Victor

@pkese Generally speaking, I am also very interested. I am also a maintainer of Python.NET. I would like to see the necessary bits implemented on PyTorch side though: https://discuss.pytorch.org/t/mixing-c-and-python-types/171129 Unfortunately,...

@pkese unsure. The storage mechanism in PyTorch seems multilayered and spans both Python and C++ side. Unless one can come up with a concise Python interface that would need to...

> As long as we don't introduce a Python VM dependency into TorchSharp... Yeah, my feed is full of links to LeCun's recent comment on Python slowing deep learning research...

Moving this to 2.x for now. Current runtime is unable to survive reload of its assemblies. Unofficially, you should be able to use TensorFlow in Unity by adding the following...

Yes, it is! The coming preview will bring 1.15, followed by a 1.15-based release shortly. After that previews will switch to 2.0 Can't promise any specific dates here, but you...

We have released [Preview 1](https://www.nuget.org/packages/LostTech.TensorFlow/) of TensorFlow 2.4 support

The model should be saved to the current directory, e.g. [Environment.CurrentDirectory](https://docs.microsoft.com/en-us/dotnet/api/system.environment.currentdirectory?view=netframework-4.8)

@Cliff-Foreman you need to reproduce model structure first. E.g. pass all layers as you did for training. If you want layers to be restored too, you need to refer to...

@solarflarefx to see if I can help you, I need to know how exactly is data currently stored in GPU? Is it a texture in DirectX/OpenGL? A CUDA buffer? This...

This is still not supported by TensorFlow directly: https://github.com/tensorflow/tensorflow/issues/29039