Downloads · 30 days
0
dtaddis/localimagegen-sdxs-mobile
localimagegen-sdxs-mobile is a text-to-image model from dtaddis. Use it when you need an image from a text prompt. It is set up for unity-inference-engine. The card lists the license as other.
Mixed-precision Unity Sentis conversion of SDXS DreamShaper for LocalImageGen: UINT8 text encoder and UNet with an FP16 TAESD decoder.
Downloads · 30 days
0
Access
Public
Updated Jul 26, 2026
Repo size
442 MB
Likes
0
Public
Click a slice to open those files.
.sentis442 MB · 100%
From the Hugging Face model README
Mixed-precision Unity Sentis conversion of SDXS DreamShaper for LocalImageGen: UINT8 text encoder and UNet with an FP16 TAESD decoder.
This is a transformed inference artifact, not a newly trained model. All training credit belongs to the upstream authors.
These files are ready for the LocalImageGen Unity asset. Use its model manager to download them; do not rename files or folders.
Converted with Unity 6000.5.5f1 and Unity Inference Engine 2.6.1. The customer-facing asset does not contain model weights.
76f720262bb051da75666b22c902a78c8e16c763This is the experimental mobile tier. It uses staged worker loading to reduce peak memory, but it is not guaranteed to run on every phone or tablet. Treat 4 GB of system RAM as a practical minimum, prefer recent devices, and test the exact target hardware before shipping. The validated range is 384x384 to 512x512; this distilled model requires exactly one denoising step and guidance 0.
The SDXS checkpoint is distributed under OpenRAIL++, while the DreamShaper teacher/model lineage also carries CreativeML OpenRAIL-M terms. Both license texts are included.
CreativeML OpenRAIL-M and OpenRAIL++. See LICENSE.md. Downstream users are responsible for
reviewing and complying with the model license and any applicable laws.