GeorgeS2019
GeorgeS2019
@zhongkaifu thnx for the valuable tip :-)
## method [CreateSubGraph(string name)](https://github.com/zhongkaifu/Seq2SeqSharp/blob/86602da16c8b660e50ac85a23e9ab6bdfc59a706/Seq2SeqSharp/Tools/ComputeGraphTensor.cs#L104) ```cs public IComputeGraph CreateSubGraph(string name) { ComputeGraphTensor subGraph = new ComputeGraphTensor(m_weightTensorFactory, m_deviceId, m_needsBackprop, m_backprop, isSubGraph: true); if (m_visNeuralNetwork) { // Create parameters for neural network visualization...
@zhongkaifu abstractive summarization tasks sounds exciting !
@zhongkaifu please FYI [a potential reusable ONNX viewer](https://github.com/fel88/Dendrite/issues/1) that may be relevant for importing ONNX models
I think rust is a suitable option to create the .NET interop. I wonder what others think?
[How we integrate Rust with C#](https://blog.datalust.co/rust-at-datalust-how-we-integrate-rust-with-csharp/)
No, speed is the aim. Not WebAPI.
@sergey-tihon ## IKVM-8.2.0 now supports .NET Core/.NET 5+ https://github.com/sergey-tihon/Stanford.NLP.NET/discussions/119 https://www.reddit.com/r/dotnet/comments/vqqpqb/ikvm_820/ 
@agonzalezm https://github.com/ikvm-revived/ikvm-maven/issues/8 This provide instruction to generate .NET Core IKVM dll based on maven repository ``` edu.stanford.nlp stanford-corenlp 4.4.0 ``` https://mvnrepository.com/artifact/edu.stanford.nlp
@atrauzzi Please close this issue :-)