sd_download_model: Download a Stable Diffusion model from Kaggle Models
In sd2R: Stable Diffusion Image Generation

sd_download_model

R Documentation

Download a Stable Diffusion model from Kaggle Models

Description

Downloads a model bundle from the public Kaggle Models registry and unpacks it into dest. Mirrors the behaviour of the Python kagglehub package (kagglehub.model_download("owner/model/framework/variation")) but uses only base R – no Python dependency.

Usage

sd_download_model(
  handle = "lbsbmsu/flux-2/gguf/default",
  dest,
  version = NULL,
  files = NULL,
  verbose = FALSE
)

Arguments

`handle`	Model handle in kagglehub form `"owner/model/framework/variation"`. Defaults to `"lbsbmsu/flux-2/gguf/default"` – a ready-to-use FLUX 2 (GGUF) model, so newcomers can call `sd_download_model(dest = "models/flux2")`.
`dest`	Destination directory for the unpacked files. Created if it does not exist. Required.
`version`	Integer version number. If `NULL` (default) the latest version is resolved automatically from Kaggle.
`files`	Optional character vector of file names to extract from the bundle. If `NULL` (default) all files are extracted.
`verbose`	Logical; print progress messages. Defaults to `FALSE`.

Details

Kaggle serves each model version as a single .tar.gz bundle; the whole bundle is downloaded even when only some files are needed. Only public models are supported.