Skip to content

vllm_omni.diffusion.model_loader.host_weight_loader ¶

Host Weight Runtime integration for diffusion model loaders.

This mixin keeps HWR identity, restore, publication, and startup-recovery policy out of the ordinary diffusers loader. The host loader supplies the canonical source-preparation and weight-loading hooks; this module owns only the optional final-layout transaction.

logger `module-attribute` ¶

logger = init_logger(__name__)

HWRLoaderMixin ¶

Optional final-layout HWR behavior shared by diffusion loaders.

load_fresh_canonical_model ¶

load_fresh_canonical_model() -> Module

Reload a disposable startup model without HWR or checkpoint mmap.