Views
No views yet
Flux2KleinPipeline, so you can use Flux2KleinPipeline.from_pretrained(...) directly without a custom train-order pipeline.1import torch
2from diffusers import Flux2KleinPipeline
3from PIL import Image
4
5model_id = "AliceKJ/BLOCKv0.6"
6
7pipe = Flux2KleinPipeline.from_pretrained(
8 model_id,
9 torch_dtype=torch.bfloat16,
10)
11pipe.to("cuda")
12
13preview = Image.open("examples/ex1_preview.png").convert("RGB").resize((512, 512))
14result = pipe(
15 prompt="Image-to-image translation using the reference image. The reference shows the same 3D Minecraft character with front and back views in a single image. Generate the corresponding Minecraft skin UV atlas in 64x64 pixel-art UV layout. High-quality anime-style. Flat shading, sharp pixel edges, no blur, no anti-aliasing. Keep consistent UV placement and mapping; match the same character design from the reference. Model type: classic (auto-detected Minecraft player model).",
16 image=preview,
17 num_inference_steps=30,
18 guidance_scale=4.0,
19).images[0]
20
21result.save("generated_skin.png")| Preview | Generated Skin |
|---|---|
 |  |
 |  |
 |  |
Flux2KleinPipeline now works directly.1@article{guo2026block,
2 title={BLOCK: An Open-Source Bi-Stage MLLM Character-to-Skin Pipeline for Minecraft},
3 author={Guo, Hengquan},
4 journal={arXiv preprint arXiv:2603.03964},
5 year={2026},
6 url={http://arxiv.org/abs/2603.03964}
7}