block concat outputs to save time? how?

[vgybhu-oss]

sorry to bother you, but gpt didn't answer this.

Can I block the exports on concat files?

of course, while they are interesting for study,
I would like to save time and only get the main outputs.

I don't know how to prohibit that when I run with python in a colab.

btw, this is a great program.
I haven't made a lot of use of it,
but I think I am going to do a round.

I will pair this for dialogue with Vidu ai for character consistency and action scenes, and hope for pretty cool animation movies.

somehow I'm starting to wish for better cohesion when the face is attached to the video, but for a rough funny animation, this is kick-ass.
I'm also seriously jonesing for Face/Off 2.
what's the hold up?

here's how I am implementing:
https://colab.research.google.com/drive/13gCm0yvakCnMmqJ2ZfXpDXzhMCpS6zgg
also,
here are the other tools I use, if anybody wants to know my "workflow" and strategies:
http://vgy.rf.gd/?i=mmfm
http://vgy.rf.gd/?i=ai

[cleardusk]

You can modify these code snippets.

LivePortrait/src/live_portrait_pipeline.py

Lines 454 to 464 in 7f755d9

	######### build the final concatenation result #########
	# driving frame | source frame | generation
	if flag_is_source_video and flag_is_driving_video:
	frames_concatenated = concat_frames(driving_rgb_crop_256x256_lst, img_crop_256x256_lst, I_p_lst)
	elif flag_is_source_video and not flag_is_driving_video:
	if flag_load_from_template:
	frames_concatenated = concat_frames(driving_rgb_crop_256x256_lst, img_crop_256x256_lst, I_p_lst)
	else:
	frames_concatenated = concat_frames(driving_rgb_crop_256x256_lst*n_frames, img_crop_256x256_lst, I_p_lst)
	else:
	frames_concatenated = concat_frames(driving_rgb_crop_256x256_lst, [img_crop_256x256], I_p_lst)