Downloads · 30 days
0
LocalMuseAI/coreml-realism-by-stable-yogi-sd15-v9-lcm-6bit
coreml-realism-by-stable-yogi-sd15-v9-lcm-6bit is a text-to-image model from LocalMuseAI. Use it when you need an image from a text prompt. It is set up for ml-stable-diffusion. The card lists the license as openrail++.
This repository contains a reproducible Core ML format conversion for the LocalMuse iOS app. The original model was created by Stable Yogi. LocalMuseAI does not claim authorship of the source weights.
Downloads · 30 days
0
Access
Public
Updated Aug 31, 2026
Repo size
1.1 GB
Likes
0
Public
Click a slice to open those files.
.bin1.1 GB · 100%
From the Hugging Face model README
This repository contains a reproducible Core ML format conversion for the LocalMuse iOS app. The original model was created by Stable Yogi. LocalMuseAI does not claim authorship of the source weights.
SD1.5_V9_LCM / Civitai version 1235013realismByStableYogi_sd15V9LCM.safetensors2,132,626,090 bytes1a16a9524427019dc513b398087035e7e36ad2f2aa9691bd71d28ad959e94180The source is an already-distilled SD 1.5 LCM checkpoint. No additional LoRA or merge was applied during conversion.
SPLIT_EINSUM_V2Unet.mlmodelc: Apple 6-bit k-means palettizationTextEncoder.mlmodelc: FP16VAEDecoder.mlmodelc: FP16VAEEncoder.mlmodelc: FP16ml-stable-diffusion revision
e12202c1f6405b83918b58a5d097cd61e3e1f702All source files are authenticated before tracing. Component-level
PyTorch/Core ML parity checks are enabled during conversion. Exact source and
artifact hashes are recorded in PROVENANCE.json.
The upstream gallery examples use Euler a at 9 steps and CFG 1. LocalMuse uses its existing LCM scheduler because this is an LCM-distilled checkpoint and the same configuration was validated before conversion. Upstream textual-inversion trigger names are separate assets and are not included in this repository.
The model and this converted representation remain subject to the CreativeML
Open RAIL++-M license in LICENSE, including its use-based restrictions.
Modified files are identified here as a Core ML format conversion with 6-bit
UNet palettization. No additional restrictions are imposed.
ml-stable-diffusion