Downloads · 30 days
0
anakhiu/realesrgan-onnx
realesrgan-onnx is a machine learning model from anakhiu. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for onnx. The card lists the license as bsd-3-clause.
ONNX export of Real-ESRGAN x4plus (RRDBNet, ~16.7M params) for 4x super-resolution, for lifting low-resolution images over a stock-site resolution floor. Runs on ONNX Runtime — DirectML on Windows GPUs, CPU elsewhere.
Downloads · 30 days
0
Access
Public
Updated Jul 27, 2026
Repo size
67.1 MB
Likes
2
Public
Click a slice to open those files.
.onnx67.1 MB · 100%
From the Hugging Face model README
ONNX export of Real-ESRGAN x4plus (RRDBNet, ~16.7M params) for 4x super-resolution, for lifting low-resolution images over a stock-site resolution floor. Runs on ONNX Runtime — DirectML on Windows GPUs, CPU elsewhere.
RealESRGAN_x4plus.pth
(xinntao/Real-ESRGAN, release v0.1.0).num_feat=64, num_block=23, num_grow_ch=32,
scale=4), defined inline and loaded from the params_ema state dict — no
basicsr dependency.torch.onnx.export, opset=17, dynamic height/width. No quantization
— full fp32.input: [1, 3, H, W] float32, RGB, values 0..1. H and W are
dynamic; the caller tiles large images itself.output: [1, 3, H*4, W*4] float32, RGB, values ~0..1 (clamp
before converting to 8-bit).No normalisation beyond /255 on the way in and *255 on the way out.
BSD-3-Clause, from xinntao/Real-ESRGAN.