We apply Parallel Decoding Distillation (PDD) 1 to MiniMax-H3, enabling efficient video generation in only a few inference steps. For more details, please refer to our GitHub repo. Set modelpath and pddlorapath to the MiniMax-H3 model and the matching acceleration LoRA checkpoint in predictt2v.py for FL2VA or predictref2v.py for Ref2VA, then run the corresponding script. Each example uses applypddlora to load the checkpoint and derive the required number of inference steps from its configuration.
Open weights
other
videox_fun
We explore the Reward Backpropagation technique 1 2 to optimized the generated videos by Wan2.2-Fun for better alignment with human preferences. We provide the following pre-trained models (i.e. LoRAs) along with the training script. You can use these LoRAs to enhance the corresponding base model as a plug-in or train your own reward LoRA. For more details, please refer to our GitHub repo. A panda eats bamboo while a monkey swings from branch to branch A dog runs through a field while a cat climbs a tree A penguin waddles on the ice, a camel treks by Pig with wings flying above a diamond mountain Set lorapath along with loraweight for the low noise reward LoRA, while specifying lorahighpath…
Open weights
apache-2.0
videox_fun
