chore: bump version to v5.4.1

Turns out a gallery image's `imageDTO` object can actually be a different object by reference. I thought this was not possible thanks to how we have a quasi-normalized cache.

Need to check against image name instead of reference equality when deciding whether or not to use the single image or the gallery selection for the dnd payload.

invokeai/app/invocations/compel.py

@@ -95,6 +95,7 @@ class CompelInvocation(BaseInvocation):
                 ti_manager,
             ),
         ):
+            context.util.signal_progress("Building conditioning")
             assert isinstance(text_encoder, CLIPTextModel)
             assert isinstance(tokenizer, CLIPTokenizer)
             compel = Compel(
@@ -191,6 +192,7 @@ class SDXLPromptInvocationBase:
                 ti_manager,
             ),
         ):
+            context.util.signal_progress("Building conditioning")
             assert isinstance(text_encoder, (CLIPTextModel, CLIPTextModelWithProjection))
             assert isinstance(tokenizer, CLIPTokenizer)
 

invokeai/app/invocations/create_denoise_mask.py

@@ -65,6 +65,7 @@ class CreateDenoiseMaskInvocation(BaseInvocation):
             img_mask = tv_resize(mask, image_tensor.shape[-2:], T.InterpolationMode.BILINEAR, antialias=False)
             masked_image = image_tensor * torch.where(img_mask < 0.5, 0.0, 1.0)
             # TODO:
+            context.util.signal_progress("Running VAE encoder")
             masked_latents = ImageToLatentsInvocation.vae_encode(vae_info, self.fp32, self.tiled, masked_image.clone())
 
             masked_latents_name = context.tensors.save(tensor=masked_latents)

invokeai/app/invocations/create_gradient_mask.py

@@ -131,6 +131,7 @@ class CreateGradientMaskInvocation(BaseInvocation):
                     image_tensor = image_tensor.unsqueeze(0)
                 img_mask = tv_resize(mask, image_tensor.shape[-2:], T.InterpolationMode.BILINEAR, antialias=False)
                 masked_image = image_tensor * torch.where(img_mask < 0.5, 0.0, 1.0)
+                context.util.signal_progress("Running VAE encoder")
                 masked_latents = ImageToLatentsInvocation.vae_encode(
                     vae_info, self.fp32, self.tiled, masked_image.clone()
                 )

invokeai/app/invocations/flux_text_encoder.py

@@ -71,6 +71,7 @@ class FluxTextEncoderInvocation(BaseInvocation):
 
             t5_encoder = HFEncoder(t5_text_encoder, t5_tokenizer, False, self.t5_max_seq_len)
 
+            context.util.signal_progress("Running T5 encoder")
             prompt_embeds = t5_encoder(prompt)
 
         assert isinstance(prompt_embeds, torch.Tensor)
@@ -111,6 +112,7 @@ class FluxTextEncoderInvocation(BaseInvocation):
 
             clip_encoder = HFEncoder(clip_text_encoder, clip_tokenizer, True, 77)
 
+            context.util.signal_progress("Running CLIP encoder")
             pooled_prompt_embeds = clip_encoder(prompt)
 
         assert isinstance(pooled_prompt_embeds, torch.Tensor)

invokeai/app/invocations/flux_vae_decode.py

@@ -41,7 +41,8 @@ class FluxVaeDecodeInvocation(BaseInvocation, WithMetadata, WithBoard):
     def _vae_decode(self, vae_info: LoadedModel, latents: torch.Tensor) -> Image.Image:
         with vae_info as vae:
             assert isinstance(vae, AutoEncoder)
-            latents = latents.to(device=TorchDevice.choose_torch_device(), dtype=TorchDevice.choose_torch_dtype())
+            vae_dtype = next(iter(vae.parameters())).dtype
+            latents = latents.to(device=TorchDevice.choose_torch_device(), dtype=vae_dtype)
             img = vae.decode(latents)
 
         img = img.clamp(-1, 1)
@@ -53,6 +54,7 @@ class FluxVaeDecodeInvocation(BaseInvocation, WithMetadata, WithBoard):
     def invoke(self, context: InvocationContext) -> ImageOutput:
         latents = context.tensors.load(self.latents.latents_name)
         vae_info = context.models.load(self.vae.vae)
+        context.util.signal_progress("Running VAE")
         image = self._vae_decode(vae_info=vae_info, latents=latents)
 
