Creative Graphic Design Lab

Exploring the intersection of 𝒉𝒖𝒎𝒂𝒏 creativity 👥 and 𝑨𝑰 🤖 to achieve new heights in design aesthetics. We're unlocking technology to reach this goal.

[Creative Graphic Design Lab @cgd_lab]

🔗 Links & Resources:

📄 The compliance study (Highlighting our Dataset Card):
https://t.co/hK3B4u781G

🗄️ Our Hugging Face release (GenPoster100K):
https://t.co/9bpnmPTbqk

📝 The original dataset paper (SAGE):
https://t.co/dfhZYJsEsp

[Creative Graphic Design Lab @cgd_lab]

Proud that our HF port of GenPoster100k (org. SAGE paper) was featured in a recent AI compliance study!

Out of 11,682 datasets, ours was named an 𝒆𝒙𝒄𝒆𝒑𝒕𝒊𝒐𝒏𝒂𝒍𝒍𝒚 𝒓𝒂𝒓𝒆, gold-standard example of a highly detailed Dataset Card.

🧵 Links to the papers & dataset below

[Creative Graphic Design Lab @cgd_lab]

What if we applied the concept 𝐾𝑎𝑟𝑝𝑎𝑡ℎ𝑦'𝑠 𝐿𝐿𝑀 𝑤𝑖𝑘𝑖 to creative graphic design?

[Creative Graphic Design Lab @cgd_lab]

@huggingface https://t.co/aEM0QEncdN

[Creative Graphic Design Lab @cgd_lab]

We’re building 𝘾𝙧𝙚𝙖𝙩𝙞𝙫𝙚 𝙂𝙧𝙖𝙥𝙝𝙞𝙘 𝘿𝙚𝙨𝙞𝙜𝙣 𝘿𝙖𝙩𝙖𝙨𝙚𝙩𝙨, a growing collection of 20+ Hugging Face🤗-ready datasets for graphic design generation and evaluation.

From posters and banners to UIs, typography, layouts, ads, evaluation, and more.

Link below 👇

[Creative Graphic Design Lab @cgd_lab]

Released a FastAPI server for image layer decomposition using Qwen-Image-Layered!

✨️ Automatically decompose images into semantic layers
⚡️ FastAPI + GPU acceleration
🛠️ Simple REST API

GitHub: https://t.co/JBGOntt6Ot

[Creative Graphic Design Lab @cgd_lab]

https://t.co/q3uPwSWo4w

[Creative Graphic Design Lab @cgd_lab]

🔥 Want LongCLIP in a familiar API? Now you can use LongCLIP with a Transformers-compatible API — load models like any other 🤗 Transformers model and run long-text vision-language inference with the `.from_pretrained` function.

← After & Before →

🚀 Link in the next tweet 👇

[Creative Graphic Design Lab @cgd_lab]

🔗 https://t.co/fbmrcOQNtc

🔗 https://t.co/r8RkqDwH15

[Creative Graphic Design Lab @cgd_lab]

🎉 Happy New Year! Start 2026 with smarter layout evaluation 🎨✨

Check out the Layout Evaluation Metrics for 🤗 Hugging Face Evaluate – tools to benchmark and score layout generation with ease. 🚀

link in the next post! 👇

[Creative Graphic Design Lab @cgd_lab]

🚀 Just dropped: gpt-graphic-design-evaluator — a LangChain-powered system to automatically evaluate graphic design via LLMs!

📐 Based on the paper “Can GPTs Evaluate Graphic Design Based on Design Principles?” https://t.co/vO8r5Hx7yS

💡 Check it out: https://t.co/L6xMEFVmrU

[Creative Graphic Design Lab @cgd_lab]

💭 I’ve seen some researchers who have been working on creative graphic design research recently move to Canva 🧐

[Creative Graphic Design Lab @cgd_lab]

> F Lite is a 10B parameter diffusion model created by Freepik and Fal, trained exclusively on copyright-safe and SFW content. The model was trained on Freepik's internal dataset comprising approximately 80 million copyright-safe images
https://t.co/Q3Bh3q9MeI

[Creative Graphic Design Lab @cgd_lab]

🚀 Introducing lighting-diffusers: your template for diffusion model training with 🤗Diffusers & ⚡PyTorch Lightning!

With full LightningCLI support, load all experiment settings from YAML files—making workflows reproducible, transparent, and manageable.
https://t.co/RnErDBPv5R

[Creative Graphic Design Lab @cgd_lab]

KoalaAI/StockImages-CC0 · Datasets at Hugging Face https://t.co/JGjbLS9jBC

[Creative Graphic Design Lab @cgd_lab]

common-canvas/commoncatalog-cc-by · Datasets at Hugging Face https://t.co/LTsmiqo86p

[Creative Graphic Design Lab @cgd_lab]

nyuuzyou/publicdomainpictures · Datasets at Hugging Face https://t.co/1KldeWnDnL

[Creative Graphic Design Lab @cgd_lab]

Thrilled to share our published #SSII2024 technical map, showcasing the evolution of layout and poster generation research from 1990 to 2024! Excited to see how the field evolves with future directions in creative design😘
🗺️ https://t.co/MsYcvvYZsl
📊 https://t.co/1YLim9dRQi

[Creative Graphic Design Lab @cgd_lab]

📣 Our partner has released 𝙪𝙠𝙞𝙮𝙤-𝙚-𝙛𝙖𝙘𝙚-𝙗𝙡𝙞𝙥𝟮-𝙘𝙖𝙥𝙩𝙞𝙤𝙣𝙨 dataset, which is a Ukiyo-e dataset with captions added using BLIP2. We intend to use this dataset in LoRA tutorials and other such materials:
🤗 https://t.co/TDKqbGtW0O
🐙 https://t.co/JPxqV6TJs7

[Creative Graphic Design Lab @cgd_lab]

laion/relaion2B-en-research-safe · Datasets at Hugging Face https://t.co/PkNopYcdjK

[Creative Graphic Design Lab @cgd_lab]

laion/relaion2B-en-research · Datasets at Hugging Face https://t.co/vwaeaiEKrS