|
Download README.md from afun-anonymous/AFUN: direct link, hf CLI and curl.
- Browser
- Download file 759 Bytes
-
https://huggingface.co/afun-anonymous/AFUN/resolve/main/README.md
- Command line
-
hf download hf://afun-anonymous/AFUN/README.md
-
curl -L -o README.md https://huggingface.co/afun-anonymous/AFUN/resolve/main/README.md
759 Bytes
| license: apache-2.0 | |
| tags: | |
| - robotics | |
| - affordance | |
| - manipulation | |
| - 3d-motion | |
| - segmentation | |
| # AFUN checkpoint | |
| `afun.pt` is the released AFUN checkpoint (Table 3 model). It loads with the inference code in the | |
| anonymous AFUN code repository; place it at `checkpoints/afun.pt` and run | |
| ```bash | |
| python demo.py --checkpoint checkpoints/afun.pt | |
| ``` | |
| Backbone weights (Qwen3-VL-8B, SAM3, Sonata) download from Hugging Face on first use. SAM3 is gated: | |
| request access at [facebook/sam3](https://huggingface.co/facebook/sam3) first. | |
| Companion datasets: [AFUN](https://huggingface.co/datasets/afun-anonymous/AFUN), | |
| [AFUN_eval](https://huggingface.co/datasets/afun-anonymous/AFUN_eval), | |
| [AFUN_pool](https://huggingface.co/datasets/afun-anonymous/AFUN_pool). | |