File size: 12,614 Bytes
98a3af2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
"""
DEIM: DETR with Improved Matching for Fast Convergence
Copyright (c) 2024 The DEIM Authors. All Rights Reserved.
---------------------------------------------------------------------------------
Modified from DETR (https://github.com/facebookresearch/detr/blob/main/engine.py)
Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved.
"""


import sys
import math
import gc
import os
from typing import Iterable

import torch
import torch.amp
from torch.utils.tensorboard import SummaryWriter
from torch.cuda.amp.grad_scaler import GradScaler
import cv2
import numpy as np

from ..optim import ModelEMA, Warmup
from ..data import CocoEvaluator
from ..misc import MetricLogger, SmoothedValue, dist_utils

try:
    import wandb
    _WANDB_AVAILABLE = True
except ImportError:
    _WANDB_AVAILABLE = False
    wandb = None


def visualize_augmented_batch(samples, targets, epoch, step):
    """
    Visualize augmented images with bounding boxes using OpenCV imshow.
    
    Args:
        samples (torch.Tensor): Batch of images [B, C, H, W]
        targets (list): List of target dictionaries
        epoch (int): Current epoch
        step (int): Current step
    """
    for idx in range(len(samples)):
        img = samples[idx].detach().cpu().permute(1, 2, 0).numpy()
        img_bgr = cv2.cvtColor(img, cv2.COLOR_RGB2BGR)
        img_bgr = (img_bgr * 255).astype(np.uint8)

        print(targets[idx]["image_id"])
        boxes = targets[idx]['boxes'].detach().cpu()
        labels = targets[idx].get('labels', None)

        h, w = img_bgr.shape[:2]

        print(len(boxes))
        boxes_np = boxes.numpy()
        for i, box in enumerate(boxes_np):
            cx, cy, bw, bh = box

            x1 = int((cx - bw / 2) * w)
            y1 = int((cy - bh / 2) * h)
            x2 = int((cx + bw / 2) * w)
            y2 = int((cy + bh / 2) * h)

            cv2.rectangle(img_bgr, (x1, y1), (x2, y2), (0, 0, 255), 3)
            print(f"box: {(x1, y1), (x2, y2)}")

            label = f"cls_{int(labels[i])}"
            label_size = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.8, 2)[0]
            label_y = max(label_size[1] + 5, y1 - 5)
            cv2.rectangle(img_bgr, (x1, label_y - label_size[1] - 5),
                          (x1 + label_size[0] + 10, label_y + 5), (0, 0, 255), -1)
            cv2.putText(img_bgr, label, (x1 + 5, label_y),
                        cv2.FONT_HERSHEY_SIMPLEX, 0.8, (255, 255, 255), 2)

        cv2.namedWindow('img', cv2.WINDOW_KEEPRATIO)
        cv2.imshow("img", img_bgr)
        cv2.waitKey(0)

def train_one_epoch(self_lr_scheduler, lr_scheduler, model: torch.nn.Module, criterion: torch.nn.Module,
                    data_loader: Iterable, optimizer: torch.optim.Optimizer,
                    device: torch.device, epoch: int, max_norm: float = 0, **kwargs):
    model.train()
    criterion.train()
    metric_logger = MetricLogger(delimiter="  ")
    metric_logger.add_meter('lr', SmoothedValue(window_size=1, fmt='{value:.6f}'))
    header = 'Epoch: [{}]'.format(epoch)

    print_freq = kwargs.get('print_freq', 10)
    writer :SummaryWriter = kwargs.get('writer', None)

    ema :ModelEMA = kwargs.get('ema', None)
    scaler :GradScaler = kwargs.get('scaler', None)
    lr_warmup_scheduler :Warmup = kwargs.get('lr_warmup_scheduler', None)
    
    # wandb parameters
    use_wandb = kwargs.get('use_wandb', False)
    wandb_run = kwargs.get('wandb_run', None)
    wandb_log_freq = kwargs.get('wandb_log_freq', 10)
    
