Image-to-3D
MLX
Safetensors
Trellis
English
apple-silicon
gaussian-splatting
quantized
8-bit precision
Instructions to use ucalyptus/trellis-image-large-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use ucalyptus/trellis-image-large-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download ucalyptus/trellis-image-large-mlx --local-dir trellis-image-large-mlx
- Trellis
How to use ucalyptus/trellis-image-large-mlx with Trellis:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Download gaussian_decoder/config.json from ucalyptus/trellis-image-large-mlx: direct link, hf CLI and curl.
- Browser
- Download file 576 Bytes
-
https://huggingface.co/ucalyptus/trellis-image-large-mlx/resolve/main/gaussian_decoder/config.json
- Command line
-
hf download hf://ucalyptus/trellis-image-large-mlx/gaussian_decoder/config.json
-
curl -L -o config.json https://huggingface.co/ucalyptus/trellis-image-large-mlx/resolve/main/gaussian_decoder/config.json
576 Bytes
| { | |
| "_class_name": "TrellisGaussianDecoderConfig", | |
| "attn_mode": "swin", | |
| "features_lr": 1.0, | |
| "filter_kernel_size_3d": 0.0009, | |
| "latent_channels": 8, | |
| "mlp_ratio": 4, | |
| "model_channels": 768, | |
| "num_blocks": 12, | |
| "num_gaussians": 32, | |
| "num_heads": 12, | |
| "opacity_bias": 0.1, | |
| "opacity_lr": 1.0, | |
| "pe_mode": "ape", | |
| "perturb_offset": true, | |
| "qk_rms_norm": false, | |
| "resolution": 64, | |
| "rotation_lr": 0.1, | |
| "scaling_activation": "softplus", | |
| "scaling_bias": 0.004, | |
| "scaling_lr": 1.0, | |
| "use_fp16": true, | |
| "voxel_size": 1.5, | |
| "window_size": 8, | |
| "xyz_lr": 1.0 | |
| } |