hidream_roman_novel is a LORA trained with hidream-i1-full, so please check your base model before using it.

Details info.

Trigger = Novel_line_art

GPU VRAM USE = 38GB(fp8) and 43.12(fp16)

Recommended weight 1.0-1.5(should not exceed 1.8)

step 45-55

basic workflow in comfyui = hidream-i1-full

If you have any problem using it, please tell me

I did not test with GGUF, only base

Enjoy my friend