Skip to content

Latest commit

 

History

History
25 lines (14 loc) · 1.44 KB

File metadata and controls

25 lines (14 loc) · 1.44 KB

Streaming output

Streams a generated article to the console as updates arrive, using the TXTextControl.AI streaming API.

Prerequisites

Install the .NET 10 SDK and .NET 8 runtime. Windows/Linux/macOS require a compatible platform-specific llama-server runtime and an application-owned GGUF chat model. See runtime setup and the license.

No MCP or Document Editor is required to execute this focused API example. Use within solutions under the repository's TX Text Control license terms.

Run

From the repository root, after provisioning the runtime:

dotnet run --project samples/LocalStreaming -- "/path/to/chat-model.gguf"

An article appears incrementally. Update the prompt in Program.cs to experiment.

Configuration and troubleshooting

The single argument is the GGUF path. Use an absolute path, especially when running from another directory. Set TXTEXTCONTROL_AI_LLAMA_SERVER to an installed executable or install a managed runtime explicitly using the package API described in the runtime guide. Startup does not download missing runtimes by default.

Model files are not included. Missing-runtime, unsupported-model and insufficient-memory errors are separate from NuGet restore errors. Try a smaller model/context if memory is limited. See the complete AI API guide for LocalModelOptions, cancellation, disposal and session behavior.