README.md

---
license: apache-2.0
tags:
- image-super-resolution
- image-to-image
- mlx
- apple-silicon
- super-resolution
- neural-heat-fields
base_model:
- prs-eth/thera-rdn-air
- prs-eth/thera-rdn-pro
library_name: mlx
language:
- en
pipeline_tag: image-to-image
---

<div align="center">

# ✨ thera-mlx

**Aliasing-Free Arbitrary-Scale Super-Resolution for Apple Silicon**

*An [MLX](https://github.com/ml-explore/mlx) port of [Thera](https://therasr.github.io) — running natively on M1/M2/M3/M4*

[![GitHub](https://img.shields.io/badge/GitHub-madeleinelmuller%2Fthera--mlx-black?logo=github)](https://github.com/madeleinelmuller/thera-mlx)
[![License](https://img.shields.io/badge/License-Apache%202.0-blue)](https://github.com/madeleinelmuller/thera-mlx/blob/main/LICENSE)
[![Paper](https://img.shields.io/badge/Paper-arXiv%3A2311.17643-red)](https://arxiv.org/abs/2311.17643)

**128×128 → 512×512 in ~1.5 seconds on M1** · Arbitrary scale factors · Web UI + CLI · Video support

</div>

---

## Model Description

**thera-mlx** contains MLX-format weights for the [Thera](https://therasr.github.io) super-resolution model, converted from the original JAX/Flax checkpoints released by [PRS-ETH](https://prs.igp.ethz.ch/) (ETH Zurich).

Thera uses **neural heat fields** to perform aliasing-free image super-resolution at *any* scale factor — not just 2× or 4×. This repo provides two variants:

| File | Model | Speed | Quality |
|------|-------|-------|---------|
| `weights-air.safetensors` | Air | ~1.5s for 128→512px on M1 | Great |
| `weights-pro.safetensors` | Pro (+ SwinIR refinement) | ~5s for 128→512px on M1 | Best |

## Usage

```bash
git clone https://github.com/madeleinelmuller/thera-mlx
cd thera-mlx
pip install -r requirements.txt

# Download weights (pulls from this repo automatically)
python run.py convert --model air
python run.py convert --model pro

# Launch web UI
python run.py

# Or use CLI
python run.py run input.png output.png --scale 4 --model air
```

See the full [GitHub repository](https://github.com/madeleinelmuller/thera-mlx) for complete documentation.

## Original Weights

The weights in this repository are converted from:
- [`prs-eth/thera-rdn-air`](https://huggingface.co/prs-eth/thera-rdn-air)
- [`prs-eth/thera-rdn-pro`](https://huggingface.co/prs-eth/thera-rdn-pro)

## Citation

If you use this in your work, please cite the original Thera paper:

```bibtex
@article{becker2023thera,
  title   = {Thera: Aliasing-Free Arbitrary-Scale Super-Resolution with Neural Heat Fields},
  author  = {Becker, Alexander and Caye Daudt, Rodrigo and Narnhofer, Dominik and
             Peters, Torben and Metzger, Nando and Wegner, Jan Dirk and Schindler, Konrad},
  journal = {arXiv preprint arXiv:2311.17643},
  year    = {2023}
}
```

**Original resources:**
- Paper: [arxiv.org/abs/2311.17643](https://arxiv.org/abs/2311.17643)
- Project page: [therasr.github.io](https://therasr.github.io)
- Original HF weights: [huggingface.co/prs-eth](https://huggingface.co/prs-eth)