How to convert BF16 safetensors Generative AI models into NVFP4 like Qwen 2512, Wan 2.2, LTX 2, FLUX SRPO or any other model

[FurkanGozukara]

Can you generalize your script to support all models

I want to compile FLUX SRPO, Wan 2.2, Qwen 2512 and 2511 and more models

Please show some instructions examples and direction