import torch
from diffusers import Flux2Pipeline

pipe = Flux2Pipeline.from_pretrained("black-forest-labs/FLUX.2-dev", torch_dtype=torch.bfloat16).to("cuda")
pipe.load_lora_weights("danildushenev/pneumothorax-lora-flux-2-dev", weight_name="pytorch_lora_weights.safetensors")
image = pipe("your prompt", num_inference_steps=24, guidance_scale=2.5).images[0]
image.save("result.png")

Usage (ComfyUI)

Download pytorch_lora_weights.safetensors from this repo.
Put it into ComfyUI/models/loras/.
In workflow use LoRA Loader and set this file.

Trigger text: a photo of Human with pneumothorax

Generated on 2026-03-11 19:44:14Z.