Jupiter notebookでのValueError: not enough values to unpack (expected 3, got 0)というエラー

実現したいこと

jupyter notebookでpythonを用いて、自作のデータセットをYOLOv8を使用して画像学習させたい

発生している問題・分からないこと

from ultralytics import YOLO
model = YOLO("yolov8n.pt")
model.train(data="datasets.yaml", epochs=5, batch=8, workers=4, degrees=45.0)
というコードを実行したら下記のエラーが出てきた

エラーメッセージ

error
1ValueError                                Traceback (most recent call last)
2Cell In[1], line 23
3     21 from ultralytics import YOLO
4     22 model = YOLO("yolov8n.pt")
5---> 23 model.train(data="datasets.yaml", epochs=5, batch=8, workers=4, degrees=45.0)
6
7File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\engine\model.py:802, in Model.train(self, trainer, **kwargs)
8    799     self.model = self.trainer.model
9    801 self.trainer.hub_session = self.session  # attach optional HUB session
10--> 802 self.trainer.train()
11    803 # Update model and cfg after training
12    804 if RANK in {-1, 0}:
13
14File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\engine\trainer.py:207, in BaseTrainer.train(self)
15    204         ddp_cleanup(self, str(file))
16    206 else:
17--> 207     self._do_train(world_size)
18
19File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\engine\trainer.py:327, in BaseTrainer._do_train(self, world_size)
20    325 if world_size > 1:
21    326     self._setup_ddp(world_size)
22--> 327 self._setup_train(world_size)
23    329 nb = len(self.train_loader)  # number of batches
24    330 nw = max(round(self.args.warmup_epochs * nb), 100) if self.args.warmup_epochs > 0 else -1  # warmup iterations
25
26File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\engine\trainer.py:291, in BaseTrainer._setup_train(self, world_size)
27    289 # Dataloaders
28    290 batch_size = self.batch_size // max(world_size, 1)
29--> 291 self.train_loader = self.get_dataloader(self.trainset, batch_size=batch_size, rank=LOCAL_RANK, mode="train")
30    292 if RANK in {-1, 0}:
31    293     # Note: When training DOTA dataset, double batch size could get OOM on images with >2000 objects.
32    294     self.test_loader = self.get_dataloader(
33    295         self.testset, batch_size=batch_size if self.args.task == "obb" else batch_size * 2, rank=-1, mode="val"
34    296     )
35
36File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\models\yolo\detect\train.py:49, in DetectionTrainer.get_dataloader(self, dataset_path, batch_size, rank, mode)
37     47 assert mode in {"train", "val"}, f"Mode must be 'train' or 'val', not {mode}."
38     48 with torch_distributed_zero_first(rank):  # init dataset *.cache only once if DDP
39---> 49     dataset = self.build_dataset(dataset_path, mode, batch_size)
40     50 shuffle = mode == "train"
41     51 if getattr(dataset, "rect", False) and shuffle:
42
43File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\models\yolo\detect\train.py:43, in DetectionTrainer.build_dataset(self, img_path, mode, batch)
44     34 """
45     35 Build YOLO Dataset.
46     36 
47   (...)
48     40     batch (int, optional): Size of batches, this is for `rect`. Defaults to None.
49     41 """
50     42 gs = max(int(de_parallel(self.model).stride.max() if self.model else 0), 32)
51---> 43 return build_yolo_dataset(self.args, img_path, batch, self.data, mode=mode, rect=mode == "val", stride=gs)
52
53File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\data\build.py:87, in build_yolo_dataset(cfg, img_path, batch, data, mode, rect, stride, multi_modal)
54     85 """Build YOLO Dataset."""
55     86 dataset = YOLOMultiModalDataset if multi_modal else YOLODataset
56---> 87 return dataset(
57     88     img_path=img_path,
58     89     imgsz=cfg.imgsz,
59     90     batch_size=batch,
60     91     augment=mode == "train",  # augmentation
61     92     hyp=cfg,  # TODO: probably add a get_hyps_from_cfg function
62     93     rect=cfg.rect or rect,  # rectangular batches
63     94     cache=cfg.cache or None,
64     95     single_cls=cfg.single_cls or False,
65     96     stride=int(stride),
66     97     pad=0.0 if mode == "train" else 0.5,
67     98     prefix=colorstr(f"{mode}: "),
68     99     task=cfg.task,
69    100     classes=cfg.classes,
70    101     data=data,
71    102     fraction=cfg.fraction if mode == "train" else 1.0,
72    103 )
73
74File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\data\dataset.py:64, in YOLODataset.__init__(self, data, task, *args, **kwargs)
75     62 self.data = data
76     63 assert not (self.use_segments and self.use_keypoints), "Can not use both segments and keypoints."
77---> 64 super().__init__(*args, **kwargs)
78
79File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\data\base.py:74, in BaseDataset.__init__(self, img_path, imgsz, cache, augment, hyp, prefix, rect, batch_size, stride, pad, single_cls, classes, fraction)
80     72 self.fraction = fraction
81     73 self.im_files = self.get_img_files(self.img_path)
82---> 74 self.labels = self.get_labels()
83     75 self.update_labels(include_class=classes)  # single_cls and include_class
84     76 self.ni = len(self.labels)  # number of images
85
86File ~\anaconda3\envs\Aidemy\Lib\site-packages\ultralytics\data\dataset.py:161, in YOLODataset.get_labels(self)
87    159 # Check if the dataset is all boxes or all segments
88    160 lengths = ((len(lb["cls"]), len(lb["bboxes"]), len(lb["segments"])) for lb in labels)
89--> 161 len_cls, len_boxes, len_segments = (sum(x) for x in zip(*lengths))
90    162 if len_segments and len_boxes != len_segments:
91    163     LOGGER.warning(
92    164         f"WARNING ⚠️ Box and segment counts should be equal, but got len(segments) = {len_segments}, "
93    165         f"len(boxes) = {len_boxes}. To resolve this only boxes will be used and all segments will be removed. "
94    166         "To avoid this please supply either a detect or segment dataset, not a detect-segment mixed dataset."
95    167     )
96ValueError: not enough values to unpack (expected 3, got 0)

該当のソースコード

python
1from ultralytics import YOLO
2model = YOLO("yolov8n.pt")
3model.train(data="datasets.yaml", epochs=5, batch=8, workers=4, degrees=45.0)

yamlファイル
1# Path
2path: ../datasets  # dataset root directory
3train: images/train  # train images (relative to 'path')
4val: images/val      # val images (relative to 'path')
5# test: images/test  # optional, if you have test data
6
7# Class names
8names:
9  0: manchester city
10  1: manchester united
11  2: real madrid