add mosaic and warmup to hyperparameters (#931)

data/hyp.finetune.yaml

@@ -12,6 +12,9 @@ lr0: 0.0032
 lrf: 0.12
 momentum: 0.843
 weight_decay: 0.00036
+warmup_epochs: 2.0
+warmup_momentum: 0.5
+warmup_bias_lr: 0.05
 giou: 0.0296
 cls: 0.243
 cls_pw: 0.631
@@ -31,4 +34,5 @@ shear: 0.602
 perspective: 0.0
 flipud: 0.00856
 fliplr: 0.5
+mosaic: 1.0
 mixup: 0.243

data/hyp.scratch.yaml

@@ -7,6 +7,9 @@ lr0: 0.01  # initial learning rate (SGD=1E-2, Adam=1E-3)
 lrf: 0.2  # final OneCycleLR learning rate (lr0 * lrf)
 momentum: 0.937  # SGD momentum/Adam beta1
 weight_decay: 0.0005  # optimizer weight decay 5e-4
+warmup_epochs: 3.0  # warmup epochs (fractions ok)
+warmup_momentum: 0.8  # warmup initial momentum
+warmup_bias_lr: 0.1  # warmup initial bias lr
 giou: 0.05  # box loss gain
 cls: 0.5  # cls loss gain
 cls_pw: 1.0  # cls BCELoss positive_weight
@@ -26,4 +29,5 @@ shear: 0.0  # image shear (+/- deg)
 perspective: 0.0  # image perspective (+/- fraction), range 0-0.001
 flipud: 0.0  # image flip up-down (probability)
 fliplr: 0.5  # image flip left-right (probability)
+mosaic: 1.0  # image mosaic (probability)
 mixup: 0.0  # image mixup (probability)

train.py

@@ -202,7 +202,7 @@ def train(hyp, opt, device, tb_writer=None):
 
     # Start training
     t0 = time.time()
-    nw = max(3 * nb, 1e3)  # number of warmup iterations, max(3 epochs, 1k iterations)
+    nw = max(round(hyp['warmup_epochs'] * nb), 1e3)  # number of warmup iterations, max(3 epochs, 1k iterations)
     # nw = min(nw, (epochs - start_epoch) / 2 * nb)  # limit warmup to < 1/2 of training
     maps = np.zeros(nc)  # mAP per class
     results = (0, 0, 0, 0, 0, 0, 0)  # 'P', 'R', 'mAP', 'F1', 'val GIoU', 'val Objectness', 'val Classification'
@@ -250,9 +250,9 @@ def train(hyp, opt, device, tb_writer=None):
                 accumulate = max(1, np.interp(ni, xi, [1, nbs / total_batch_size]).round())
                 for j, x in enumerate(optimizer.param_groups):
                     # bias lr falls from 0.1 to lr0, all other lrs rise from 0.0 to lr0
-                    x['lr'] = np.interp(ni, xi, [0.1 if j == 2 else 0.0, x['initial_lr'] * lf(epoch)])
+                    x['lr'] = np.interp(ni, xi, [hyp['warmup_bias_lr'] if j == 2 else 0.0, x['initial_lr'] * lf(epoch)])
                     if 'momentum' in x:
-                        x['momentum'] = np.interp(ni, xi, [0.9, hyp['momentum']])
+                        x['momentum'] = np.interp(ni, xi, [hyp['warmup_momentum'], hyp['momentum']])
 
