yolov8_comicbook / train_log.txt
jongkook90's picture
Fixed bug
13ed643
Logging results to runs/detect/train5
Starting training for 3 epochs...
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
1/3 8.35G 0.9681 0.8524 1.109 449 640: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 533/533 [03:36<00:00, 2.47it/s]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 267/267 [01:08<00:00, 3.89it/s]
all 8519 401038 0.87 0.815 0.885 0.635
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
2/3 9.16G 0.8795 0.6299 1.058 600 640: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 533/533 [01:36<00:00, 5.54it/s]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 267/267 [01:08<00:00, 3.90it/s]
all 8519 401038 0.891 0.845 0.907 0.663
Epoch GPU_mem box_loss cls_loss dfl_loss Instances Size
3/3 10.2G 0.8375 0.5747 1.037 590 640: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 533/533 [01:34<00:00, 5.63it/s]
Class Images Instances Box(P R mAP50 mAP50-95): 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 267/267 [02:00<00:00, 2.21it/s]Bn
all 8519 401038 0.903 0.864 0.923 0.689
3 epochs completed in 0.191 hours.
Optimizer stripped from runs/detect/train5/weights/last.pt, 52.0MB
Optimizer stripped from runs/detect/train5/weights/best.pt, 52.0MB
Validating runs/detect/train5/weights/best.pt...
Ultralytics YOLOv8.0.120 πŸš€ Python-3.9.7 torch-2.0.0+cu118 CUDA:0 (NVIDIA GeForce RTX 3090, 24265MiB)
Model summary (fused): 218 layers, 25842076 parameters, 0 gradients
Class Images Instances Box(P R mAP50 mAP50-95): 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 267/267 [01:56<00:00, 2.29it/s]
all 8519 401038 0.903 0.864 0.923 0.689
frame 8519 82124 0.944 0.972 0.989 0.942
face 8519 87174 0.903 0.77 0.873 0.467
body 8519 118126 0.852 0.811 0.885 0.642
text 8519 113614 0.915 0.9 0.944 0.704
Speed: 0.1ms preprocess, 2.1ms inference, 0.0ms loss, 0.8ms postprocess per image
Results saved to runs/detect/train5
Ultralytics YOLOv8.0.120 πŸš€ Python-3.9.7 torch-2.0.0+cu118 CUDA:0 (NVIDIA GeForce RTX 3090, 24265MiB)
Model summary (fused): 218 layers, 25842076 parameters, 0 gradients
val: Scanning /media/jongkook90/Morpho DB/dataset-manga109-ko/yolov8-frame/datasets/manga109/labels/train.cache... 8155 images, 364 backgrounds, 0 corrupt: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 8519/8519 [00:00<?, ?it/s]
Class Images Instances Box(P R mAP50 mAP50-95): 0%| | 2/533 [00:08<43:38, 4.93s/it]WARNING ⚠️ NMS time limit 1.300s exceeded
Class Images Instances Box(P R mAP50 mAP50-95): 1%| | 3/533 [00:14<47:58, 5.43s/it]WARNING ⚠️ NMS time limit 1.300s exceeded
Class Images Instances Box(P R mAP50 mAP50-95): 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 533/533 [02:06<00:00, 4.22it/s]
all 8519 401038 0.904 0.861 0.92 0.688
frame 8519 82124 0.944 0.969 0.986 0.941
face 8519 87174 0.903 0.769 0.87 0.467
body 8519 118126 0.852 0.809 0.882 0.641
text 8519 113614 0.916 0.897 0.941 0.704
Speed: 0.1ms preprocess, 4.1ms inference, 0.0ms loss, 1.5ms postprocess per image
Results saved to runs/detect/val3