[Perf][Diffusion] Optimize HSDP startup via Rank-0 shared weight loading and accelerated LoRA delta computation (#7005)

Signed-off-by: samithuang <285365963@qq.com>
This commit is contained in:
Samit authored and GitHub committed 2026-09-11 08:28:42 +00:00
1 parent bad50980c9
commit 96360ccb00
13 files changed
+637 -16

No files matched your search

+1
View File
@@ -572,6 +572,7 @@ class OrchestratorArgs:
vae_use_slicing: bool = False
vae_use_tiling: bool = False
enable_multithread_weight_load: bool = True
enable_broadcast_weight_load: bool = False
num_weight_load_threads: int = 4
diffusion_offload_config: dict[str, Any] | None = None
# Compatibility aliases for existing callers and model-specific stage