Hello! I liked your approach to solve the problem with dedicated training rather than relying entirely on the imputation or classifier guidance.
BTW, I observed jitterness of the xz trajectory (on the plain) when I ran
python -m sample.conditional_synthesis --model_path ./save/condmdi_randomframes/model000750000.pt --edit_mode uncond --num_samples 10 --num_repetitions 3
and saw in sample01_rep00.mp4. Am I doing something wrong?
Regards
Hello! I liked your approach to solve the problem with dedicated training rather than relying entirely on the imputation or classifier guidance.
BTW, I observed jitterness of the xz trajectory (on the plain) when I ran
and saw in
sample01_rep00.mp4. Am I doing something wrong?Regards