AI.Text.Completion.Ollama

This repository contains a C# Console application that demonstrates a modern integration of Artificial Intelligence using local Large Language Models (LLMs). The project leverages the latest .NET AI abstractions to communicate with Ollama and the Llama 3.2 model.

🚀 Overview

The goal of this project is to showcase how to implement a clean, decoupled AI client using standard Microsoft interfaces. By using Microsoft.Extensions.AI, the code remains flexible and ready for professional environments.

Tech Stack

• Language: C# / .NET
• AI Model: Llama 3.2 (via Ollama)
• Key Libraries:
  • Microsoft.Extensions.AI: Standardized abstractions for AI services.
  • OllamaSharp: The connector for Ollama integration.

---

🛠 Prerequisites

To run this project, you need to have Ollama running locally. You can choose one of the following two methods:

Option 1: Direct Installation (Recommended for Local Dev)

1. Download and install from ollama.com.
2. Open your terminal and download the model:

   ollama pull llama3.2
   

Option 2: Docker Desktop

1. Run the Ollama container:

   docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama

2. Pull the model into the container

   docker exec -it ollama ollama pull llama3.2
   

---

💻 Installation & Usage

1. Clone the repo:

   git clone [https://github.com/your-username/AI.Text.Completion.Ollama.git](https://github.com/your-username/AI.Text.Completion.Ollama.git)
   

2. Restore and Build: The project will automatically restore the Microsoft.Extensions.AI and OllamaSharp NuGet packages.

   dotnet build
   

3. Run the App:

   dotnet run
   

---

📝 Example Code

The implementation focuses on simplicity and readability, using the IChatClient interface

 using Microsoft.Extensions.AI;
 using OllamaSharp;
 
 // Initialize the client pointing to the local Ollama instance
 IChatClient client = new OllamaApiClient(new Uri("http://localhost:11434"), "llama3.2");
 
 // Send prompt to model
 string prompt = "What is Domain-Driven Design? Explain it in 20 words.";
 ChatResponse response = await client.GetResponseAsync(prompt);
 
 // Output result and metadata
 Console.WriteLine($"Model response -> {response}");
 Console.WriteLine($"Tokens used: In={response.Usage?.InputTokenCount}, Out={response.Usage?.OutputTokenCount}");