Like the title says, the lora removes watermarks or subtitle text. Will be able to do most jobs, but I am expanding the dataset and will continue training for a more robust v2 which will be able to handle more complex tasks. If it struggles with the removal, increase the strength of the lora.
This is a V2V IC-Lora, so you have to input your video via the LTX Add Video IC-Lora Guide node (workflow in example video).