|
Download README.md from jonathanst29/tinier-upscale-models: direct link, hf CLI and curl.
- Browser
- Download file 1.03 kB
-
https://huggingface.co/jonathanst29/tinier-upscale-models/resolve/main/README.md
- Command line
-
hf download hf://jonathanst29/tinier-upscale-models/README.md
-
curl -L -o README.md https://huggingface.co/jonathanst29/tinier-upscale-models/resolve/main/README.md
1.03 kB
| license: bsd-3-clause | |
| tags: | |
| - super-resolution | |
| - real-esrgan | |
| - onnx | |
| - image-upscaling | |
| library_name: onnx | |
| # Tinier upscale models | |
| ONNX exports of the official [Real-ESRGAN](https://github.com/xinntao/Real-ESRGAN) | |
| weights, used by the in-browser AI upscaler at [tinier.app/upscale-image](https://tinier.app/upscale-image) | |
| (onnxruntime-web, WebGPU/WASM). | |
| | File | Source weights | Params | Size | | |
| |---|---|---|---| | |
| | `realesr-general-x4v3.onnx` | realesr-general-x4v3.pth (SRVGGNetCompact) | 1.2M | 4.6 MB | | |
| | `realesrgan-x4plus.onnx` | RealESRGAN_x4plus.pth (RRDBNet) | 16.7M | 64 MB | | |
| Export details: fp32, opset 17, dynamic H/W axes. | |
| Contract: input `float32 [1, 3, h, w]`, RGB in `[0, 1]`, NCHW → output `float32 [1, 3, 4h, 4w]`. | |
| Both exports are numerically verified against the PyTorch originals | |
| (max abs diff ~1e-6). Conversion script: | |
| [`scripts/export-onnx-models.py`](https://github.com/jonathanst29/tinier) in the Tinier repo. | |
| License: BSD-3-Clause, inherited from the upstream Real-ESRGAN weights. | |