Improve model card: add metadata, paper link, and usage instructions

#1
by nielsr HF Staff - opened
Files changed (1) hide show
  1. README.md +44 -4
README.md CHANGED
@@ -5,13 +5,21 @@ tags:
5
  - image-editing
6
  - cref
7
  - sref
 
 
8
  ---
9
 
10
- # FreeStyle Checkpoints
11
 
12
- This repository contains LoRA checkpoints for FreeStyle CRef/SRef image editing.
 
 
 
 
 
 
13
 
14
- ## Checkpoint Folders
15
 
16
  | Folder | What it is | Inference note |
17
  | --- | --- | --- |
@@ -21,4 +29,36 @@ This repository contains LoRA checkpoints for FreeStyle CRef/SRef image editing.
21
 
22
  Each folder contains a `model.safetensors` LoRA weight file.
23
 
24
- The previous `freestyle-cref-sref-40000-no-rope/` folder name was corrected to `freestyle-cref-sref-40000-rope/` to reflect the RoPE inference requirement.
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
5
  - image-editing
6
  - cref
7
  - sref
8
+ pipeline_tag: image-to-image
9
+ license: cc-by-nc-4.0
10
  ---
11
 
12
+ # FreeStyle: Free Control of Style-Content Dual-Reference Generation from Community LoRA Mining
13
 
14
+ FreeStyle is a scalable dual-reference generation framework that allows for synthesizing images by combining the structure and semantics of a content reference with the style of a separate reference image. It leverages community LoRA weights as compositional anchors and introduces mechanisms like attention-level enrichment constraints and frequency-aware RoPE modulation to prevent content leakage from style references.
15
+
16
+ - **Paper:** [FreeStyle: Free Control of Style-Content Dual-Reference Generation from Community LoRA Mining](https://huggingface.co/papers/2606.20506)
17
+ - **Project Page:** [https://blue2giant.github.io/FreeStyle/](https://blue2giant.github.io/FreeStyle/)
18
+ - **GitHub Repository:** [https://github.com/Blue2Giant/FreeStyle](https://github.com/Blue2Giant/FreeStyle)
19
+
20
+ ## FreeStyle Checkpoints
21
 
22
+ This repository contains LoRA checkpoints for FreeStyle CRef/SRef image editing.
23
 
24
  | Folder | What it is | Inference note |
25
  | --- | --- | --- |
 
29
 
30
  Each folder contains a `model.safetensors` LoRA weight file.
31
 
32
+ ## Sample Usage
33
+
34
+ To generate an image using the FreeStyle model, you can use the inference script provided in the [official GitHub repository](https://github.com/Blue2Giant/FreeStyle).
35
+
36
+ ```bash
37
+ cd model_infer
38
+ # Follow environment setup instructions in the GitHub repository
39
+ conda activate Sref
40
+
41
+ python3 cref_sref_core_infer.py \
42
+ assets/00-cref.jpg \
43
+ assets/00-sref.jpg \
44
+ 'Transfer the style of image 2 onto image 1, keeping image 1\'s layout.' \
45
+ --weight_preset sref_14000 \
46
+ --recaption_task_type style_transfer \
47
+ --out_dir outputs/demo \
48
+ --steps 28 --cfg 8 --seed 42 --overwrite
49
+ ```
50
+
51
+ ## Citation
52
+
53
+ ```bibtex
54
+ @article{lan2026freestyle,
55
+ title = {FreeStyle: Free Control of Style-Content Dual-Reference Generation from Community LoRA Mining},
56
+ author = {Lan, Jinghong and Cheng, Wei and Chen, Yunuo and Ye, Ziqi and Xing, Peng and Fang, Yixiao and Wang, Rui and Yang, Yufeng and Zhang, Xuanyang and Zou, Difan and Zeng, Xianfang and Yu, Gang and Zhang, Chi},
57
+ journal = {arXiv preprint arXiv:2606.20506},
58
+ year = {2026}
59
+ }
60
+ ```
61
+
62
+ ## Disclaimer
63
+
64
+ This project is released **strictly for academic research and non-commercial use only**. Users are responsible for ensuring that their use of this project complies with all applicable laws and third-party terms of service.