Lightricks' production recipe for Dev is Stage 1 + 2x upsample + Stage 2
refine. Until now the UI reset Refine to False on every model change, so
users picking Dev got a single-pass generation that does not match the
recommended flow. Default Refine to on for variants that support the
canonical Stage 2 recipe; multi-condition variants stay off.
Modules already inventoried at load time; repeating the six-line dump at
each upsample or refine boundary is redundant. silent=True suppresses
the per-module DEBUG lines; op=init and Model class= INFO stay intact.
The upsample pipes receive shared.sd_model.vae as a constructor formality;
the forward pass is pure latent to latent. Main pipe already owns that
VAE's accelerate hook lifecycle, so walking it again from the upsample
pipe raises "Cannot copy out of meta tensor" when the main pipe has
offloaded params to meta. Skip it in the walk.
Collapses the scheduler + shared.opts snapshot/mutate/restore block in
run_ltx into a single `with temp_scheduler_opts(...)` statement. Same
five pieces of state are snapshotted and restored as before; the
mechanism is unchanged per PR #4783 thread 2 feedback, only readability.
LTXCaps gains a `variant` field ('0.9', '2.0', '2.3') replacing
`is_ltx_2_3`; variant-specific branches check `caps.variant == '2.3'`
instead of grepping the model name.
ltx_util.load_upsample_2x mirrors load_upsample's contract so the 2.x
path owns a module-level cache and stops reloading ~2.3 GB every run.
The cached pipe is stamped with a synthetic
`CheckpointInfo('ltx-upsampler-2.x')` so it gets its own OffloadHook
slot and can go through apply_balanced_offload without invalidating
the main pipe's module map. The hardcoded `.to('cpu')` and post-pass
torch_gc are gone; the second apply_balanced_offload handles spill.
video_overrides comment on OzzyGT LTX-2.3 connectors states plainly
that mirrors pack weights twice by design; ignore_patterns is the
surgical workaround, not an hf_hub bug.
Drive-by: load_upsample log line used `__class__.__name__` (always
'type'); switched to `__name__`.
Each engine tab (Generic, FramePack, LTX) now owns its own size, seed,
frame count, and init imagery. The previously shared Size and Inputs
subtabs are removed; persistent encoding settings move to a single
Output subtab constructed first so engine tabs can consume its widgets.
FramePack drops its local fps/interpolate sliders and consumes the
shared Output widgets; LTX drops the redundant shared init_image
parameter and keeps only its own ltx_init_image, with run_ltx's
condition_image fallback removed in turn.
Wire the dynamic_shift checkbox into _model_change so its value resets
to caps.default_dynamic_shift on model swap, and narrow the
refine_strength interactive gate to the 0.9.x family. On 2.x refine_strength
is unused: Distilled runs a pre-baked sigma schedule, Dev runs the
canonical Stage 2 distilled-LoRA schedule, neither consumes the slider.
Implement the Lightricks two-stage recipe (diffusers PR #13217) for the
LTX-2.x Dev family: Stage 1 at half-res with full four-way guidance,
2x latent upsample, Stage 2 with distilled LoRA + scheduler swap + identity
guidance on STAGE_2_DISTILLED_SIGMA_VALUES.
Extends to both LTX-2.0 and LTX-2.3 Dev via per-family distilled-LoRA
repos carried on the caps; Distilled variants take the same flow minus
the LoRA swap. Auto-couples Refine with a fixed 2x upsample on any Dev
variant with a known LoRA when the user enables Refine without Upsample.
- caps: is_ltx_2_3, use_cross_timestep, default_dynamic_shift,
stage2_dev_lora_repo, supports_canonical_stage2, modality_default_scale,
guidance_rescale_default; LTX-2.x defaults realigned to canonical
cfg=3.0 / steps=30; per-variant STG block and four-way guidance wired
for non-distilled 2.x
- process: canonical Stage 1/Stage 2 helpers, scheduler + opts snapshot
under try/finally, per-family upsampler repo, audio latents threaded
from Stage 1 into Stage 2, use_cross_timestep gated per caps
- overrides: skip the redundant unsharded LTX-2.3 connectors blob and
share LTX2TextConnectors weights across 2.3 variants when te_shared_t5
- load: Gemma3 shared-TE path for LTX-2.3; gate use_dynamic_shifting=False
override to 0.9.x only so LTX-2.x stays on its canonical token-count
dynamic shift
Move cache tracking from ltx_util into video_load where shared.sd_model
lives, and invalidate the name-based hit when the cached class no longer
matches the current pipeline (e.g. after Unload Models triggers an
auto-reload of the default checkpoint).
- Drop the duplicate module-level loaded_model cache in ltx_util
- Add a pipe-class isinstance check around the cache hit in video_load
Vladimir Mandic
awsr
CalamitousFelicitousness
Oleksandr Liutyi