尊敬的作者您好，很抱歉又来打扰您，当我用服务器运行python inference_resshift.py -i /mnt/ResShift-journal/testdata/inpainting/face/lq_mark -o /mnt/RS --mask_path /mnt/ResShift-journal/testdata/inpainting/face/mask --task inpaint_face --scale 1
这一命令时，出现Traceback (most recent call last):
File "/mnt/ResShift-journal/inference_resshift.py", line 197, in
main()
File "/mnt/ResShift-journal/inference_resshift.py", line 188, in main
resshift_sampler.inference(
File "/mnt/ResShift-journal/sampler.py", line 275, in inference
for data in dataloader:
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 630, in next
data = self._next_data()
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 674, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 51, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 51, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/mnt/ResShift-journal/datapipe/datasets.py", line 471, in getitem
im_mask = util_image.imread(im_path, chn='gray', dtype='float32')
File "/mnt/ResShift-journal/utils/util_image.py", line 465, in imread
assert im.ndim == 2, f"{str(path)} has {im.ndim} channels!"
AttributeError: 'NoneType' object has no attribute 'ndim'
我找到该项目下utils的util_image.py，运行发现NameError: name 'zero_pad' is not defined，请问该如何解决这些问题，到扰您了。
尊敬的作者您好，很抱歉又来打扰您，当我用服务器运行python inference_resshift.py -i /mnt/ResShift-journal/testdata/inpainting/face/lq_mark -o /mnt/RS --mask_path /mnt/ResShift-journal/testdata/inpainting/face/mask --task inpaint_face --scale 1 这一命令时，出现Traceback (most recent call last): File "/mnt/ResShift-journal/inference_resshift.py", line 197, in
main()
File "/mnt/ResShift-journal/inference_resshift.py", line 188, in main
resshift_sampler.inference(
File "/mnt/ResShift-journal/sampler.py", line 275, in inference
for data in dataloader:
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 630, in next
data = self._next_data()
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/dataloader.py", line 674, in _next_data
data = self._dataset_fetcher.fetch(index) # may raise StopIteration
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 51, in fetch
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/root/miniconda3/envs/resshift/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py", line 51, in
data = [self.dataset[idx] for idx in possibly_batched_index]
File "/mnt/ResShift-journal/datapipe/datasets.py", line 471, in getitem
im_mask = util_image.imread(im_path, chn='gray', dtype='float32')
File "/mnt/ResShift-journal/utils/util_image.py", line 465, in imread
assert im.ndim == 2, f"{str(path)} has {im.ndim} channels!"
AttributeError: 'NoneType' object has no attribute 'ndim'
我找到该项目下utils的util_image.py，运行发现NameError: name 'zero_pad' is not defined，请问该如何解决这些问题，到扰您了。