|
--- |
|
license: cc-by-nc-4.0 |
|
--- |
|
|
|
# Model Card for Tailor3D |
|
|
|
## π Overview |
|
|
|
This model card is for the **Tailor3D** project. You can see the more information below. |
|
- **Github Code**: [Link to Github](https://github.com/Qi-Zhangyang/Tailor3D/) |
|
- **Arxiv Paper**: [Link to Arxiv](https://arxiv.org/abs/2407.06191) |
|
- **Project Page**: [Link to Project](https://tailor3d-2024.github.io/) |
|
- **Youtube Video**: [Link to Youtube](https://youtu.be/z9w5CN8mUJQ/) |
|
- **Hugging Face Demo**: [Link to Hugging Face Demo](https://huggingface.co/spaces/alexzyqi/Tailor3D) |
|
|
|
## ποΈ Model Details |
|
|
|
- Pretrained Models |
|
|
|
| Model | Pretained Model | Training Data | |
|
| :---: | :---: | :---: | |
|
| [tailor3d-small-1.0](https://huggingface.co/alexzyqi/tailor3d-small-1.0) | [openlrm-mix-small-1.1](https://huggingface.co/zxhezexin/openlrm-mix-small-1.1) | Objaverse + MVImgNet | |
|
| [tailor3d-base-1.0](https://huggingface.co/alexzyqi/tailor3d-base-1.0) | [openlrm-mix-base-1.1](https://huggingface.co/zxhezexin/openlrm-mix-base-1.1) | Objaverse + MVImgNet | |
|
| [tailor3d-large-1.0](https://huggingface.co/alexzyqi/tailor3d-large-1.0)| [openlrm-mix-large-1.1](https://huggingface.co/zxhezexin/openlrm-mix-large-1.1) | Objaverse + MVImgNet | |
|
|
|
- Model architecture (same as [OpenLRM](https://huggingface.co/zxhezexin/openlrm-mix-large-1.1)). |
|
|
|
| Type | Layers | Feat. Dim | Attn. Heads | Triplane Dim. | Input Res. | Image Encoder | |
|
| :---: | :----: | :-------: | :---------: | :-----------: | :--------: | :---------------: | |
|
| small | 12 | 512 | 8 | 32 | 224 | dinov2_vits14_reg | |
|
| base | 12 | 768 | 12 | 48 | 336 | dinov2_vitb14_reg | |
|
| large | 16 | 1024 | 16 | 80 | 448 | dinov2_vitb14_reg | |
|
|
|
- Training settings (same as [OpenLRM](https://huggingface.co/zxhezexin/openlrm-mix-large-1.1)). |
|
|
|
| Type | Rend. Res. | Rend. Patch | Ray Samples | |
|
| :---: | :--------: | :---------: | :---------: | |
|
| small | 192 | 64 | 96 | |
|
| base | 288 | 96 | 96 | |
|
| large | 384 | 128 | 128 | |
|
|
|
## π OpenLRM as the Pretained Model |
|
This repo, we use the LRM as the pretrained model. These are the link of its [arxiv paper](https://arxiv.org/abs/2311.04400) and [open-source implementation](https://github.com/3DTopia/OpenLRM). |
|
|
|
All regulations are the same as the OpenLRM. |
|
|
|
## π License |
|
|
|
- The model weights are released under the Creative Commons Attribution-NonCommercial 4.0 International License. |
|
- They are provided for research purposes only, and CANNOT be used commercially. |
|
|