|
Download README.md from 32V/MUNI: direct link, hf CLI and curl.
- Browser
- Download file 623 Bytes
-
https://huggingface.co/32V/MUNI/resolve/main/README.md
- Command line
-
hf download hf://32V/MUNI/README.md
-
curl -L -o README.md https://huggingface.co/32V/MUNI/resolve/main/README.md
623 Bytes
| library_name: pytorch | |
| license: gemma | |
| # MUNI | |
| MUNI-PoE checkpoint and FID/FAD reference statistics for image-text-audio generation. | |
| Use with the [official MUNI code](https://github.com/KAIST-Visual-AI-Group/MUNI). | |
| - `muni_poe.pth`: MUNI-PoE EMA weights for inference and evaluation, not training resume. | |
| - `stats/`: FID/FAD references for an MS-COCO 2014 validation subset, the AudioCaps test split, and the VGGSound test split. | |
| ## Download | |
| ```bash | |
| hf download 32V/MUNI --local-dir ckpts/muni | |
| ``` | |
| Use `ckpts/muni/muni_poe.pth` for `--checkpoint` and set the benchmark's | |
| `reference_root` to `ckpts/muni/stats`. | |