数据集文件路径问题

 

data_root = '/root/autodl-tmp/COCO/whole_images/whole_images/'

train_dataset = dict(
    type='MultiImageMixDataset',
    dataset=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_train2017.json',
        img_prefix=data_root + 'train2017/',
        pipeline=[
            dict(type='LoadImageFromFile'),
            dict(type='LoadAnnotations', with_bbox=True)
        ],
        filter_empty_gt=False,
    ),
    pipeline=train_pipeline)

于是得到train_dataset这个字典里面的dataset里面的ann_file是下面这个路径

ann_file = '/root/autodl-tmp/COCO/whole_images/whole_images/annotations/instances_train2017.json'

从打印的日志进行验证自己的猜测也可以

img_prefix也是同样的道理

img_prefix = '/root/autodl-tmp/COCO/whole_images/whole_images/train2017/'

 下面这里data也是同样的道理

data = dict(
    samples_per_gpu=8,
    workers_per_gpu=4,
    persistent_workers=True,
    train=train_dataset,
    val=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017.json',
        img_prefix=data_root + 'val2017/',
        pipeline=test_pipeline),
    test=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017.json',
        img_prefix=data_root + 'val2017/',
        pipeline=test_pipeline))

 train=train_dataset,
    val=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017.json',
        img_prefix=data_root + 'val2017/',
        pipeline=test_pipeline),
    test=dict(
        type=dataset_type,
        ann_file=data_root + 'annotations/instances_val2017.json',
        img_prefix=data_root + 'val2017/',
        pipeline=test_pipeline))

比对一下文件夹结构