* feat(flux): add scheduler selection for Flux models
Add support for alternative diffusers Flow Matching schedulers:
- Euler (default, 1st order)
- Heun (2nd order, better quality, 2x slower)
- LCM (optimized for few steps)
Backend:
- Add schedulers.py with scheduler type definitions and class mapping
- Modify denoise.py to accept optional scheduler parameter
- Add scheduler InputField to flux_denoise invocation (v4.2.0)
Frontend:
- Add fluxScheduler to Redux state and paramsSlice
- Create ParamFluxScheduler component for Linear UI
- Add scheduler to buildFLUXGraph for generation
* fix(flux): prevent progress percentage overflow with LCM scheduler
LCM scheduler may have more internal timesteps than user-facing steps,
causing user_step to exceed total_steps. This resulted in progress
percentage > 1.0, which caused a pydantic validation error.
Fix: Only call step_callback when user_step <= total_steps.
* Ruff format
* fix(flux): remove initial step-0 callback for consistent step count
Remove the initial step_callback at step=0 to match SD/SDXL behavior.
Previously Flux showed N+1 steps (step 0 + N denoising steps), while
SD/SDXL showed only N steps. Now all models display N steps consistently.
* feat(flux): add scheduler support with metadata recall
- Handle LCM scheduler by using num_inference_steps instead of custom sigmas
- Fix progress bar to show user-facing steps instead of internal scheduler steps
- Pass scheduler parameter to Flux denoise node in graph builder
- Add model-aware metadata recall for Flux scheduler
---------
Co-authored-by: Jonathan <34005131+JPPhoto@users.noreply.github.com>
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
* feat(flux): add scheduler selection for Flux models
Add support for alternative diffusers Flow Matching schedulers:
- Euler (default, 1st order)
- Heun (2nd order, better quality, 2x slower)
- LCM (optimized for few steps)
Backend:
- Add schedulers.py with scheduler type definitions and class mapping
- Modify denoise.py to accept optional scheduler parameter
- Add scheduler InputField to flux_denoise invocation (v4.2.0)
Frontend:
- Add fluxScheduler to Redux state and paramsSlice
- Create ParamFluxScheduler component for Linear UI
- Add scheduler to buildFLUXGraph for generation
* feat(z-image): add scheduler selection for Z-Image models
Add support for alternative diffusers Flow Matching schedulers for Z-Image:
- Euler (default) - 1st order, optimized for Z-Image-Turbo (8 steps)
- Heun (2nd order) - Better quality, 2x slower
- LCM - Optimized for few-step generation
Backend:
- Extend schedulers.py with Z-Image scheduler types and mapping
- Add scheduler InputField to z_image_denoise invocation (v1.3.0)
- Refactor denoising loop to support diffusers schedulers
Frontend:
- Add zImageScheduler to Redux state in paramsSlice
- Create ParamZImageScheduler component for Linear UI
- Add scheduler to buildZImageGraph for generation
* fix ruff check
* fix(schedulers): prevent progress percentage overflow with LCM scheduler
LCM scheduler may have more internal timesteps than user-facing steps,
causing user_step to exceed total_steps. This resulted in progress
percentage > 1.0, which caused a pydantic validation error.
Fix: Only call step_callback when user_step <= total_steps.
* Ruff format
* fix(schedulers): remove initial step-0 callback for consistent step count
Remove the initial step_callback at step=0 to match SD/SDXL behavior.
Previously Flux/Z-Image showed N+1 steps (step 0 + N denoising steps),
while SD/SDXL showed only N steps. Now all models display N steps
consistently in the server log.
* feat(z-image): add scheduler support with metadata recall
- Handle LCM scheduler by using num_inference_steps instead of custom sigmas
- Fix progress bar to show user-facing steps instead of internal scheduler steps
- Pass scheduler parameter to Z-Image denoise node in graph builder
- Add model-aware metadata recall for Flux and Z-Image schedulers
---------
Co-authored-by: Jonathan <34005131+JPPhoto@users.noreply.github.com>
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
When using GGUF-quantized models on MPS (Apple Silicon), the
dequantized tensors could end up on a different device than the
other operands in math operations, causing "Expected all tensors
to be on the same device" errors.
This fix ensures that after dequantization, tensors are moved to
the same device as the other tensors in the operation.
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
Add local_files_only fallback for Qwen3 tokenizer loading in both
Checkpoint and GGUF loaders. This ensures Z-Image models can generate
images offline after the initial tokenizer download.
