🥯 BAGEL-NHR-Edit

🌐 NHR Website | 📜 NHR Paper on arXiv | 🤗 NHR-Edit Dataset |

This repository hosts the model weights for BAGEL, fine-tuned on the NHR-Edit dataset. For installation, usage instructions, and further documentation, please visit the official BAGEL GitHub repository.

🛠️ Training Setup

We performed parameter-efficient adaptation on the generation expert’s attention and FFN projection layers using LoRA.

LoRA parameters:

r = 16
lora_alpha = 16
dropout = 0.05
bias = "none"
target_modules = [
  "v_proj_moe_gen",
  "k_proj_moe_gen",
  "mlp_moe_gen.down_proj",
  "mlp_moe_gen.gate_proj",
  "q_proj_moe_gen",
  "mlp_moe_gen.up_proj",
  "o_proj_moe_gen"
]

📊 Image Editing Metrics

Metrics for GEdit-Bench-EN:

Model	GEdit-Bench-EN (SC) ↑	GEdit-Bench-EN (PQ) ↑	GEdit-Bench-EN (O) ↑
BAGEL-7B-MoT	7.983	6.570	6.921
BAGEL-NHR-Edit	8.067	6.881	7.115

Scoring model: gpt-4.1-2025-04-14 (with default temperature)

Metrics for ImgEdit-Bench:

Model	Style	Extract	Remove	Background	Action	Adjust	Add	Replace	Compose	Overall ↑
BAGEL-7B-MoT	4.22	1.53	3.04	3.3	4.07	3.67	3.98	3.5	3.0	3.3
BAGEL-NHR-Edit	4.3	1.62	3.18	3.42	3.95	3.55	4.19	3.77	2.94	3.39

Scoring model: gpt-4o-2024-11-20 (with temperature = 0.0)

🖼️ Image Editing Results

Generated images for ImgEdit-Bench and GEdit-Bench are included in this repository.

Results comparison between original Bagel-7B-MoT and BAGEL-NHR-EDIT on samples from ImgEdit and GEdit benches:

License

BAGEL-NHR-Edit is licensed under the Apache 2.0 license. It is finetuned from ByteDance-Seed/BAGEL-7B-MoT, which is also licensed under Apache 2.0.

✍️ Citation

@article{Layer2025NoHumansRequired,
    arxivId = {2507.14119},
    author = {Maksim Kuprashevich and Grigorii Alekseenko and Irina Tolstykh and Georgii Fedorov and Bulat Suleimanov and Vladimir Dokholyan and Aleksandr Gordeev},
    title = {{NoHumansRequired: Autonomous High-Quality Image Editing Triplet Mining}},
    year = {2025},
    eprint = {2507.14119},
    archivePrefix = {arXiv},
    primaryClass = {cs.CV},
    url = {https://arxiv.org/abs/2507.14119},
    journal={arXiv preprint arXiv:2507.14119}
}