         TorchDevice.empty_cache()

invokeai/app/invocations/flux_vae_encode.py

@@ -44,9 +44,8 @@ class FluxVaeEncodeInvocation(BaseInvocation):
         generator = torch.Generator(device=TorchDevice.choose_torch_device()).manual_seed(0)
         with vae_info as vae:
             assert isinstance(vae, AutoEncoder)
-            image_tensor = image_tensor.to(
-                device=TorchDevice.choose_torch_device(), dtype=TorchDevice.choose_torch_dtype()
-            )
+            vae_dtype = next(iter(vae.parameters())).dtype
+            image_tensor = image_tensor.to(device=TorchDevice.choose_torch_device(), dtype=vae_dtype)
             latents = vae.encode(image_tensor, sample=True, generator=generator)
             return latents
 
@@ -60,6 +59,7 @@ class FluxVaeEncodeInvocation(BaseInvocation):
         if image_tensor.dim() == 3:
             image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
 
+        context.util.signal_progress("Running VAE")
         latents = self.vae_encode(vae_info=vae_info, image_tensor=image_tensor)
 
         latents = latents.to("cpu")

invokeai/app/invocations/image_to_latents.py

@@ -117,6 +117,7 @@ class ImageToLatentsInvocation(BaseInvocation):
         if image_tensor.dim() == 3:
             image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
 
+        context.util.signal_progress("Running VAE encoder")
         latents = self.vae_encode(
             vae_info=vae_info, upcast=self.fp32, tiled=self.tiled, image_tensor=image_tensor, tile_size=self.tile_size
         )

invokeai/app/invocations/latents_to_image.py

@@ -60,6 +60,7 @@ class LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
         vae_info = context.models.load(self.vae.vae)
         assert isinstance(vae_info.model, (AutoencoderKL, AutoencoderTiny))
         with SeamlessExt.static_patch_model(vae_info.model, self.vae.seamless_axes), vae_info as vae:
+            context.util.signal_progress("Running VAE decoder")
             assert isinstance(vae, (AutoencoderKL, AutoencoderTiny))
             latents = latents.to(vae.device)
             if self.fp32:

invokeai/app/invocations/metadata.py

@@ -147,6 +147,10 @@ GENERATION_MODES = Literal[
     "flux_img2img",
     "flux_inpaint",
     "flux_outpaint",
+    "sd3_txt2img",
+    "sd3_img2img",
+    "sd3_inpaint",
+    "sd3_outpaint",
 ]
 
 

invokeai/app/invocations/sd3_denoise.py

@@ -1,16 +1,19 @@
-from typing import Callable, Tuple
+from typing import Callable, Optional, Tuple
 
 import torch
+import torchvision.transforms as tv_transforms
 from diffusers.models.transformers.transformer_sd3 import SD3Transformer2DModel
-from diffusers.schedulers.scheduling_flow_match_euler_discrete import FlowMatchEulerDiscreteScheduler
+from torchvision.transforms.functional import resize as tv_resize
 from tqdm import tqdm
 
 from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
 from invokeai.app.invocations.constants import LATENT_SCALE_FACTOR
 from invokeai.app.invocations.fields import (
+    DenoiseMaskField,
     FieldDescriptions,
     Input,
     InputField,
+    LatentsField,
     SD3ConditioningField,
     WithBoard,
     WithMetadata,
@@ -19,7 +22,9 @@ from invokeai.app.invocations.model import TransformerField
 from invokeai.app.invocations.primitives import LatentsOutput
 from invokeai.app.invocations.sd3_text_encoder import SD3_T5_MAX_SEQ_LEN
 from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.sampling_utils import clip_timestep_schedule_fractional
 from invokeai.backend.model_manager.config import BaseModelType
+from invokeai.backend.sd3.extensions.inpaint_extension import InpaintExtension
 from invokeai.backend.stable_diffusion.diffusers_pipeline import PipelineIntermediateState
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import SD3ConditioningInfo
 from invokeai.backend.util.devices import TorchDevice
@@ -30,16 +35,24 @@ from invokeai.backend.util.devices import TorchDevice
     title="SD3 Denoise",
     tags=["image", "sd3"],
     category="image",
-    version="1.0.0",
+    version="1.1.0",
     classification=Classification.Prototype,
 )
 class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
     """Run denoising process with a SD3 model."""
 