The tokenizer is now loaded with local_files_only=True first, falling
back to network download only if files aren't cached yet.
Fixes#8716
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
* feat: Implement PBR Maps Generation Node
* feat(ui): Add PBR Maps Generation to UI
* chore: fix typegen checks
* chore: possible fix for nvidia 5000 series cards
* fix: Use safetensor models for PBR maps instead of pickles.
* fix: incorrect naming of upconv_block for PBR network
* fix: incorrect naming of displacement map variable
* chore: add relevant docs to the PBR generate function
* fix: clear cuda cache after loading state_dict for PBR maps
* fix: load torch_device only once as multiple models are loaded
* chore(ui): update the filter icon for PBR to CubeBold
More relevant
---------
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
* Fix an issue with multiple quick-queued generations after moving bbox
After moving the canvas bbox we still handed out the previous regional-guidance mask because only two parts of the system knew anything had changed. The adapter’s
cache key doesn’t include the bbox, so the next few graph builds reused the stale mask from before the move; if the user queued several runs back‑to‑back, every
background enqueue except the last skipped rerasterizing altogether because another raster job was still in flight. The fix makes the canvas manager invalidate each
region adapter’s cached mask whenever the bbox (or a related setting) changes, and—if a reraster is already running—queues up and waits instead of bailing. Now the
first run after a bbox edit forces a new mask, and rapid-fire enqueues just wait their turn, so every queued generation gets the correct regional prompt.
* (fix) Update invokeai/frontend/web/src/features/controlLayers/konva/CanvasStateApiModule.ts
Fixes race condition identified during copilot review.
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Update invokeai/frontend/web/src/features/controlLayers/konva/CanvasStateApiModule.ts
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* Apply suggestions from code review
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
---------
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
* fix(ui): make Z-Image model selects mutually exclusive
VAE and Qwen3 Encoder selects are disabled when Qwen3 Source is selected,
and vice versa. This prevents invalid model combinations.
* feat(ui): auto-select Z-Image component models on model change
When switching to a Z-Image model, automatically set valid defaults
if no configuration exists:
- Prefers Qwen3 Source (Diffusers model) if available
- Falls back to Qwen3 Encoder + FLUX VAE combination
This ensures the generate button is enabled immediately after selecting
a Z-Image model, without requiring manual configuration.
* fix(ui): save and restore Qwen3 Source model in metadata
Qwen3 Source (Diffusers Z-Image) model was not being saved to image
metadata or restored during Remix. This adds:
- Saving qwen3_source to metadata in buildZImageGraph
- ZImageQwen3SourceModel metadata handler for parsing and recall
- i18n translation for qwen3Source
Changes image self-attention from restricted (region-isolated) to unrestricted
(all image tokens can attend to each other), similar to the FLUX approach.
This fixes the issue where ZImage-Turbo with multiple regional guidance layers
would generate two separate/disconnected images instead of compositing them
into a single unified image.
The regional text-image attention remains restricted so that each region still
responds to its corresponding prompt.
Fixes#8715
Changed the default value of model_cache_keep_alive from 0 (indefinite)
to 5 minutes as requested. This means models will now be automatically
cleared from cache after 5 minutes of inactivity by default, unless
users explicitly configure a different value.
Users can still set it to 0 in their config to get the old behavior
of keeping models indefinitely.
Co-authored-by: lstein <111189+lstein@users.noreply.github.com>
## Summary
Fix Z-Image LoRA/DoRA model detection failing during installation.
Z-Image LoRAs use different key patterns than SD/SDXL LoRAs. The base
`LoRA_LyCORIS_Config_Base` class only checked for key suffixes like
`lora_A.weight` and `lora_B.weight`, but Z-Image LoRAs (especially those
in DoRA format) use:
- `lora_down.weight` / `lora_up.weight` (standard LoRA format)
- `dora_scale` (DoRA weight decomposition)
This PR overrides `_validate_looks_like_lora` in
`LoRA_LyCORIS_ZImage_Config` to recognize Z-Image specific patterns:
- Keys starting with `diffusion_model.layers.` (Z-Image S3-DiT
architecture)
- Keys ending with `lora_down.weight`, `lora_up.weight`,
`lora_A.weight`, `lora_B.weight`, or `dora_scale`
## Related Issues / Discussions
Fixes installation of Z-Image LoRAs trained with DoRA (Weight-Decomposed
Low-Rank Adaptation).
## QA Instructions
1. Download a Z-Image LoRA in DoRA format (e.g., from CivitAI with keys
like `diffusion_model.layers.X.attention.to_k.lora_down.weight`)
2. Try to install the LoRA via Model Manager
3. Verify the model is recognized as a Z-Image LoRA and installs
successfully
4. Verify the LoRA can be applied when generating with Z-Image
## Merge Plan
Standard merge, no special considerations.
## Checklist
- [x] _The PR has a short but descriptive title, suitable for a
changelog_
- [ ] _Tests added / updated (if applicable)_
- [ ] _❗Changes to a redux slice have a corresponding migration_
- [ ] _Documentation added / updated (if applicable)_
- [ ] _Updated `What's New` copy (if doing a release after this PR)_
Two fixes for Z-Image LoRA support:
1. Override _validate_looks_like_lora in LoRA_LyCORIS_ZImage_Config to
recognize Z-Image specific LoRA formats that use different key patterns
than SD/SDXL LoRAs. Z-Image LoRAs use lora_down.weight/lora_up.weight
and dora_scale suffixes instead of lora_A.weight/lora_B.weight.
2. Fix _group_by_layer in z_image_lora_conversion_utils.py to correctly
group LoRA keys by layer name. The previous logic used rsplit with
maxsplit=2 which incorrectly grouped keys like:
- "to_k.alpha" -> layer "diffusion_model.layers.17.attention"
- "lora_down.weight" -> layer "diffusion_model.layers.17.attention.to_k"
Now uses suffix matching to ensure all keys for a layer are grouped
together (alpha, dora_scale, lora_down.weight, lora_up.weight).
Override _validate_looks_like_lora in LoRA_LyCORIS_ZImage_Config to
recognize Z-Image specific LoRA formats that use different key patterns
than SD/SDXL LoRAs.
Z-Image LoRAs (including DoRA format) use keys like:
- diffusion_model.layers.X.attention.to_k.lora_down.weight
- diffusion_model.layers.X.attention.to_k.dora_scale
The base LyCORIS config only checked for lora_A.weight/lora_B.weight
suffixes, missing the lora_down.weight/lora_up.weight and dora_scale
patterns used by Z-Image LoRAs.
* feat: Add Regional Guidance support for Z-Image model
Implements regional prompting for Z-Image (S3-DiT Transformer) allowing
different prompts to affect different image regions using attention masks.
Backend changes:
- Add ZImageRegionalPromptingExtension for mask preparation
- Add ZImageTextConditioning and ZImageRegionalTextConditioning data classes
- Patch transformer forward to inject 4D regional attention masks
- Use additive float mask (0.0 attend, -inf block) in bfloat16 for compatibility
- Alternate regional/full attention layers for global coherence
Frontend changes:
- Update buildZImageGraph to support regional conditioning collectors
- Update addRegions to create z_image_text_encoder nodes for regions
- Update addZImageLoRAs to handle optional negCond when guidance_scale=0
- Add Z-Image validation (no IP adapters, no autoNegative)
* @Pfannkuchensack
Fix windows path again
* ruff check fix
* ruff formating
* fix(ui): Z-Image CFG guidance_scale check uses > 1 instead of > 0
Changed the guidance_scale check from > 0 to > 1 for Z-Image models.
Since Z-Image uses guidance_scale=1.0 as "no CFG" (matching FLUX convention),
negative conditioning should only be created when guidance_scale > 1.
---------
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
* (bugfix)(mm) work around Windows being unable to rmtree tmp directories after GGUF install
* (style) fix ruff error
* (fix) add workaround for Windows Permission Denied on GGUF file move() call
* (fix) perform torch copy() in GGUF reader to avoid deletion failures on Windows
* (style) fix ruff formatting issues
Add support for loading Flux LoRA models in the xlabs format, which uses
keys like `double_blocks.X.processor.{qkv|proj}_lora{1|2}.{down|up}.weight`.
The xlabs format maps:
- lora1 -> img_attn (image attention stream)
- lora2 -> txt_attn (text attention stream)
- qkv -> query/key/value projection
- proj -> output projection
Changes:
- Add FluxLoRAFormat.XLabs enum value
- Add flux_xlabs_lora_conversion_utils.py with detection and conversion
- Update formats.py to detect xlabs format
- Update lora.py loader to handle xlabs format
- Update model probe to accept recognized Flux LoRA formats
- Add unit tests for xlabs format detection and conversion
Co-authored-by: Lincoln Stein <lincoln.stein@gmail.com>
