Hello,
I am trying out the below tutorial by @shahizat for safety helmet detection. I got upto the point of installing yolo and converting PASCAL VOC format to the format required by YOLOv8.
Now I am stuck at the point of running yolo detect task as there is no mention of the contents of data.yaml.
Please help.
Detection task is up and running with batch size of 12 and epoch of 100. Looks like each epoch takes about 3.5 hours. So, will take about 15 days to complete.
Any thoughts about speeding this up? is it using GPU at all?
Hey @ngr,
Could you please provide a log for us to review?
It should be using the GPU. Also, keep in mind that not all epochs take the same amount of time.
Please find the logs below. It is up to epoch 13 in ~2 days.
Thanks for looking into this.
yolo task=detect mode=train model=yolov8n.pt data=dataset.yaml epochs=100 imgsz=640 batch=12
New ultralytics · PyPI available Update with ‘pip install -U ultralytics’
Ultralytics YOLOv8.2.79 Python-3.11.2 torch-2.4.0 CPU (Cortex-A76)
engine/trainer: task=detect, mode=train, model=yolov8n.pt, data=dataset.yaml, epochs=100, time=None, patience=100, batch=12, imgsz=640, save=True, save_period=-1, cache=False, device=None, workers=8, project=None, name=train10, exist_ok=False, pretrained=True, optimizer=auto, verbose=True, seed=0, deterministic=True, single_cls=False, rect=False, cos_lr=False, close_mosaic=10, resume=False, amp=True, fraction=1.0, profile=False, freeze=None, multi_scale=False, overlap_mask=True, mask_ratio=4, dropout=0.0, val=True, split=val, save_json=False, save_hybrid=False, conf=None, iou=0.7, max_det=300, half=False, dnn=False, plots=True, source=None, vid_stride=1, stream_buffer=False, visualize=False, augment=False, agnostic_nms=False, classes=None, retina_masks=False, embed=None, show=False, save_frames=False, save_txt=False, save_conf=False, save_crop=False, show_labels=True, show_conf=True, show_boxes=True, line_width=None, format=torchscript, keras=False, optimize=False, int8=False, dynamic=False, simplify=False, opset=None, workspace=4, nms=False, lr0=0.01, lrf=0.01, momentum=0.937, weight_decay=0.0005, warmup_epochs=3.0, warmup_momentum=0.8, warmup_bias_lr=0.1, box=7.5, cls=0.5, dfl=1.5, pose=12.0, kobj=1.0, label_smoothing=0.0, nbs=64, hsv_h=0.015, hsv_s=0.7, hsv_v=0.4, degrees=0.0, translate=0.1, scale=0.5, shear=0.0, perspective=0.0, flipud=0.0, fliplr=0.5, bgr=0.0, mosaic=1.0, mixup=0.0, copy_paste=0.0, auto_augment=randaugment, erasing=0.4, crop_fraction=1.0, cfg=None, tracker=botsort.yaml, save_dir=runs/detect/train10
Overriding model.yaml nc=80 with nc=3
from n params module arguments
0 -1 1 464 ultralytics.nn.modules.conv.Conv [3, 16, 3, 2]
1 -1 1 4672 ultralytics.nn.modules.conv.Conv [16, 32, 3, 2]
2 -1 1 7360 ultralytics.nn.modules.block.C2f [32, 32, 1, True]
3 -1 1 18560 ultralytics.nn.modules.conv.Conv [32, 64, 3, 2]
4 -1 2 49664 ultralytics.nn.modules.block.C2f [64, 64, 2, True]
5 -1 1 73984 ultralytics.nn.modules.conv.Conv [64, 128, 3, 2]
6 -1 2 197632 ultralytics.nn.modules.block.C2f [128, 128, 2, True]
7 -1 1 295424 ultralytics.nn.modules.conv.Conv [128, 256, 3, 2]
8 -1 1 460288 ultralytics.nn.modules.block.C2f [256, 256, 1, True]
9 -1 1 164608 ultralytics.nn.modules.block.SPPF [256, 256, 5]
10 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, ‘nearest’]
11 [-1, 6] 1 0 ultralytics.nn.modules.conv.Concat [1]
12 -1 1 148224 ultralytics.nn.modules.block.C2f [384, 128, 1]
13 -1 1 0 torch.nn.modules.upsampling.Upsample [None, 2, ‘nearest’]
14 [-1, 4] 1 0 ultralytics.nn.modules.conv.Concat [1]
15 -1 1 37248 ultralytics.nn.modules.block.C2f [192, 64, 1]
16 -1 1 36992 ultralytics.nn.modules.conv.Conv [64, 64, 3, 2]
17 [-1, 12] 1 0 ultralytics.nn.modules.conv.Concat [1]
18 -1 1 123648 ultralytics.nn.modules.block.C2f [192, 128, 1]
19 -1 1 147712 ultralytics.nn.modules.conv.Conv [128, 128, 3, 2]
20 [-1, 9] 1 0 ultralytics.nn.modules.conv.Concat [1]
21 -1 1 493056 ultralytics.nn.modules.block.C2f [384, 256, 1]
22 [15, 18, 21] 1 751897 ultralytics.nn.modules.head.Detect [3, [64, 128, 256]]
Model summary: 225 layers, 3,011,433 parameters, 3,011,417 gradients, 8.2 GFLOPs
Transferred 319/355 items from pretrained weights
Freezing layer ‘model.22.dfl.conv.weight’
/home/rpi/work/helmet/lib/python3.11/site-packages/ultralytics/engine/trainer.py:271: FutureWarning: torch.cuda.amp.GradScaler(args...) is deprecated. Please use torch.amp.GradScaler('cuda', args...) instead.
