Updated VOC hyperparameters (#6732)

data/hyps/hyp.finetune_objects365.yaml → data/hyps/hyp.Objects365.yaml

 # YOLOv5 🚀 by Ultralytics, GPL-3.0 license
+# Hyperparameters for Objects365 training
+# python train.py --weights yolov5m.pt --data Objects365.yaml --evolve
+# See Hyperparameter Evolution tutorial for details https://github.com/ultralytics/yolov5#tutorials
 
 lr0: 0.00258
 lrf: 0.17

data/hyps/hyp.VOC.yaml

+# YOLOv5 🚀 by Ultralytics, GPL-3.0 license
+# Hyperparameters for VOC training
+# python train.py --batch 128 --weights yolov5m6.pt --data VOC.yaml --epochs 50 --img 512 --hyp hyp.scratch-med.yaml --evolve
+# See Hyperparameter Evolution tutorial for details https://github.com/ultralytics/yolov5#tutorials
+
+# YOLOv5 Hyperparameter Evolution Results
+# Best generation: 319
+# Last generation: 434
+#    metrics/precision,       metrics/recall,      metrics/mAP_0.5, metrics/mAP_0.5:0.95,         val/box_loss,         val/obj_loss,         val/cls_loss
+#              0.86236,              0.86184,              0.91274,              0.72647,            0.0077056,            0.0042449,            0.0013846
+
+lr0: 0.0033
+lrf: 0.15184
+momentum: 0.74747
+weight_decay: 0.00025
+warmup_epochs: 3.4278
+warmup_momentum: 0.59032
+warmup_bias_lr: 0.18742
+box: 0.02
+cls: 0.21563
+cls_pw: 0.5
+obj: 0.50843
+obj_pw: 0.6729
+iou_t: 0.2
+anchor_t: 3.4172
+fl_gamma: 0.0
+hsv_h: 0.01032
+hsv_s: 0.5562
+hsv_v: 0.28255
+degrees: 0.0
+translate: 0.04575
+scale: 0.73711
+shear: 0.0
+perspective: 0.0
+flipud: 0.0
+fliplr: 0.5
+mosaic: 0.87158
+mixup: 0.04294
+copy_paste: 0.0
+anchors: 3.3556

data/hyps/hyp.finetune.yaml

-# YOLOv5 🚀 by Ultralytics, GPL-3.0 license
-# Hyperparameters for VOC finetuning
-# python train.py --batch 64 --weights yolov5m.pt --data VOC.yaml --img 512 --epochs 50
-# See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials
-
-# Hyperparameter Evolution Results
-# Generations: 306
-#                   P         R     mAP.5 mAP.5:.95       box       obj       cls
-# Metrics:        0.6     0.936     0.896     0.684    0.0115   0.00805   0.00146
-
-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
-box: 0.0296
-cls: 0.243
-cls_pw: 0.631
-obj: 0.301
-obj_pw: 0.911
-iou_t: 0.2
-anchor_t: 2.91
-# anchors: 3.63
-fl_gamma: 0.0
-hsv_h: 0.0138
-hsv_s: 0.664
-hsv_v: 0.464
-degrees: 0.373
-translate: 0.245
-scale: 0.898
-shear: 0.602
-perspective: 0.0
-flipud: 0.00856
-fliplr: 0.5
-mosaic: 1.0
-mixup: 0.243
-copy_paste: 0.0

data/hyps/hyp.scratch.yaml

-# YOLOv5 🚀 by Ultralytics, GPL-3.0 license
-# Hyperparameters for COCO training from scratch
-# python train.py --batch 40 --cfg yolov5m.yaml --weights '' --data coco.yaml --img 640 --epochs 300
-# See tutorials for hyperparameter evolution https://github.com/ultralytics/yolov5#tutorials
-
-lr0: 0.01  # initial learning rate (SGD=1E-2, Adam=1E-3)
-lrf: 0.1  # 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
-box: 0.05  # box loss gain
-cls: 0.5  # cls loss gain
-cls_pw: 1.0  # cls BCELoss positive_weight
-obj: 1.0  # obj loss gain (scale with pixels)
-obj_pw: 1.0  # obj BCELoss positive_weight
-iou_t: 0.20  # IoU training threshold
-anchor_t: 4.0  # anchor-multiple threshold
-# anchors: 3  # anchors per output layer (0 to ignore)
-fl_gamma: 0.0  # focal loss gamma (efficientDet default gamma=1.5)
-hsv_h: 0.015  # image HSV-Hue augmentation (fraction)
-hsv_s: 0.7  # image HSV-Saturation augmentation (fraction)
-hsv_v: 0.4  # image HSV-Value augmentation (fraction)
-degrees: 0.0  # image rotation (+/- deg)
-translate: 0.1  # image translation (+/- fraction)
-scale: 0.5  # image scale (+/- gain)
-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)
-copy_paste: 0.0  # segment copy-paste (probability)

train.py

@@ -456,7 +456,7 @@ def parse_opt(known=False):
     parser.add_argument('--weights', type=str, default=ROOT / 'yolov5s.pt', help='initial weights path')
     parser.add_argument('--cfg', type=str, default='', help='model.yaml path')
     parser.add_argument('--data', type=str, default=ROOT / 'data/coco128.yaml', help='dataset.yaml path')
-    parser.add_argument('--hyp', type=str, default=ROOT / 'data/hyps/hyp.scratch.yaml', help='hyperparameters path')
+    parser.add_argument('--hyp', type=str, default=ROOT / 'data/hyps/hyp.scratch-low.yaml', help='hyperparameters path')
     parser.add_argument('--epochs', type=int, default=300)
     parser.add_argument('--batch-size', type=int, default=16, help='total batch size for all GPUs, -1 for autobatch')
     parser.add_argument('--imgsz', '--img', '--img-size', type=int, default=640, help='train, val image size (pixels)')

utils/general.py

@@ -795,7 +795,7 @@ def print_mutation(results, hyp, save_dir, bucket, prefix=colorstr('evolve: ')):
     # Download (optional)
     if bucket:
         url = f'gs://{bucket}/evolve.csv'
-        if gsutil_getsize(url) > (os.path.getsize(evolve_csv) if os.path.exists(evolve_csv) else 0):
+        if gsutil_getsize(url) > (evolve_csv.stat().st_size if evolve_csv.exists() else 0):
             os.system(f'gsutil cp {url} {save_dir}')  # download evolve.csv if larger than local
 
     # Log to evolve.csv