cafe_aesthetic

Model overview

cafe_aesthetic is published by cafeai as an image-classification model. The captured configuration identifies BeitForImageClassification with a beit model type. According to the model card, it is an image classifier fine-tuned on microsoft/beit-base-patch16-384 for dataset conditioning in the Waifu Diffusion project.

Recorded capabilities

Use cases in the source record

• Dataset-conditioning filtering that classifies images as not_aesthetic to remove them before Stable Diffusion fine-tuning, as described in the model card.
• Cautious inclusion workflows where the classifier tends to keep images unless they are manga-like, sketchy, or text-obscured, according to the model card.

Limitations and unknowns

• According to the model card, the classifier is not perfect and misclassifications are expected; large-dataset averaging is the publisher's stated mitigation.
• No parameter count was extracted from this record.
• No evaluation results were extracted from this record.
• Provider state is historical snapshot data, not independently refreshed current availability.

Source and provenance

Source: cafeai/cafe_aesthetic

Captured: Unknown. Processed: 2026-09-07T19:34:42.277077+00:00.

Info Since people are downloading this and I don't know why, I'll add some information. This model is an image classifier fine-tuned on microsoft/beit-base-patch16-384 . Its purpose is to be used in the dataset conditioning step for the Waifu Diffusion project , a fine-tune effort for Stable Diffusion. As WD1.4 is planned to have a significantly large dataset (~15m images), it is infeasible to analyze every image manually to determine whether or not it should be included in the final training dataset. This image classifier is trained on approximately 3.5k real-life and anime/manga images. Its purpose is to remove aesthetically worth…