+    # If latents is provided, this means we are doing image-to-image.
+    latents: Optional[LatentsField] = InputField(
+        default=None, description=FieldDescriptions.latents, input=Input.Connection
+    )
+    # denoise_mask is used for image-to-image inpainting. Only the masked region is modified.
+    denoise_mask: Optional[DenoiseMaskField] = InputField(
+        default=None, description=FieldDescriptions.denoise_mask, input=Input.Connection
+    )
+    denoising_start: float = InputField(default=0.0, ge=0, le=1, description=FieldDescriptions.denoising_start)
+    denoising_end: float = InputField(default=1.0, ge=0, le=1, description=FieldDescriptions.denoising_end)
     transformer: TransformerField = InputField(
-        description=FieldDescriptions.sd3_model,
-        input=Input.Connection,
-        title="Transformer",
+        description=FieldDescriptions.sd3_model, input=Input.Connection, title="Transformer"
     )
     positive_conditioning: SD3ConditioningField = InputField(
         description=FieldDescriptions.positive_cond, input=Input.Connection
@@ -61,6 +74,41 @@ class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
         name = context.tensors.save(tensor=latents)
         return LatentsOutput.build(latents_name=name, latents=latents, seed=None)
 
+    def _prep_inpaint_mask(self, context: InvocationContext, latents: torch.Tensor) -> torch.Tensor | None:
+        """Prepare the inpaint mask.
+        - Loads the mask
+        - Resizes if necessary
+        - Casts to same device/dtype as latents
+
+        Args:
+            context (InvocationContext): The invocation context, for loading the inpaint mask.
+            latents (torch.Tensor): A latent image tensor. Used to determine the target shape, device, and dtype for the
+                inpaint mask.
+
+        Returns:
+            torch.Tensor | None: Inpaint mask. Values of 0.0 represent the regions to be fully denoised, and 1.0
+                represent the regions to be preserved.
+        """
+        if self.denoise_mask is None:
+            return None
+        mask = context.tensors.load(self.denoise_mask.mask_name)
+
+        # The input denoise_mask contains values in [0, 1], where 0.0 represents the regions to be fully denoised, and
+        # 1.0 represents the regions to be preserved.
+        # We invert the mask so that the regions to be preserved are 0.0 and the regions to be denoised are 1.0.
+        mask = 1.0 - mask
+
+        _, _, latent_height, latent_width = latents.shape
+        mask = tv_resize(
+            img=mask,
+            size=[latent_height, latent_width],
+            interpolation=tv_transforms.InterpolationMode.BILINEAR,
+            antialias=False,
+        )
+
+        mask = mask.to(device=latents.device, dtype=latents.dtype)
+        return mask
+
     def _load_text_conditioning(
         self,
         context: InvocationContext,
@@ -170,14 +218,20 @@ class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
         prompt_embeds = torch.cat([neg_prompt_embeds, pos_prompt_embeds], dim=0)
         pooled_prompt_embeds = torch.cat([neg_pooled_prompt_embeds, pos_pooled_prompt_embeds], dim=0)
 
-        # Prepare the scheduler.
-        scheduler = FlowMatchEulerDiscreteScheduler()
-        scheduler.set_timesteps(num_inference_steps=self.steps, device=device)
-        timesteps = scheduler.timesteps
-        assert isinstance(timesteps, torch.Tensor)
+        # Prepare the timestep schedule.
+        # We add an extra step to the end to account for the final timestep of 0.0.
+        timesteps: list[float] = torch.linspace(1, 0, self.steps + 1).tolist()
+        # Clip the timesteps schedule based on denoising_start and denoising_end.
+        timesteps = clip_timestep_schedule_fractional(timesteps, self.denoising_start, self.denoising_end)
+        total_steps = len(timesteps) - 1
 
