This model is based on the conceptual slider LoRA model trained with FLUX.1D, which makes the character's hair in the generated pictures gradually change from long hair to baldness.weight: -5.0 to 5.0 （Not limited to this interval, you can try other ranges.）<pre><code> target_prompt: "female person"
 positive_prompt: "female person, bald"
 negative_prompt: "female person, longer hair"</code></pre>Code used for training: <a target="_blank" rel="ugc" href="https://github.com/SWHL/flux-sliders">https://github.com/SWHL/flux-sliders</a>Use this model:<pre><code>from datetime import datetime
from pathlib import Path

import torch
from diffusers import FluxPipeline

lora_path = "flux-hair_sliders_latest.safetensors"
pipe = FluxPipeline.from_pretrained("models/FLUX.1-dev", torch_dtype=torch.bfloat16)
pipe.to("cuda")
pipe.load_lora_weights(lora_path)

time_stamp = datetime.strftime(datetime.now(), "%Y-%m-%d-%H-%M-%S")
save_dir = Path("outputs") / time_stamp
save_dir.mkdir(parents=True, exist_ok=True)

scales = (-5, -2.5, 0, 2.5, 5)
prompt = "female person"

for scale in scales:
 out = pipe(
 prompt=prompt,
 guidance_scale=3.5,
 height=512,
 width=512,
 num_inference_steps=25,
 joint_attention_kwargs={"scale": scale * 1 / 16},
 generator=torch.Generator().manual_seed(42),
 ).images[0]

 save_img_path = save_dir / f"{time_stamp}_scale_{scale}.jpg"
 out.save(save_img_path)</code></pre>Enjoy it!⚠️ Notion:<ul><li>The current model does not support direct use of comfy.</li><li>Since the model has been converted, you need to try the slider scale multiple times to select the range you need.</li><li>ComfyUI will be supported later, so stay tuned.</li></ul>

Files with SHA256: 0bf6dd31c3114614f5e294a1f14f68346bc5bfb462fa21a8510193563d6304d5