5k.txt / 5k.part file extension

[CF2220160244]

when i run :
~/PycharmProjects/yolov3-master$ python test.py -weights_path checkpoints/latest.pt
there is error in dataset.py:

img_all = np.stack(img_all)[:, :, :, ::-1].transpose(0, 3, 1, 2) #row 218(dataset.py)

Traceback (most recent call last):
File "test.py", line 59, in
for batch_i, (imgs, targets) in enumerate(dataloader):
File "/home/chenfei/PycharmProjects/yolov3-master/utils/datasets.py", line 218, in next
img_all = np.stack(img_all)[:, :, :, ::-1].transpose(0, 3, 1, 2) # BGR to RGB and cv2 to pytorch
File "/home/chenfei/anaconda3/lib/python3.6/site-packages/numpy/core/shape_base.py", line 349, in stack
raise ValueError('need at least one array to stack')
ValueError: need at least one array to stack

can you help me ?

[CF2220160244]

i find why there is such a problem,that is in ubuntu16.04 i should use
5k.txt rather than 5k.part in test.py.

test_path = '~/coco/images/5k.txt' ###row 31

[glenn-jocher]

Ah yes, be careful of the extension used here. COCO naturally downloads as 5k.part I believe, but I have the 5k.txt extension coded into codo.data. I understand your confusion. You can modify this section to point to your 5k.* file:

yolov3/test.py

Lines 27 to 30 in 68de92f

	if platform == 'darwin': # MacOS (local)
	test_path = data_config['valid']
	else: # linux (cloud, i.e. gcp)
	test_path = '../coco/5k.part'

If you are using coco.data then make sure to update these two lines as well to point to your coco directory:

yolov3/cfg/coco.data

Lines 2 to 3 in 68de92f

	train=/Users/glennjocher/Downloads/DATA/coco/trainvalno5k.txt
	valid=/Users/glennjocher/Downloads/DATA/coco/5k.txt

[TommeyChang]

I have tryed to download the coco dataset, but occured a problem. How I can solve it? Thank you
get_coco_dataset.sh: 31: get_coco_dataset.sh: Syntax error: "(" unexpected

[CF2220160244]

thanks! @glenn-jocher

[jaelim]

This solution still does not resolve the same problem I'm having.

[jaelim]

The difficulty I'm having is that there are many hard coded directory paths in dataset.py that are conflicting with my data directory defined in my cfg file.

Also, the platform check in training.py is probably not even necessary.

[glenn-jocher]

I've resolved this issue now:

• *.part files are the local image paths (with respect to the *.part file)
• *.txt files are the absolute image paths