Start the local RegCheck server — regcheck_start

Starts the bundled RegCheck server as a background process, either via Docker (recommended) or a manual Python virtual environment. The server runs at http://localhost:8000 and is used automatically by module_run(paper, "reg_check") (the default client = "ollama").

Usage

regcheck_start_local(method = c("docker", "python"), model = NULL, port = 8000)

Arguments

method: "docker" (default) or "python"
model: Ollama language model to use. Defaults to NULL, which auto-selects the largest language model currently pulled in Ollama. Any model you have pulled with ollama pull can be used here – larger models generally produce better judgements.
port: port to run the server on (default 8000)

Value

invisibly, the process object

Details

Docker path (method = "docker", recommended): Requires Docker to be installed. The first run builds the image (~5 minutes); subsequent starts are instant. No Python installation needed.

Python path (method = "python"): Requires Python 3.12+ and Ollama. Run regcheck_setup_local() once first.

In both cases, Ollama must be running with the required models pulled. The embedding model is fixed; the language model can be any Ollama model you have pulled – use the largest one that fits your hardware:

ollama pull nomic-embed-text-v2-moe
ollama pull llama3.2          # or llama3.1, mistral, etc.

The API token is set automatically to "metacheck-local" – no manual token configuration needed.

Examples

if (FALSE) { # \dontrun{
# needs Docker (or Python, after regcheck_setup_local()) and Ollama
regcheck_start_local(method = "docker")
regcheck_stop_local()
} # }