We explore the Reward Backpropagation technique <a target="_blank" rel="nofollow" href="https://huggingface.co/alibaba-pai/Wan2.1-Fun-Reward-LoRAs#ref1">1</a> <a target="_blank" rel="nofollow" href="https://huggingface.co/alibaba-pai/Wan2.1-Fun-Reward-LoRAs#ref2">2</a> to optimized the generated videos by <a target="_blank" rel="nofollow" href="https://github.com/aigc-apps/VideoX-Fun">Wan2.1-Fun</a> for better alignment with human preferences. We provide the following pre-trained models (i.e. LoRAs) along with <a target="_blank" rel="nofollow" href="https://github.com/aigc-apps/VideoX-Fun/blob/main/scripts/wan2.1_fun/train_reward_lora.py">the training script</a>. 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 <a target="_blank" rel="nofollow" href="https://github.com/aigc-apps/VideoX-Fun">GitHub repo</a>.

Files with SHA256: d5a8582d7a1a671e0485724fad9fe70caf7b061a01d2ec352e4998b86e5764c1