You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
Lack of Computational resource consumption in paper #12
Specifically, For the described two-stage training (CogVideoX-5B-I2V + ControlNet) with the given hyperparameters, what are the training duration (days/hours), VRAM usage, and GPU(s) used? Particularly interested in the details related to the Style Transfer task.
We use CogVideoX-5B-I2V(Yang et al., 2024) as the base
model and integrate it with ControlNet to leverage the edited
first frame to guide the editing process. The batch size of
the editing model’s training is 32, the learning rate is 1e-5,
and weight decay is 1e-4. We train model for 2 epoch. We
sample 33 frames of the videos to train with a resolution
of 336×592 in first stage. Different from the first stage, we
use higher resolution of 448×768 and batch size of 16 in
stage two, fintuning with 1 epoch to help model edit high
resolution
Activity
zibojia commentedon Oct 31, 2025
The final version has these information.
greasebig commentedon Nov 3, 2025
Specifically, For the described two-stage training (CogVideoX-5B-I2V + ControlNet) with the given hyperparameters, what are the training duration (days/hours), VRAM usage, and GPU(s) used? Particularly interested in the details related to the Style Transfer task.