I have recreated this from my XL Lora for better improving eyes. If you have issues with generating in text prompt to image, try doing Inpainting.
Eye_Detail_Flux_Lora_-_Inpainting.safetensors