         # Prepare the CFG scale list.
-        cfg_scale = self._prepare_cfg_scale(len(timesteps))
+        cfg_scale = self._prepare_cfg_scale(total_steps)
+
+        # Load the input latents, if provided.
+        init_latents = context.tensors.load(self.latents.latents_name) if self.latents else None
+        if init_latents is not None:
+            init_latents = init_latents.to(device=device, dtype=inference_dtype)
 
         # Generate initial latent noise.
         num_channels_latents = transformer_info.model.config.in_channels
@@ -191,9 +245,34 @@ class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
             device=device,
             seed=self.seed,
         )
-        latents: torch.Tensor = noise
 
-        total_steps = len(timesteps)
+        # Prepare input latent image.
+        if init_latents is not None:
+            # Noise the init_latents by the appropriate amount for the first timestep.
+            t_0 = timesteps[0]
+            latents = t_0 * noise + (1.0 - t_0) * init_latents
+        else:
+            # init_latents are not provided, so we are not doing image-to-image (i.e. we are starting from pure noise).
+            if self.denoising_start > 1e-5:
+                raise ValueError("denoising_start should be 0 when initial latents are not provided.")
+            latents = noise
+
+        # If len(timesteps) == 1, then short-circuit. We are just noising the input latents, but not taking any
+        # denoising steps.
+        if len(timesteps) <= 1:
+            return latents
+
+        # Prepare inpaint extension.
+        inpaint_mask = self._prep_inpaint_mask(context, latents)
+        inpaint_extension: InpaintExtension | None = None
+        if inpaint_mask is not None:
+            assert init_latents is not None
+            inpaint_extension = InpaintExtension(
+                init_latents=init_latents,
+                inpaint_mask=inpaint_mask,
+                noise=noise,
+            )
+
         step_callback = self._build_step_callback(context)
 
         step_callback(
@@ -210,11 +289,12 @@ class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
             assert isinstance(transformer, SD3Transformer2DModel)
 
             # 6. Denoising loop
-            for step_idx, t in tqdm(list(enumerate(timesteps))):
+            for step_idx, (t_curr, t_prev) in tqdm(list(enumerate(zip(timesteps[:-1], timesteps[1:], strict=True)))):
                 # Expand the latents if we are doing CFG.
                 latent_model_input = torch.cat([latents] * 2) if do_classifier_free_guidance else latents
                 # Expand the timestep to match the latent model input.
-                timestep = t.expand(latent_model_input.shape[0])
+                # Multiply by 1000 to match the default FlowMatchEulerDiscreteScheduler num_train_timesteps.
+                timestep = torch.tensor([t_curr * 1000], device=device).expand(latent_model_input.shape[0])
 
                 noise_pred = transformer(
                     hidden_states=latent_model_input,
@@ -232,21 +312,19 @@ class SD3DenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
 
                 # Compute the previous noisy sample x_t -> x_t-1.
                 latents_dtype = latents.dtype
-                latents = scheduler.step(model_output=noise_pred, timestep=t, sample=latents, return_dict=False)[0]
+                latents = latents.to(dtype=torch.float32)
+                latents = latents + (t_prev - t_curr) * noise_pred
+                latents = latents.to(dtype=latents_dtype)
 
-                # TODO(ryand): This MPS dtype handling was copied from diffusers, I haven't tested to see if it's
-                # needed.
-                if latents.dtype != latents_dtype:
-                    if torch.backends.mps.is_available():
-                        # some platforms (eg. apple mps) misbehave due to a pytorch bug: https://github.com/pytorch/pytorch/pull/99272
-                        latents = latents.to(latents_dtype)
+                if inpaint_extension is not None:
+                    latents = inpaint_extension.merge_intermediate_latents_with_init_latents(latents, t_prev)
 
