Skip to content

How much paired video data is needed to train Pose-Control IC-LoRA? #217

@QChencq

Description

@QChencq

Thank you for releasing the LTX-2 models and the Pose-Control IC-LoRA checkpoint.
I am interested in training a similar Pose-Control IC-LoRA model based on LTX-2, and I would like to ask about the approximate amount of training data required. Specifically, for the released model:https://huggingface.co/Lightricks/LTX-2-19b-IC-LoRA-Pose-Control. Could you share roughly how many paired videos were used for training the Pose-Control IC-LoRA? For example, the number of reference-target video pairs, total video duration, or any recommended data scale for reproducing a similar pose-control capability.
I understand if the exact dataset is not publicly available, but even an approximate range or practical recommendation would be very helpful.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions