Clarify Python/onnx inference usage

[Ivan-267]

Clarified instructions for which Sync node mode to select when using Python based vs ONNX inference. Partially addresses edbeeching/godot_rl_agents_plugin#69.

[stefanfausser]

LGTM.

Maybe the two sentences with "(onnx inference uses C# instead)" could be changed to "(ONNX inference uses a C# library in Godot instead)" because ONNX is an abbreviation for Open Neural Network Exchange and to clarify where ONNX inference is running. The original webpage (https://onnx.ai/onnx/intro/) also writes ONNX in capital cases.

[Ivan-267]

LGTM.

Maybe the two sentences with "(onnx inference uses C# instead)" could be changed to "(ONNX inference uses a C# library in Godot instead)" because ONNX is an abbreviation for Open Neural Network Exchange and to clarify where ONNX inference is running. The original webpage (https://onnx.ai/onnx/intro/) also writes ONNX in capital cases.

Thanks, I agree with that and used ONNX in other places where applicable.

In the latest commit I changed it to:

Set Onnx Inference if you want to run inference in Godot without Python (this mode uses C# instead).

I clarified it's talking about this "mode" in this change as ONNX inference can be done using other languages as well.
Onnx in that part (I know you were not referring to this one) is because that's how the property name shows in the inspector.