self.scaler = torch.cuda.amp.GradScaler(enabled=self.amp)
train: Scanning /home/rpi/work/helmet/datasets/dataset/labels.cache… 5000 images, 0 backgrounds, 0 corrupt: 100%|██████████| 5000/5000 [00:00<?, ?it/s]
val: Scanning /home/rpi/work/helmet/datasets/dataset/labels.cache… 5000 images, 0 backgrounds, 0 corrupt: 100%|██████████| 5000/5000 [00:00<?, ?it/s]
Plotting labels to runs/detect/train10/labels.jpg…
optimizer: ‘optimizer=auto’ found, ignoring ‘lr0=0.01’ and ‘momentum=0.937’ and determining best ‘optimizer’, ‘lr0’ and ‘momentum’ automatically…
optimizer: AdamW(lr=0.001429, momentum=0.9) with parameter groups 57 weight(decay=0.0), 64 weight(decay=0.00046875), 63 bias(decay=0.0)
Image sizes 640 train, 640 val
Using 0 dataloader workers
Logging results to runs/detect/train10
Starting training for 100 epochs…
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
1/100 0G 1.52 1.688 1.216 48 640: 100%|██████████| 417/417 [2:32:25<00:00, 21.93s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:08:35<00:00, 19.69s/it]
all 5000 25502 0.912 0.496 0.558 0.328
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
2/100 0G 1.448 1.164 1.186 70 640: 100%|██████████| 417/417 [2:45:06<00:00, 23.76s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:06:52<00:00, 19.20s/it]
all 5000 25502 0.897 0.517 0.575 0.334
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
3/100 0G 1.433 1.061 1.181 82 640: 100%|██████████| 417/417 [2:47:54<00:00, 24.16s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:10:39<00:00, 20.28s/it]
all 5000 25502 0.902 0.493 0.557 0.3
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
4/100 0G 1.418 1.011 1.178 59 640: 100%|██████████| 417/417 [2:36:40<00:00, 22.54s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:09:09<00:00, 19.86s/it]
all 5000 25502 0.919 0.528 0.589 0.359
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
5/100 0G 1.395 0.9661 1.174 88 640: 100%|██████████| 417/417 [2:40:22<00:00, 23.08s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:09:17<00:00, 19.89s/it]
all 5000 25502 0.93 0.551 0.606 0.373
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
6/100 0G 1.374 0.9265 1.163 68 640: 100%|██████████| 417/417 [2:39:58<00:00, 23.02s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:11:39<00:00, 20.57s/it]
all 5000 25502 0.934 0.55 0.605 0.373
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
7/100 0G 1.362 0.9035 1.156 91 640: 100%|██████████| 417/417 [2:54:14<00:00, 25.07s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:13:28<00:00, 21.09s/it]
all 5000 25502 0.928 0.562 0.61 0.386
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
8/100 0G 1.362 0.8804 1.147 67 640: 100%|██████████| 417/417 [2:55:44<00:00, 25.29s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:12:19<00:00, 20.76s/it]
all 5000 25502 0.933 0.564 0.612 0.385
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
9/100 0G 1.342 0.8569 1.148 40 640: 100%|██████████| 417/417 [3:00:41<00:00, 26.00s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:11:05<00:00, 20.41s/it]
all 5000 25502 0.941 0.561 0.618 0.387
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
10/100 0G 1.338 0.8457 1.139 33 640: 100%|██████████| 417/417 [2:46:12<00:00, 23.92s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:12:17<00:00, 20.75s/it]
all 5000 25502 0.939 0.573 0.621 0.399
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
11/100 0G 1.316 0.8324 1.133 42 640: 100%|██████████| 417/417 [2:52:32<00:00, 24.83s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:12:30<00:00, 20.82s/it]
all 5000 25502 0.942 0.573 0.624 0.393
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
12/100 0G 1.327 0.8214 1.14 49 640: 100%|██████████| 417/417 [3:05:05<00:00, 26.63s/it]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|██████████| 209/209 [1:11:21<00:00, 20.49s/it]
all 5000 25502 0.939 0.567 0.624 0.389
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
13/100 0G 1.325 0.8227 1.13 70 640: 61%|██████ | 253/417 [2:03:43<2:26:17, 53.52s/it]
Finally the model is ready. However it wont run on rpi/hailo because it was compiled using 3.28 
have to re-do the hef part.
Got the model working. however it is only detecting person. No helmet and head detection.
What could be wrong??