Reputation: 1
I'm trying to run a few keyframe prompts for an animation test, however after the run starts and a haze based image forms, it only completes the first few passes and then hits me with this:
Seed used: 1412665177
---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
<ipython-input-19-95fd5446980a> in <module>()
171 torch.cuda.empty_cache()
172 try:
--> 173 do_run()
174 except KeyboardInterrupt:
175 pass
<ipython-input-8-f0c50060b30a> in do_run()
366 else:
367 img_0 = cv2.imread('prevFrame.png')
--> 368 center = (1*img_0.shape[1]//2, 1*img_0.shape[0]//2)
369 trans_mat = np.float32(
370 [[1, 0, translation_x],
AttributeError: 'NoneType' object has no attribute 'shape'
So far I've not been able to figure it out through normal StackOverflow answers.
Upvotes: 0
Views: 2960
Reputation: 1
I think I may have found the issue: My prompts looked like this, divided by so many (""). So I strung the prompt together using the commas and less (""):
0:["A beautiful ultradetailed painting of a single moon setting behind a mountain", "Joe Sorren", "J. Frederick Smith", "4k", "trending on artstation:3", "Aerial Photography", "Ultrawide lens", "People:-2", "Dof:-1", "blur:-1"],
\ TO THIS > \
0:["A beautiful ultradetailed painting of a single moon setting behind a mountain, Joe Sorren, J. Frederick Smith, 4k, trending on artstation:3", "Aerial Photography", "Ultrawide lens", "People:-2", "Dof:-1", "blur:-1"],
Now it runs.
Upvotes: 0