                 step_callback(
                     PipelineIntermediateState(
                         step=step_idx + 1,
                         order=1,
                         total_steps=total_steps,
-                        timestep=int(t),
+                        timestep=int(t_curr),
                         latents=latents,
                     ),
                 )

invokeai/app/invocations/sd3_image_to_latents.py

@@ -0,0 +1,65 @@
+import einops
+import torch
+from diffusers.models.autoencoders.autoencoder_kl import AutoencoderKL
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    ImageField,
+    Input,
+    InputField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import VAEField
+from invokeai.app.invocations.primitives import LatentsOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.model_manager.load.load_base import LoadedModel
+from invokeai.backend.stable_diffusion.diffusers_pipeline import image_resized_to_grid_as_tensor
+
+
+@invocation(
+    "sd3_i2l",
+    title="SD3 Image to Latents",
+    tags=["image", "latents", "vae", "i2l", "sd3"],
+    category="image",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class SD3ImageToLatentsInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Generates latents from an image."""
+
+    image: ImageField = InputField(description="The image to encode")
+    vae: VAEField = InputField(description=FieldDescriptions.vae, input=Input.Connection)
+
+    @staticmethod
+    def vae_encode(vae_info: LoadedModel, image_tensor: torch.Tensor) -> torch.Tensor:
+        with vae_info as vae:
+            assert isinstance(vae, AutoencoderKL)
+
+            vae.disable_tiling()
+
+            image_tensor = image_tensor.to(device=vae.device, dtype=vae.dtype)
+            with torch.inference_mode():
+                image_tensor_dist = vae.encode(image_tensor).latent_dist
+                # TODO: Use seed to make sampling reproducible.
+                latents: torch.Tensor = image_tensor_dist.sample().to(dtype=vae.dtype)
+
+            latents = vae.config.scaling_factor * latents
+
+        return latents
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        image = context.images.get_pil(self.image.image_name)
+
+        image_tensor = image_resized_to_grid_as_tensor(image.convert("RGB"))
+        if image_tensor.dim() == 3:
+            image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
+
+        vae_info = context.models.load(self.vae.vae)
+        latents = self.vae_encode(vae_info=vae_info, image_tensor=image_tensor)
+
+        latents = latents.to("cpu")
+        name = context.tensors.save(tensor=latents)
+        return LatentsOutput.build(latents_name=name, latents=latents, seed=None)

invokeai/app/invocations/sd3_latents_to_image.py

@@ -47,6 +47,7 @@ class SD3LatentsToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
         vae_info = context.models.load(self.vae.vae)
         assert isinstance(vae_info.model, (AutoencoderKL))
         with SeamlessExt.static_patch_model(vae_info.model, self.vae.seamless_axes), vae_info as vae:
+            context.util.signal_progress("Running VAE")
             assert isinstance(vae, (AutoencoderKL))
             latents = latents.to(vae.device)
 

invokeai/app/invocations/sd3_text_encoder.py

@@ -95,6 +95,7 @@ class Sd3TextEncoderInvocation(BaseInvocation):
             t5_text_encoder_info as t5_text_encoder,
             t5_tokenizer_info as t5_tokenizer,
         ):
+            context.util.signal_progress("Running T5 encoder")
             assert isinstance(t5_text_encoder, T5EncoderModel)
             assert isinstance(t5_tokenizer, (T5Tokenizer, T5TokenizerFast))
 
@@ -137,6 +138,7 @@ class Sd3TextEncoderInvocation(BaseInvocation):
             clip_tokenizer_info as clip_tokenizer,
             ExitStack() as exit_stack,
         ):
+            context.util.signal_progress("Running CLIP encoder")
             assert isinstance(clip_text_encoder, (CLIPTextModel, CLIPTextModelWithProjection))
             assert isinstance(clip_tokenizer, CLIPTokenizer)
 

invokeai/app/services/shared/invocation_context.py

@@ -160,6 +160,10 @@ class LoggerInterface(InvocationContextInterface):
 
 
 class ImagesInterface(InvocationContextInterface):
+    def __init__(self, services: InvocationServices, data: InvocationContextData, util: "UtilInterface") -> None:
+        super().__init__(services, data)
+        self._util = util
+
     def save(
         self,
         image: Image,
@@ -186,6 +190,8 @@ class ImagesInterface(InvocationContextInterface):
             The saved image DTO.
         """
 
