support

TensorFlow.NET: Machine Learning in C#

Name: TensorFlow.NET: Machine Learning in C#
Author: SciSharp

Develop ML models in C# and F# with TensorFlow.NET, for founders building AI-driven apps, backed by 3.4k+ GitHub stars

3,393 stars541 forksC#Health Score 9/10Updated 1/22/2025100% free · open source

What it does

TensorFlow.NET allows developers to build and run machine learning models in C# and F# using the TensorFlow framework

Install / run

Install-Package TensorFlow.NET -Version 2.4.0

When to use it

•Building AI-driven Windows desktop or mobile apps
•Developing machine learning models for .NET-based web applications
•Integrating TensorFlow models into existing C# or F# projects

Quick start

1Create a new C# console application using `dotnet new console`
2Install the TensorFlow.NET NuGet package using `Install-Package TensorFlow.NET`
3Import the TensorFlow.NET namespace with `using TensorFlow;`
4Load a pre-trained TensorFlow model using `TF.Graph.Import`
5Use the `TF.Session` class to run the model and get predictions

Ready-to-paste prompt

TF.Graph.Import("path/to/model.pb"); var session = new TF.Session(); var input = TF Placeholder("input:0"); var output = session.Run(new[] { input }, new[] { "output:0" });

Heads up: Make sure to install the Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017, or 2019, as it is a prerequisite for running TensorFlow.NET

Saves to your device