WZMIAOMIAO/deep-learning-for-image-processing · error · FileNotFoundError

VOCdevkit dose not in path:'{}'.

Error message

VOCdevkit dose not in path:'{}'.

What it means

train.py's main() verifies that args.data_path contains a VOCdevkit subdirectory before constructing VOCDataSet. If the path does not exist, it raises FileNotFoundError naming the configured root. This is an early, explicit guard against running with a wrong/empty data root, since the dataset classes would otherwise fail later with more confusing errors.

Source

Thrown at pytorch_object_detection/retinaNet/train.py:53

    return model


def main(args):
    device = torch.device(args.device if torch.cuda.is_available() else "cpu")
    print("Using {} device training.".format(device.type))

    results_file = "results{}.txt".format(datetime.datetime.now().strftime("%Y%m%d-%H%M%S"))

    data_transform = {
        "train": transforms.Compose([transforms.ToTensor(),
                                     transforms.RandomHorizontalFlip(0.5)]),
        "val": transforms.Compose([transforms.ToTensor()])
    }

    VOC_root = args.data_path
    # check voc root
    if os.path.exists(os.path.join(VOC_root, "VOCdevkit")) is False:
        raise FileNotFoundError("VOCdevkit dose not in path:'{}'.".format(VOC_root))

    # load train data set
    # VOCdevkit -> VOC2012 -> ImageSets -> Main -> train.txt
    train_dataset = VOCDataSet(VOC_root, "2012", data_transform["train"], "train.txt")
    train_sampler = None

    # 是否按图片相似高宽比采样图片组成batch
    # 使用的话能够减小训练时所需GPU显存，默认使用
    if args.aspect_ratio_group_factor >= 0:
        train_sampler = torch.utils.data.RandomSampler(train_dataset)
        # 统计所有图像高宽比例在bins区间中的位置索引
        group_ids = create_aspect_ratio_groups(train_dataset, k=args.aspect_ratio_group_factor)
        # 每个batch图片从同一高宽比例区间中取
        train_batch_sampler = GroupedBatchSampler(train_sampler, group_ids, args.batch_size)

    # 注意这里的collate_fn是自定义的，因为读取的数据包括image和targets，不能直接使用默认的方法合成batch
    batch_size = args.batch_size
    nw = min([os.cpu_count(), batch_size if batch_size > 1 else 0, 8])  # number of workers

View on GitHub (pinned to 1ec3fe6f37)

Solutions

1. Pass the directory that directly contains VOCdevkit: --data-path /data (with /data/VOCdevkit/VOC2012/...).
2. Download and extract VOC2012 into the data root so VOCdevkit exists.
3. Fix extraction nesting: if you got root/VOCdevkit/VOCdevkit/..., move the inner folder up.
4. Verify with ls $DATA_PATH/VOCdevkit before launching training.

Example fix

// before
python train.py --data-path /data/VOCdevkit   # wrong: points inside the root
// after
python train.py --data-path /data             # /data/VOCdevkit must exist

import os
VOC_root = args.data_path
if not os.path.isdir(os.path.join(VOC_root, "VOCdevkit")):
    raise FileNotFoundError(f"VOCdevkit not found under {VOC_root}; pass the parent directory")

def has_voc_root(data_path: str) -> bool:
    import os
    return os.path.isdir(os.path.join(data_path, "VOCdevkit", "VOC2012"))

try:
    main(args)
except FileNotFoundError as e:
    if "VOCdevkit" in str(e):
        print(f"Fix --data-path (currently {args.data_path}); it must contain VOCdevkit/")
        sys.exit(1)
    raise

When it happens

Trigger: Running python train.py --data-path <wrong dir>; data_path pointing at the parent of the parent (should be the dir containing VOCdevkit/); dataset not yet downloaded/extracted; typo in path or running from a different working directory with a relative path.

Common situations: Cloning the repo without downloading VOC2012; extracting the archive so the structure becomes root/VOC2012/... without the VOCdevkit folder level; Docker/colab setups where the dataset volume is mounted elsewhere; passing the VOCdevkit path itself instead of its parent.

Related errors

• image: {} isn't RGB mode.
• dataset have {} classes, but input {}
• dataset have {} classes, but input {}
• image: {} isn't RGB mode.
• image: {} isn't RGB mode.

AI-assisted analysis of WZMIAOMIAO/deep-learning-for-image-processing@1ec3fe6f37 (2026-08-30). Data as JSON: /api/errors/2d49561e170d5421. Report an issue: GitHub.