Pickle error when train on own data

[MarcusAnnaeus]

Hi!
If I try to train on a cumstom dataset I get the error:

PicklingError: Can't pickle <function rebuild_pipe_connection at 0x000001BDC94789D0>: it's not the same object as multiprocessing.connection.rebuild_pipe_connection

Traceback (most recent call last):
File "", line 1, in
File "C:\Program Files\Python310\lib\multiprocessing\spawn.py", line 116, in spawn_main
exitcode = _main(fd, parent_sentinel)
File "C:\Program Files\Python310\lib\multiprocessing\spawn.py", line 126, in _main
self = reduction.pickle.load(from_parent)
EOFError: Ran out of input

Could there be a problem with the input data?

I altered the "coco_detection.yml" in:

task: detection

num_classes: 1
remap_mscoco_category: False

train_dataloader:
type: DataLoader
dataset:
type: CocoDetection
img_folder: "[.....]/images/"
ann_file: "[.....]/result.json"
transforms:
type: Compose
ops: ~
shuffle: True
batch_size: 2
num_workers: 1
drop_last: True

val_dataloader:
type: DataLoader
dataset:
type: CocoDetection
img_folder: "[.....]/val/"
ann_file: "[.....]/val/result.json"
transforms:
type: Compose
ops: ~

shuffle: False
batch_size: 1
num_workers: 1

Does someone have an idea?
Thanks a lot!