Pytorch-yolov3: Error while training

[codeyogi911]

!python3 train.py --model_def config/yolov3-custom.cfg --data_config config/custom.data --pretrained_weights weights/darknet53.conv.74

Namespace(batch_size=8, checkpoint_interval=1, compute_map=False, data_config='config/custom.data', epochs=100, evaluation_interval=1, gradient_accumulations=2, img_size=416, model_def='config/yolov3-custom.cfg', multiscale_training=True, n_cpu=8, pretrained_weights='weights/darknet53.conv.74')
Traceback (most recent call last):
  File "train.py", line 99, in <module>
    for batch_i, (_, imgs, targets) in enumerate(dataloader):
  File "/home/ubuntu/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 637, in __next__
    return self._process_next_batch(batch)
  File "/home/ubuntu/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 658, in _process_next_batch
    raise batch.exc_type(batch.exc_msg)
TypeError: Traceback (most recent call last):
  File "/home/ubuntu/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 138, in _worker_loop
    samples = collate_fn([dataset[i] for i in batch_indices])
  File "/home/ubuntu/anaconda3/envs/pytorch_p36/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 138, in <listcomp>
    samples = collate_fn([dataset[i] for i in batch_indices])
  File "/home/ubuntu/PyTorch-YOLOv3/utils/datasets.py", line 130, in __getitem__
    img, targets = horisontal_flip(img, targets)
  File "/home/ubuntu/PyTorch-YOLOv3/utils/augmentations.py", line 8, in horisontal_flip
    targets[:, 2] = 1 - targets[:, 2]
TypeError: 'NoneType' object is not subscriptable

I am getting this error for all images in my dataset except for some. Is there a contraint on the type or size of images? Am i missing something?

[codeyogi911]

I found the fix by changing
path.replace("images", "labels").replace(".png", ".txt").replace(".jpg", ".txt")
with
path.replace("images", "labels").replace(".png", ".txt").replace(".jpg", ".txt").replace(".JPG", ".txt")

in datasets.py line 65

[pxg0507]

hi I have the same error with you but when I try to use your way to fix it still not work

[TonightHitTiger1]

me too

[vansanety]

I found the fix by changing
path.replace("images", "labels").replace(".png", ".txt").replace(".jpg", ".txt")
with
path.replace("images", "labels").replace(".png", ".txt").replace(".jpg", ".txt").replace(".JPG", ".txt")

in datasets.py line 65

Thank you for this information and it solved my problem.