[Diffusion] Preserve non-DiT loading with DLO plans

Signed-off-by: Hongsheng Liu <liuhongsheng4@huawei.com>
This commit is contained in:
Hongsheng Liu committed 2026-08-15 23:57:03 +08:00
1 parent 11687f73bd
commit fa2f5b4571
6 files changed
+155 -6

No files matched your search

@@ -63,6 +63,14 @@ files, repeat the capability decision, or reconstruct names from its block
topology. If preflight fails, the loader materializes weights normally and DLO
consumes those runtime tensors.
The plan owns only dedicated DiT component sources. If a pipeline also exposes
ordinary sources for a text encoder or another non-DiT component, the loader
still consumes those sources and includes their loaded names in its strict
coverage check. Only the source prefixes covered by the plan skip ordinary
materialization. A source that mixes DiT and non-DiT weights fails closed to
the complete ordinary-loader path because it cannot be skipped safely as a
unit.
This boundary keeps checkpoint semantics out of DLO and avoids model-pipeline
flags such as `_supports_mmap_loading` or parameter attributes for mmap-only
transforms. Model-specific direct-layout knowledge, when required, lives in a
@@ -72,6 +72,12 @@ coverage, shape, dtype, topology, or loader-callback compatibility cannot be
proven, it runs the ordinary model loader instead. DLO consumes that result and
does not make a second checkpoint-compatibility decision.
The mmap plan skips only dedicated DiT weight sources. Other component sources,
such as a text encoder loaded through the shared diffusion loader, continue to
use their ordinary component loader. A checkpoint source that mixes DiT and
non-DiT weights falls back completely rather than leaving an unplanned
component uninitialized.
With direct checkpoint mmap, the loader:
1. saves non-persistent buffers such as RoPE frequencies;