README.md

---
title: OppaiOracle
emoji: ⚡
colorFrom: pink
colorTo: purple
sdk: gradio
sdk_version: 6.14.0
python_version: "3.12"
app_file: app.py
pinned: false
license: apache-2.0
short_description: Multi-label anime tagger — ViT, 19,294 tags
models:
- Grio43/OppaiOracle
tags:
- anime
- anime-tagger
- tagger
- image-tagging
- multi-label
- multi-label-classification
- vision-transformer
- vit
- illustration
- danbooru
---

# OppaiOracle — live demo

Multi-label tagger for anime / illustration images. ViT trained from scratch on a \~5.9M-image cleaned corpus, predicting over 19,000 tags.

- **Model card:** [Grio43/OppaiOracle](https://huggingface.co/Grio43/OppaiOracle)
- **License:** Apache 2.0
- **Checkpoint served here:** V1.1 (448×448 fine-tune of V1)
- **Native input resolution:** 448×448 (letterboxed)

## How to use

1. Drop or upload an anime / illustration image.
2. The Space returns the predicted tags with their probabilities. Per-tag thresholds from `pr_thresholds.json` are applied where available.
3. Treat the output as a *first pass* — see the [model card](https://huggingface.co/Grio43/OppaiOracle) for the limitations, especially around color, hair-length, size-bucket, and small-accessory tags.

## Notes

- This Space runs the **V1.1 ONNX** checkpoint via ONNX Runtime on the assigned hardware (ZeroGPU when available, CPU fallback otherwise).
- For local inference (ONNX Runtime, DirectML, CUDA EP), grab `V1.1_onnx/model.onnx` from the model repo (or `V1_onnx/model.onnx` if you want the 320×320 sibling).
- For PyTorch / custom inference, grab `V1.1_safetensors/model.safetensors` plus `config.json` and `preprocessing.json`.