+        self._util.signal_progress("Saving image")
+
         # If `metadata` is provided directly, use that. Else, use the metadata provided by `WithMetadata`, falling back to None.
         metadata_ = None
         if metadata:
@@ -336,6 +342,10 @@ class ConditioningInterface(InvocationContextInterface):
 class ModelsInterface(InvocationContextInterface):
     """Common API for loading, downloading and managing models."""
 
+    def __init__(self, services: InvocationServices, data: InvocationContextData, util: "UtilInterface") -> None:
+        super().__init__(services, data)
+        self._util = util
+
     def exists(self, identifier: Union[str, "ModelIdentifierField"]) -> bool:
         """Check if a model exists.
 
@@ -368,11 +378,15 @@ class ModelsInterface(InvocationContextInterface):
 
         if isinstance(identifier, str):
             model = self._services.model_manager.store.get_model(identifier)
-            return self._services.model_manager.load.load_model(model, submodel_type)
         else:
-            _submodel_type = submodel_type or identifier.submodel_type
+            submodel_type = submodel_type or identifier.submodel_type
             model = self._services.model_manager.store.get_model(identifier.key)
-            return self._services.model_manager.load.load_model(model, _submodel_type)
+
+        message = f"Loading model {model.name}"
+        if submodel_type:
+            message += f" ({submodel_type.value})"
+        self._util.signal_progress(message)
+        return self._services.model_manager.load.load_model(model, submodel_type)
 
     def load_by_attrs(
         self, name: str, base: BaseModelType, type: ModelType, submodel_type: Optional[SubModelType] = None
@@ -397,6 +411,10 @@ class ModelsInterface(InvocationContextInterface):
         if len(configs) > 1:
             raise ValueError(f"More than one model found with name {name}, base {base}, and type {type}")
 
+        message = f"Loading model {name}"
+        if submodel_type:
+            message += f" ({submodel_type.value})"
+        self._util.signal_progress(message)
         return self._services.model_manager.load.load_model(configs[0], submodel_type)
 
     def get_config(self, identifier: Union[str, "ModelIdentifierField"]) -> AnyModelConfig:
@@ -467,6 +485,7 @@ class ModelsInterface(InvocationContextInterface):
         Returns:
             Path to the downloaded model
         """
+        self._util.signal_progress(f"Downloading model {source}")
         return self._services.model_manager.install.download_and_cache_model(source=source)
 
     def load_local_model(
@@ -489,6 +508,8 @@ class ModelsInterface(InvocationContextInterface):
         Returns:
             A LoadedModelWithoutConfig object.
         """
+
+        self._util.signal_progress(f"Loading model {model_path.name}")
         return self._services.model_manager.load.load_model_from_path(model_path=model_path, loader=loader)
 
     def load_remote_model(
@@ -514,6 +535,8 @@ class ModelsInterface(InvocationContextInterface):
             A LoadedModelWithoutConfig object.
         """
         model_path = self._services.model_manager.install.download_and_cache_model(source=str(source))
+
+        self._util.signal_progress(f"Loading model {source}")
         return self._services.model_manager.load.load_model_from_path(model_path=model_path, loader=loader)
 
 
@@ -707,12 +730,12 @@ def build_invocation_context(
     """
 
     logger = LoggerInterface(services=services, data=data)
-    images = ImagesInterface(services=services, data=data)
     tensors = TensorsInterface(services=services, data=data)
-    models = ModelsInterface(services=services, data=data)
     config = ConfigInterface(services=services, data=data)
     util = UtilInterface(services=services, data=data, is_canceled=is_canceled)
     conditioning = ConditioningInterface(services=services, data=data)
+    models = ModelsInterface(services=services, data=data, util=util)
+    images = ImagesInterface(services=services, data=data, util=util)
     boards = BoardsInterface(services=services, data=data)
 