             # Multi-scale
             if opt.multi_scale:
@@ -460,8 +460,11 @@ if __name__ == '__main__':
         # Hyperparameter evolution metadata (mutation scale 0-1, lower_limit, upper_limit)
         meta = {'lr0': (1, 1e-5, 1e-1),  # initial learning rate (SGD=1E-2, Adam=1E-3)
                 'lrf': (1, 0.01, 1.0),  # final OneCycleLR learning rate (lr0 * lrf)
-                'momentum': (0.1, 0.6, 0.98),  # SGD momentum/Adam beta1
+                'momentum': (0.3, 0.6, 0.98),  # SGD momentum/Adam beta1
                 'weight_decay': (1, 0.0, 0.001),  # optimizer weight decay
+                'warmup_epochs': (1, 0.0, 5.0),  # warmup epochs (fractions ok)
+                'warmup_momentum': (1, 0.0, 0.95),  # warmup initial momentum
+                'warmup_bias_lr': (1, 0.0, 0.2),  # warmup initial bias lr
                 'giou': (1, 0.02, 0.2),  # GIoU loss gain
                 'cls': (1, 0.2, 4.0),  # cls loss gain
                 'cls_pw': (1, 0.5, 2.0),  # cls BCELoss positive_weight
@@ -469,7 +472,7 @@ if __name__ == '__main__':
                 'obj_pw': (1, 0.5, 2.0),  # obj BCELoss positive_weight
                 'iou_t': (0, 0.1, 0.7),  # IoU training threshold
                 'anchor_t': (1, 2.0, 8.0),  # anchor-multiple threshold
-                'anchors': (1, 2.0, 10.0),  # anchors per output grid (0 to ignore)
+                'anchors': (2, 2.0, 10.0),  # anchors per output grid (0 to ignore)
                 'fl_gamma': (0, 0.0, 2.0),  # focal loss gamma (efficientDet default gamma=1.5)
                 'hsv_h': (1, 0.0, 0.1),  # image HSV-Hue augmentation (fraction)
                 'hsv_s': (1, 0.0, 0.9),  # image HSV-Saturation augmentation (fraction)
@@ -481,6 +484,7 @@ if __name__ == '__main__':
                 'perspective': (0, 0.0, 0.001),  # image perspective (+/- fraction), range 0-0.001
                 'flipud': (1, 0.0, 1.0),  # image flip up-down (probability)
                 'fliplr': (0, 0.0, 1.0),  # image flip left-right (probability)
+                'mosaic': (1, 0.0, 1.0),  # image mixup (probability)
                 'mixup': (1, 0.0, 1.0)}  # image mixup (probability)
 
         assert opt.local_rank == -1, 'DDP mode not implemented for --evolve'
@@ -490,7 +494,7 @@ if __name__ == '__main__':
         if opt.bucket:
             os.system('gsutil cp gs://%s/evolve.txt .' % opt.bucket)  # download evolve.txt if exists
 
-        for _ in range(1):  # generations to evolve
+        for _ in range(300):  # generations to evolve
             if os.path.exists('evolve.txt'):  # if evolve.txt exists: select best hyps and mutate
                 # Select parent(s)
                 parent = 'single'  # parent selection method: 'single' or 'weighted'
@@ -505,7 +509,7 @@ if __name__ == '__main__':
                     x = (x * w.reshape(n, 1)).sum(0) / w.sum()  # weighted combination
 
                 # Mutate
-                mp, s = 0.9, 0.2  # mutation probability, sigma
+                mp, s = 0.8, 0.2  # mutation probability, sigma
                 npr = np.random
                 npr.seed(int(time.time()))
                 g = np.array([x[0] for x in meta.values()])  # gains 0-1

utils/datasets.py

@@ -516,7 +516,8 @@ class LoadImagesAndLabels(Dataset):  # for training/testing
             index = self.indices[index]
 
         hyp = self.hyp
-        if self.mosaic:
+        mosaic = self.mosaic and random.random() < hyp['mosaic']
+        if mosaic:
             # Load mosaic
             img, labels = load_mosaic(self, index)
             shapes = None
@@ -550,7 +551,7 @@ class LoadImagesAndLabels(Dataset):  # for training/testing
 
         if self.augment:
             # Augment imagespace
-            if not self.mosaic:
+            if not mosaic:
                 img, labels = random_perspective(img, labels,
                                                  degrees=hyp['degrees'],
                                                  translate=hyp['translate'],