I want to fine-tune a First-Last Frame2Video model with WAN. How do I do that?

[ydove0324]

As stated in the prompt, I specifically want to train starting from WanT2V1.3B (so it’s smaller). In this framework, how should I do that? If I need to make any changes, what should I modify?”

[a-r-r-o-w]

You can take a look at this folder containing examples for Wan T2V: https://github.com/a-r-r-o-w/finetrainers/tree/main/examples/training/sft/wan

I believe the only change needed to make FLF2V training to work is to change the --pretrained_model_name_or_path to https://huggingface.co/Wan-AI/Wan2.1-FLF2V-14B-720P-diffusers