     ctx = InvocationContext(

invokeai/backend/lora/conversions/flux_diffusers_lora_conversion_utils.py

@@ -45,8 +45,9 @@ def lora_model_from_flux_diffusers_state_dict(state_dict: Dict[str, torch.Tensor
     # Constants for FLUX.1
     num_double_layers = 19
     num_single_layers = 38
-    # inner_dim = 3072
-    # mlp_ratio = 4.0
+    hidden_size = 3072
+    mlp_ratio = 4.0
+    mlp_hidden_dim = int(hidden_size * mlp_ratio)
 
     layers: dict[str, AnyLoRALayer] = {}
 
@@ -62,30 +63,43 @@ def lora_model_from_flux_diffusers_state_dict(state_dict: Dict[str, torch.Tensor
             layers[dst_key] = LoRALayer.from_state_dict_values(values=value)
             assert len(src_layer_dict) == 0
 
-    def add_qkv_lora_layer_if_present(src_keys: list[str], dst_qkv_key: str) -> None:
+    def add_qkv_lora_layer_if_present(
+        src_keys: list[str],
+        src_weight_shapes: list[tuple[int, int]],
+        dst_qkv_key: str,
+        allow_missing_keys: bool = False,
+    ) -> None:
         """Handle the Q, K, V matrices for a transformer block. We need special handling because the diffusers format
         stores them in separate matrices, whereas the BFL format used internally by InvokeAI concatenates them.
         """
-        # We expect that either all src keys are present or none of them are. Verify this.
-        keys_present = [key in grouped_state_dict for key in src_keys]
-        assert all(keys_present) or not any(keys_present)
-
         # If none of the keys are present, return early.
+        keys_present = [key in grouped_state_dict for key in src_keys]
         if not any(keys_present):
             return
 
-        src_layer_dicts = [grouped_state_dict.pop(key) for key in src_keys]
         sub_layers: list[LoRALayer] = []
-        for src_layer_dict in src_layer_dicts:
-            values = {
-                "lora_down.weight": src_layer_dict.pop("lora_A.weight"),
-                "lora_up.weight": src_layer_dict.pop("lora_B.weight"),
-            }
-            if alpha is not None:
-                values["alpha"] = torch.tensor(alpha)
-            sub_layers.append(LoRALayer.from_state_dict_values(values=values))
-            assert len(src_layer_dict) == 0
-        layers[dst_qkv_key] = ConcatenatedLoRALayer(lora_layers=sub_layers, concat_axis=0)
+        for src_key, src_weight_shape in zip(src_keys, src_weight_shapes, strict=True):
+            src_layer_dict = grouped_state_dict.pop(src_key, None)
+            if src_layer_dict is not None:
+                values = {
+                    "lora_down.weight": src_layer_dict.pop("lora_A.weight"),
+                    "lora_up.weight": src_layer_dict.pop("lora_B.weight"),
+                }
+                if alpha is not None:
+                    values["alpha"] = torch.tensor(alpha)
+                assert values["lora_down.weight"].shape[1] == src_weight_shape[1]
+                assert values["lora_up.weight"].shape[0] == src_weight_shape[0]
+                sub_layers.append(LoRALayer.from_state_dict_values(values=values))
+                assert len(src_layer_dict) == 0
+            else:
+                if not allow_missing_keys:
+                    raise ValueError(f"Missing LoRA layer: '{src_key}'.")
+                values = {
+                    "lora_up.weight": torch.zeros((src_weight_shape[0], 1)),
+                    "lora_down.weight": torch.zeros((1, src_weight_shape[1])),
+                }
+                sub_layers.append(LoRALayer.from_state_dict_values(values=values))
+        layers[dst_qkv_key] = ConcatenatedLoRALayer(lora_layers=sub_layers)
 
