basiclab / cpcstoryvisualization-pytorch Goto Github PK

I want to evaluate the model but I can't find './script_inference.sh' file. (based on README.md)

So I set the cfg file into './output/final_model/setting.yml', but SameFileError occurs like this.

Traceback (most recent call last):
  File "/home/user01/wooseok/CPCStoryVisualization-Pytorch/main_pororo.py", line 137, in <module>
    algo = GANTrainer(output_dir, args, ratio=1.0)
  File "/home/user01/wooseok/CPCStoryVisualization-Pytorch/trainer.py", line 56, in __init__
    copyfile(args.cfg_file, output_dir + 'setting.yml')
  File "/home/user01/anaconda3/envs/video1/lib/python3.7/shutil.py", line 104, in copyfile
    raise SameFileError("{!r} and {!r} are the same file".format(src, dst))
shutil.SameFileError: './output/final_model/setting.yml' and './output/final_model/setting.yml' are the same file

Can you tell the command or upload 'script_inference.sh' file to evaluate the model?

Hello! I'm doing some research and would like to evaluate the model but the files are unavailable when I click on the links. Is there any way I can get access to it? 你好! 我在做研究，想使用你的模型，但當我點鏈接時，文件是不可用的。有什麼辦法能讓我獲得它嗎？

Hi
Please assist getting the below error when running the main_pororo.py
File "main_pororo.py", line 87, in
video_transforms = functools.partial(video_transform, image_transform=image_transforms) # Only need to feed video later
TypeError: the first argument must be callable

Hi, I am very interested in your paper and the task to generate a story from an image sequence.
I have some questions about the preprocess_pororo.py.

1. The code does not preprocess
   I found the code process nothing except for printing file names in the Pororo dataset.

if __name__ == "__main__":
	import os, glob, PIL
	for filename in glob.glob('/mnt/storage/pororoSV/Scenes_Dialogues/*/*/*.gif'):
		img_name = os.path.basename(filename)
		print(img_name)

How do I modify the code to preprocess the dataset?

2. How can I get pororo_png/descriptions_vec.npy file?
   At L62 on preprocess_pororo.py, numpy file is loaded as:

descriptions_vec = np.load("pororo_png/descriptions_vec.npy", encoding='bytes').item()

But, I have no idea how to get this file. Could you clarify it (any documents)?