boxes: [num_ins, num_frames, 4]
if 'segm' in postprocessors.keys():
target_sizes = torch.stack([t["size"] for t in targets], dim=0)
results = postprocessors['segm'](results, all_outputs, orig_target_sizes, target_sizes)
res_img = {}
# evaluate results
if 'coco' in eval_types:
for target, output in zip(targets, results):
for fid in range(num_frames):
res_img[target['image_id'][fid].item()] = {}
for k, v in output.items():
if k == 'masks':
res_img[target['image_id'][fid].item()][k] = v[:,fid].unsqueeze(1)
elif k == 'boxes':
res_img[target['image_id'][fid].item()][k] = v[:,fid]
else:
res_img[target['image_id'][fid].item()][k] = v
if coco_evaluator is not None:
coco_evaluator.update(res_img)
# gather the stats from all processes
metric_logger.synchronize_between_processes()
print("Averaged stats:", metric_logger)
if coco_evaluator is not None:
coco_evaluator.synchronize_between_processes()
# accumulate predictions from all images
if coco_evaluator is not None:
coco_evaluator.accumulate()
coco_evaluator.summarize()
stats = {k: meter.global_avg for k, meter in metric_logger.meters.items()}
if coco_evaluator is not None:
if 'bbox' in postprocessors.keys():
stats['coco_eval_bbox'] = coco_evaluator.coco_eval['bbox'].stats.tolist()
if 'segm' in postprocessors.keys():
stats['coco_eval_masks'] = coco_evaluator.coco_eval['segm'].stats.tolist()
return stats, coco_evaluator