---
license: mit
base_model: VAST-AI/TripoSG
pipeline_tag: image-to-3d
tags:
- diffusers-3d
- triposg
- 3d
- image-to-3d
---

# TripoSG for diffusers-3d

[VAST-AI/TripoSG](https://huggingface.co/VAST-AI/TripoSG) converted into a [diffusers-3d](https://github.com/suvadityamuk/diffusers-3d)
pipeline: ordinary Diffusers and Transformers component folders plus the `object3d_model_index.json` sidecar that
the package's auto-loader validates before downloading anything. Nothing here requires remote code.

## Install

```bash
pip install git+https://github.com/suvadityamuk/diffusers-3d.git
pip install "git+https://github.com/suvadityamuk/diffusers-3d.git#subdirectory=packages/diffusers-3d"
```

`diffusers-3d` runs every network in plain PyTorch on CPU or GPU. Rendering Gaussian splats needs the optional
`gsplat` backend; the PBR-textured mesh for TRELLIS.2 needs CuMesh (see the package docs).

## Usage

```python
import torch
from diffusers_3d import AutoPipelineForImageTo3D, ImageCondition

pipeline = AutoPipelineForImageTo3D.from_pretrained("suvadityamuk/TripoSG-diffusers-3d").to("cuda")
mesh = pipeline(ImageCondition(image=rgba), generator=torch.Generator("cuda").manual_seed(0)).objects[0]
```

`rgba` is a `(4, H, W)` tensor in `[0, 1]` whose alpha channel masks the object; the pipeline crops and pads it as the released code does. The result is an untextured `MeshAsset` extracted with marching cubes.

## Components

| Folder | Class |
|---|---|
| `feature_extractor` | `transformers.BitImageProcessorPil` |
| `image_encoder` | `transformers.Dinov2Model` |
| `scheduler` | `diffusers.FlowMatchEulerDiscreteScheduler` |
| `transformer` | `diffusers_3d.families.triposg.models.TripoSGTransformerModel` |
| `vae` | `diffusers_3d.families.triposg.vae.TripoSGVAEDecoder` |

## Provenance

Converted with `diffusers-3d-convert-triposg` from `diffusers-3d 0.1.0.dev0`; the package's family README lists the pinned upstream
revisions and how the conversion maps each tensor. Parity tests against the pinned upstream code are part of the
package test suite.

## License and attribution

TripoSG weights and code: MIT License, Copyright (c) VAST-AI-Research. The DINOv2 image encoder
weights are Apache-2.0, Copyright (c) Meta Platforms, Inc. TripoSG's transformer is derived from Tencent HunyuanDiT,
whose Tencent Hunyuan Community License does not apply in the European Union, the United Kingdom, or South Korea;
you are responsible for making sure your use complies with it. See `LICENSE-MIT` and `NOTICE`. Not affiliated with
or endorsed by VAST, Tencent, or Meta.