Is it possible to get the FSRCNNX models in tensorflow/pytorch/onnx format for experimenting?

[kristoftunner]

Expected behavior 预期的功能

FSRCNNX models are uploaded as an artifact or directly to the source repo in order to experiment with it instead of only storing it in the shader file.

Alternative behavior (optional) 近似的功能（可选）

No response

[hooke007]

You must misunderstood sth. https://github.com/yifanw90/FSRCNN-TensorFlow
This repo just port its fork version.

GitHub

GitHub - yifanw90/FSRCNN-TensorFlow: An implementation of the fast super-resolution convolutional neural network in TensorFlow

An implementation of the fast super-resolution convolutional neural network in TensorFlow - GitHub - yifanw90/FSRCNN-TensorFlow: An implementation of the fast super-resolution convolutional neural ...

[Blinue]

The statement 'stored in the shader file' is not entirely accurate; we need to port FSRCNNX into a format understandable by Magpie, namely HLSL. There's a plan in place to support ONNX, but other formats won't be accommodated.

[kristoftunner]

onnx format would be super heplful!