README.md

---
{}
---
# AM-RADIO: Reduce All Domains Into One

Mike Ranzinger, Greg Heinrich, Jan Kautz, Pavlo Molchanov

[NVIDIA Research](https://www.nvidia.com/en-us/research/)

\[[Paper](https://arxiv.org/abs/2312.06709)\]\[[BibTex](#citing-radio)\]

## Pretrained Models

Refer to `model_results.csv` for model versions and their metrics.

### HuggingFace Hub

In order to pull the model from HuggingFace, you need to be logged in:

```Bash
huggingface-cli login
```

Then you can pull the model from a Python script:

```Python
from transformers import AutoModel
model = AutoModel.from_pretrained("nvidia/RADIO", trust_remote_code=True)
```

Alternatively, you can specify an access token:

```Python
access_token = "<YOUR ACCESS TOKEN"
model = AutoModel.from_pretrained("nvidia/RADIO", trust_remote_code=True, token=access_token)
```

### Usage

RADIO will return a tuple with two tensors. The `summary` is similar to the `cls_token` in ViT and is meant to represent the general concept of the entire image. It has shape $(B,C)$ with $B$ being the batch dimension, and $C$ being some number of channels. The `spatial_features` represent more localized content which should be suitable for dense tasks such as semantic segmentation, or for integration into an LLM. It has shape $(B,T,D)$ with $T$ being the flattened spatial tokens, and $D$ being the channels for spatial features. Note that $C \neq D$ in general.

Converting to a spatial tensor format can be done using the downsampling size of the model, combined with the input tensor shape. For 'radio_v1', the patch size is 14.
```Python
from einops import rearrange
spatial_features = rearrange(spatial_features, 'b (h w) d -> b d h w', h=x.shape[-2] // patch_size, w=x.shape[-1] // patch_size)
```

The resulting tensor will have shape $(B,D,H,W)$, as is typically seen with computer vision models.

### RADIOv1 Notes

We have trained this model to be flexible in input dimension. It supports inputs with both width and height in the range $[14, 1008]$ as long as both axes are divisible by 14. We have found that summarization tokens work best at $H=W=378$ (although the range $[192, 448]$ works well). For spatial tasks, we used $H=W=518$ to perform linear probing for semantic segmentation, and may perform better for more high-resolution tasks. Going up to $1008$, the model may need additional fine tuning at that resolution for best results.

It is not required that $H=W$ although we have not specifically trained or testing the model in this setting.


## Training

_Coming Soon_

## License

RADIO code and weights are released under the [NSCLv1 License](LICENSE).

## Citing RADIO

If you find this repository useful, please consider giving a star and citation:
```
@misc{ranzinger2023amradio,
      title={AM-RADIO: Agglomerative Model -- Reduce All Domains Into One},
      author={Mike Ranzinger and Greg Heinrich and Jan Kautz and Pavlo Molchanov},
      year={2023},
      eprint={2312.06709},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}
```