Framepack as an instruct/image edit model

I saw people who use WAN I2V as I2I instructions model and decided to use for the same framepack/hunyuan video. I wrote down the results on HF: by/u/Neph1010