Closed luisquintanilla closed 7 months ago
In order for ML.NET to support this, TorchSharp and/or ONNX would have to support it. Currently, ONNX does through the Execution Providers.
However, they don't publish a NuGet similar to DirectML and OpenVINO.
In order to get this working in ML.NET and even then, this would only work for inference scenarios using ONNX Execution Providers, you'd need to:
Closing this issue for now.
Issue originally reported in Model Builder repo by @Selbyl
dotnet/machinelearning-modelbuilder#2764