9.13 实验记录

工位电脑复现yolov8    并训练  FLIR数据集

• 问题1：数据集划分标签label文件内容应为绝对路径

(yolov8) wzy@wzy-Victus-by-HP-15L-Gaming-Desktop-TG02-1xxx:~/yolo/ultralytics-main$ yolo task=detect mode=train model=yolov8n.pt data=ultralytics/yolo/data/dataset/FLIR.yaml epochs=150 batch=16

Ultralytics YOLOv8.0.176 ��� Python-3.8.18 torch-1.11.0+cu113 CUDA:0 (NVIDIA GeForce RTX 3060 Ti, 7965MiB)

WARNING ⚠️ Upgrade to torch>=2.0.0 for deterministic training.

engine/trainer: task=detect, mode=train, model=yolov8n.pt, data=ultralytics/yolo/data/dataset/FLIR.yaml, epochs=150, patience=50, batch=16, imgsz=640, save=True, save_period=-1, cache=False, device=None, workers=8, project=None, name=None, 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, 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, show=False, save_txt=False, save_conf=False, save_crop=False, show_labels=True, show_conf=True, vid_stride=1, stream_buffer=False, line_width=None, visualize=False, augment=False, agnostic_nms=False, classes=None, retina_masks=False, boxes=True, 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, mosaic=1.0, mixup=0.0, copy_paste=0.0, cfg=None, tracker=botsort.yaml, save_dir=runs/detect/train9

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, 3011433 parameters, 3011417 gradients

Transferred 319/355 items from pretrained weights

Freezing layer 'model.22.dfl.conv.weight'

AMP: running Automatic Mixed Precision (AMP) checks with YOLOv8n...

AMP: checks passed ✅

Traceback (most recent call last):

  File "/home/wzy/anaconda3/envs/yolov8/bin/yolo", line 8, in <module>

    sys.exit(entrypoint())

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/cfg/__init__.py", line 445, in entrypoint

    getattr(model, mode)(**overrides)  # default args from model

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/engine/model.py", line 341, in train

    self.trainer.train()

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/engine/trainer.py", line 195, in train

    self._do_train(world_size)

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/engine/trainer.py", line 293, in _do_train

    self._setup_train(world_size)

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/engine/trainer.py", line 258, in _setup_train

    self.train_loader = self.get_dataloader(self.trainset, batch_size=batch_size, rank=RANK, mode='train')

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/models/yolo/detect/train.py", line 46, in get_dataloader

    dataset = self.build_dataset(dataset_path, mode, batch_size)

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/models/yolo/detect/train.py", line 40, in build_dataset

    return build_yolo_dataset(self.args, img_path, batch, self.data, mode=mode, rect=mode == 'val', stride=gs)

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/data/build.py", line 74, in build_yolo_dataset

    return YOLODataset(

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/data/dataset.py", line 40, in __init__

    super().__init__(*args, **kwargs)

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/data/base.py", line 72, in __init__

    self.labels = self.get_labels()

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/site-packages/ultralytics/data/dataset.py", line 97, in get_labels

    cache_path = Path(self.label_files[0]).parent.with_suffix('.cache')

  File "/home/wzy/anaconda3/envs/yolov8/lib/python3.8/pathlib.py", line 872, in with_suffix

    raise ValueError("%r has an empty name" % (self,))

ValueError: PosixPath('.') has an empty name

解决方法：

更改划分数据集的代码如下：

• 问题2：pytorch安装

30系列安装cuda11.3 （别装11.4，没有对应版本的torch和torchvision）

cuda11.3安装pytorch命令！！！

pip install torch==1.11.0+cu113 torchvision==0.12.0+cu113 torchaudio==0.11.0 --extra-index-url https://download.pytorch.org/whl/cu113

记得在requirements.txt文件中改为指定版本名称