| """Agate by LogoLabs: a 0.26B text-to-image model at 512 px (multi-resolution thinker-steered convolutional | |
| flow model + fine-tuned Ettin-68M text encoder + SD-VAE decoder). See AgatePipeline.""" | |
| from .marking import detect_watermark, provenance, save | |
| from .pipeline import AgatePipeline | |
| __all__ = ["AgatePipeline", "detect_watermark", "provenance", "save"] | |