qwen-image
Qwen-Image-Bench
Qwen-Image-Bench
A creator-centric benchmark for evaluating Text-to-Image models beyond semantic alignment.
Links
Resource
Link
📑 Paper
http://arxiv.org/abs/2605.28091
📊 Benchmark Dataset (HuggingFace)
https://huggingface.co/datasets/Qwen/Qwen-Image-Bench
📊 Benchmark Dataset (ModelScope)
https://www.modelscope.cn/datasets/Qwen/Qwen-Image-Bench
💻 GitHub
https://github.com/QwenLM/Qwen-Image-Bench
🧑⚖️ Q-Judger Model… See the full description on the dataset page: https://huggingface.co/datasets/Qwen/Qwen-Image-Bench.Qwen-Image-2512_samplesThis dataset is a highly diverse set of high quality images generated with Qwen Image 2512.
Possible uses
Regularization images for training models based on Qwen Image 2512
Quality testing
Data source
The images were created in ComfyUI with the
bf16 version
of Qwen Image 2512. For each prompt were four images generated, all are (without any cherry picking) included in the corresponding dataset directories.
bf16 - full model weights
1328x1328 pixels - native resolution… See the full description on the dataset page: https://huggingface.co/datasets/stablellama/Qwen-Image-2512_samples.ImageNet1K-T2I-QwenVL-QwenImageQwen-Image-2.1-rewriter-distill
Qwen-Image-2.1-rewriter-distill
A distillation corpus for the Qwen-Image-2.1 text-to-image prompt rewriter.
Qwen/Qwen-Image-2.1-PE-T2I is a 9B
thinking model that turns a short image request in any language into one long English
paragraph plus an aspect ratio. It needs a ~1,700-word system prompt, thinking mode, and
up to 16,256 new tokens to emit a single JSON object. This dataset records what it produced
for 8,797 short requests, so that much smaller students can be trained to… See the full description on the dataset page: https://huggingface.co/datasets/ML-Intern-lab/Qwen-Image-2.1-rewriter-distill.qwen-image21-t4-samples
Qwen-Image-2.1 INT8 samples from 2xT4
Images generated on Kaggle's free 2xTesla T4 (sm_75) with the Comfy-Org INT8 ConvRot
checkpoints at revision ace0edeb, ComfyUI c194dd00, Comfy Kitchen 0.2.35.
1024x1024, 40 steps, CFG 1, euler/simple, Comfy Kitchen INT8 attention.
Hosted to illustrate this discussion:
https://huggingface.co/Comfy-Org/Qwen-Image-2.1/discussions/8
Prompt for all three: "A capybara wearing a wizard hat, reading a book by candlelight,
detailed oil painting"… See the full description on the dataset page: https://huggingface.co/datasets/kowappa/qwen-image21-t4-samples.imagenet-latents-qwen-image-vae