Transfer from huggingface

https://huggingface.co/alibaba-pai/Wan2.2-Fun-Reward-LoRAs/tree/main

sample of workflow: https://www.runninghub.ai/post/1966777795378655234

This is one file from a LoRA set, which needs to be used in combination with other LoRAs. It is recommended to refer to the example workflow before downloading and using it.

Introduction

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.

Official MPS reward LoRA (rank=128 and network_alpha=64) for Wan2.2-Fun-A14B-InP (high noise). It is trained with a batch size of 8 for 5,000 steps.