sd_load_model: Load a registered model

In sd2R: Stable Diffusion Image Generation

Load a registered model

Description

Loads a model by its registry id. Returns a cached context if already loaded, otherwise creates a new sd_ctx. Additional arguments override registry defaults.

Usage

sd_load_model(id, ...)

Arguments

id	Model identifier from registry
...	Additional arguments passed to sd_ctx, overriding registry defaults (e.g. vae_decode_only = FALSE)

Details

Before loading, the estimated VRAM need (on-disk weight size times a headroom factor plus a reserve) is compared against free GPU memory; if it would not fit, least-recently-used models are unloaded first. If loading still fails due to insufficient VRAM, the LRU model is unloaded and the load is retried once. VRAM estimation/eviction is skipped when GPU memory cannot be queried (e.g. CPU backend). Tunable via environment variables SD2R_VRAM_HEADROOM (default 1.2) and SD2R_VRAM_RESERVE_MB (default 512).

Value

SD context (external pointer)

Examples

## Not run: 
ctx <- sd_load_model("flux-dev")
imgs <- sd_txt2img(ctx, "a cat in space")

# Override defaults
ctx <- sd_load_model("flux-dev", vae_decode_only = FALSE, verbose = TRUE)

## End(Not run)

sd2R documentation built on June 19, 2026, 9:08 a.m.