    # gradient accumulation parameters
    gradient_accumulation_steps = kwargs.get('gradient_accumulation_steps', 1)

    cur_iters = epoch * len(data_loader)
    
    # Zero gradients at the beginning of epoch
    optimizer.zero_grad()

    for i, (samples, targets) in enumerate(metric_logger.log_every(data_loader, print_freq, header)):
        samples = samples.to(device)
        targets = [{k: v.to(device) for k, v in t.items()} for t in targets]
        global_step = epoch * len(data_loader) + i
        
        # Create metas as a simple dict without accumulating references
        metas = {'epoch': epoch, 'step': i, 'global_step': global_step, 'epoch_step': len(data_loader)}
        
        # Periodic garbage collection to prevent memory accumulation
        if i > 0 and i % 100 == 0:
            gc.collect()
            if torch.cuda.is_available():
                torch.cuda.empty_cache()

        # --- VISUALIZE AUGMENTED IMAGES & TARGETS (before model inference) ---
        enable_vis = kwargs.get('enable_vis', False)

        if enable_vis and dist_utils.is_main_process():
            visualize_augmented_batch(samples, targets, epoch, i)
        # ------------------------------------------------------------------------

        if scaler is not None:
            with torch.autocast(device_type=str(device), cache_enabled=True):
                outputs = model(samples, targets=targets)

            if torch.isnan(outputs['pred_boxes']).any() or torch.isinf(outputs['pred_boxes']).any():
                print(outputs['pred_boxes'])
                state = model.state_dict()
                new_state = {}
                for key, value in model.state_dict().items():
                    # Replace 'module' with 'model' in each key
                    new_key = key.replace('module.', '')
                    # Add the updated key-value pair to the state dictionary
                    state[new_key] = value
                new_state['model'] = state
                dist_utils.save_on_master(new_state, "./NaN.pth")

            with torch.autocast(device_type=str(device), enabled=False):
                loss_dict = criterion(outputs, targets, **metas)

            # Store keys for later reconstruction
            loss_keys = list(loss_dict.keys())
            loss_values = list(loss_dict.values())
            loss = sum(loss_values)
            # Scale loss for gradient accumulation
            loss = loss / gradient_accumulation_steps
            scaler.scale(loss).backward()

            # Only step and zero gradients every gradient_accumulation_steps
            if (i + 1) % gradient_accumulation_steps == 0:
                if max_norm > 0:
                    scaler.unscale_(optimizer)
                    torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm)

                scaler.step(optimizer)
                scaler.update()
                optimizer.zero_grad()

        else:
            # --- VISUALIZE AUGMENTED IMAGES & TARGETS (before model inference) ---
            enable_vis = kwargs.get('enable_vis', False)

            if enable_vis and dist_utils.is_main_process():
                visualize_augmented_batch(samples, targets, epoch, i)
            # ------------------------------------------------------------------------
            
            outputs = model(samples, targets=targets)
            loss_dict = criterion(outputs, targets, **metas)

            # Store keys for later reconstruction
            loss_keys = list(loss_dict.keys())
            loss_values = list(loss_dict.values())
            loss : torch.Tensor = sum(loss_values)
            # Scale loss for gradient accumulation
            loss = loss / gradient_accumulation_steps
            loss.backward()

            # Only step and zero gradients every gradient_accumulation_steps
            if (i + 1) % gradient_accumulation_steps == 0:
                if max_norm > 0:
                    torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm)

                optimizer.step()
                optimizer.zero_grad()

        # ema
        if ema is not None:
            ema.update(model)

        if self_lr_scheduler:
            optimizer = lr_scheduler.step(cur_iters + i, optimizer)
        else:
            if lr_warmup_scheduler is not None:
                lr_warmup_scheduler.step()

        # Recreate loss_dict for logging
        loss_dict = dict(zip(loss_keys, loss_values))
        loss_dict_reduced = dist_utils.reduce_dict(loss_dict)
        loss_value = sum(loss_dict_reduced.values())
        
