Downloads · 30 days
0
yonosplat-vknn
katolikov/yonosplat-vknn
yonosplat-vknn is a machine learning model from katolikov. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for vknn. The card lists the license as mit.
The YoNoSplat feed-forward 3D-Gaussian-Splatting encoder (8 context views), exported to ONNX and compiled to a fp16 vknn .vxm for on-device Vulkan inference.
Access
Public
Updated Jul 8, 2026
Repo size
28.9 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.bin3.9 GB · 62%
At a glance
- Library
- vknn
- License
- mit
- Access
- Public
- Created
- Jun 28, 2026
- Updated
- Jul 8, 2026
- SHA
- 2d148c90
Model card
From the Hugging Face model README
YoNoSplat encoder — ONNX + vknn (.vxm)
The YoNoSplat feed-forward 3D-Gaussian-Splatting encoder (8 context views), exported to ONNX and
compiled to a fp16 vknn .vxm for on-device Vulkan inference.
| file | what |
|---|---|
yonosplat_encoder.onnx | encoder graph, opset 17 (weights external in weights.bin) |
weights.bin | fp32 weights for the ONNX (~3.6 GB) |
encoder8_fp16.vxm | compiled fp16 vknn model, ready to run (~2.9 GB) |
image8.npy, intr8.npy | sample real 8-frame input (RealEstate10K scene) |
*_gold.npy | onnxruntime goldens for the six outputs |
Input: image [1,8,3,224,224], intrinsics [1,8,3,3].
Output: means [1,401408,3], covariances, harmonics, opacities, rotations, scales
(401408 = 8 views × 224×224 Gaussians).
Use
pip install huggingface_hub
python benchmark/scripts/fetch_model.py --repo katolikov/yonosplat-vknn --out benchmark/models
# then run + validate on device with the benchmark tool (see vknn/benchmark/USAGE.md)
python benchmark/run.py run benchmark/configs/yonosplat.json
On a current flagship phone GPU (Vulkan, fp16) the compiled encoder matches an onnxruntime golden at cosine ≥ 0.9997 on all six outputs.
Provenance & license
- Architecture + checkpoint: YoNoSplat by Botao Ye et al. (code MIT; checkpoint
re10k_224x224_ctx2to32). The ONNX here is a faithful export (analytic 3×3 inverse, Newton–Schulz SVD, baked DINOv2 pos-embed) validated cos = 1.0 against the original. - The weights are trained on RealEstate10K, whose dataset terms are research / non-commercial — use accordingly.
- This export/compilation is provided under MIT, with attribution to the original authors.
- Library
- vknn
- License
- mit
- Created
- Jun 28, 2026
- Updated
- Jul 8, 2026