        # Clean up references
        del loss_dict, outputs

        if not math.isfinite(loss_value):
            print("Loss is {}, stopping training".format(loss_value))
            print(loss_dict_reduced)
            sys.exit(1)

        metric_logger.update(loss=loss_value, **loss_dict_reduced)
        metric_logger.update(lr=optimizer.param_groups[0]["lr"])

        if writer and dist_utils.is_main_process() and global_step % 10 == 0:
            writer.add_scalar('Loss/total', loss_value.item(), global_step)
            for j, pg in enumerate(optimizer.param_groups):
                writer.add_scalar(f'Lr/pg_{j}', pg['lr'], global_step)
            for k, v in loss_dict_reduced.items():
                writer.add_scalar(f'Loss/{k}', v.item(), global_step)
        
        # wandb logging
        if (use_wandb and wandb_run is not None and _WANDB_AVAILABLE and 
            dist_utils.is_main_process() and global_step % wandb_log_freq == 0):
            log_dict = {
                'train/loss_total': loss_value.item(),
                'train/learning_rate': optimizer.param_groups[0]['lr'],
                'train/epoch': epoch,
                'train/step': global_step
            }
            # Add individual loss components
            for k, v in loss_dict_reduced.items():
                log_dict[f'train/loss_{k}'] = v.item()
            
            wandb.log(log_dict, step=global_step)

    # Step optimizer if there are remaining accumulated gradients at the end of epoch
    if (i + 1) % gradient_accumulation_steps != 0:
        if scaler is not None:
            if max_norm > 0:
                scaler.unscale_(optimizer)
                torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm)
            scaler.step(optimizer)
            scaler.update()
        else:
            if max_norm > 0:
                torch.nn.utils.clip_grad_norm_(model.parameters(), max_norm)
            optimizer.step()
        optimizer.zero_grad()

    # gather the stats from all processes
    metric_logger.synchronize_between_processes()
    print("Averaged stats:", metric_logger)
    
    # Final cleanup at end of epoch
    gc.collect()
    if torch.cuda.is_available():
        torch.cuda.empty_cache()
    
    return {k: meter.global_avg for k, meter in metric_logger.meters.items()}


@torch.no_grad()
def evaluate(model: torch.nn.Module, criterion: torch.nn.Module, postprocessor, data_loader, coco_evaluator: CocoEvaluator, device):
    model.eval()
    criterion.eval()
    coco_evaluator.cleanup()

    metric_logger = MetricLogger(delimiter="  ")
    # metric_logger.add_meter('class_error', SmoothedValue(window_size=1, fmt='{value:.2f}'))
    header = 'Test:'

    # iou_types = tuple(k for k in ('segm', 'bbox') if k in postprocessor.keys())
    iou_types = coco_evaluator.iou_types
    # coco_evaluator = CocoEvaluator(base_ds, 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, 10, header):
        samples = samples.to(device)
        targets = [{k: v.to(device) for k, v in t.items()} for t in targets]

        outputs = model(samples)

        orig_target_sizes = torch.stack([t["orig_size"] for t in targets], dim=0)

        results = postprocessor(outputs, orig_target_sizes)

        # if 'segm' in postprocessor.keys():
        #     target_sizes = torch.stack([t["size"] for t in targets], dim=0)
        #     results = postprocessor['segm'](results, outputs, orig_target_sizes, target_sizes)

        res = {target['image_id'].item(): output for target, output in zip(targets, results)}
        if coco_evaluator is not None:
            coco_evaluator.update(res)

    # 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 = {}
    # stats = {k: meter.global_avg for k, meter in metric_logger.meters.items()}
    if coco_evaluator is not None:
        if 'bbox' in iou_types:
            stats['coco_eval_bbox'] = coco_evaluator.coco_eval['bbox'].stats.tolist()
        if 'segm' in iou_types:
            stats['coco_eval_masks'] = coco_evaluator.coco_eval['segm'].stats.tolist()

    return stats, coco_evaluator