     # time_text_embed.timestep_embedder -> time_in.
     add_lora_layer_if_present("time_text_embed.timestep_embedder.linear_1", "time_in.in_layer")
@@ -118,6 +132,7 @@ def lora_model_from_flux_diffusers_state_dict(state_dict: Dict[str, torch.Tensor
                 f"transformer_blocks.{i}.attn.to_k",
                 f"transformer_blocks.{i}.attn.to_v",
             ],
+            [(hidden_size, hidden_size), (hidden_size, hidden_size), (hidden_size, hidden_size)],
             f"double_blocks.{i}.img_attn.qkv",
         )
         add_qkv_lora_layer_if_present(
@@ -126,6 +141,7 @@ def lora_model_from_flux_diffusers_state_dict(state_dict: Dict[str, torch.Tensor
                 f"transformer_blocks.{i}.attn.add_k_proj",
                 f"transformer_blocks.{i}.attn.add_v_proj",
             ],
+            [(hidden_size, hidden_size), (hidden_size, hidden_size), (hidden_size, hidden_size)],
             f"double_blocks.{i}.txt_attn.qkv",
         )
 
@@ -175,7 +191,14 @@ def lora_model_from_flux_diffusers_state_dict(state_dict: Dict[str, torch.Tensor
                 f"single_transformer_blocks.{i}.attn.to_v",
                 f"single_transformer_blocks.{i}.proj_mlp",
             ],
+            [
+                (hidden_size, hidden_size),
+                (hidden_size, hidden_size),
+                (hidden_size, hidden_size),
+                (mlp_hidden_dim, hidden_size),
+            ],
             f"single_blocks.{i}.linear1",
+            allow_missing_keys=True,
         )
 
         # Output projections.

invokeai/backend/model_manager/load/load_default.py

@@ -35,6 +35,7 @@ class ModelLoader(ModelLoaderBase):
         self._logger = logger
         self._ram_cache = ram_cache
         self._torch_dtype = TorchDevice.choose_torch_dtype()
+        self._torch_device = TorchDevice.choose_torch_device()
 
     def load_model(self, model_config: AnyModelConfig, submodel_type: Optional[SubModelType] = None) -> LoadedModel:
         """

invokeai/backend/model_manager/load/model_loaders/flux.py

@@ -84,7 +84,15 @@ class FluxVAELoader(ModelLoader):
             model = AutoEncoder(ae_params[config.config_path])
             sd = load_file(model_path)
             model.load_state_dict(sd, assign=True)
-            model.to(dtype=self._torch_dtype)
+            # VAE is broken in float16, which mps defaults to
+            if self._torch_dtype == torch.float16:
+                try:
+                    vae_dtype = torch.tensor([1.0], dtype=torch.bfloat16, device=self._torch_device).dtype
+                except TypeError:
+                    vae_dtype = torch.float32
+            else:
+                vae_dtype = self._torch_dtype
+            model.to(vae_dtype)
 
         return model
 

invokeai/backend/model_manager/starter_models.py

@@ -300,7 +300,7 @@ ip_adapter_sdxl = StarterModel(
 ip_adapter_flux = StarterModel(
     name="Standard Reference (XLabs FLUX IP-Adapter)",
     base=BaseModelType.Flux,
-    source="https://huggingface.co/XLabs-AI/flux-ip-adapter/resolve/main/flux-ip-adapter.safetensors",
+    source="https://huggingface.co/XLabs-AI/flux-ip-adapter/resolve/main/ip_adapter.safetensors",
     description="References images with a more generalized/looser degree of precision.",
     type=ModelType.IPAdapter,
     dependencies=[clip_vit_l_image_encoder],

invokeai/backend/sd3/extensions/inpaint_extension.py

@@ -0,0 +1,58 @@
+import torch
+
+
+class InpaintExtension:
+    """A class for managing inpainting with SD3."""
+
+    def __init__(self, init_latents: torch.Tensor, inpaint_mask: torch.Tensor, noise: torch.Tensor):
+        """Initialize InpaintExtension.
+
+        Args:
+            init_latents (torch.Tensor): The initial latents (i.e. un-noised at timestep 0).
+            inpaint_mask (torch.Tensor): A mask specifying which elements to inpaint. Range [0, 1]. Values of 1 will be
+                re-generated. Values of 0 will remain unchanged. Values between 0 and 1 can be used to blend the
+                inpainted region with the background.
+            noise (torch.Tensor): The noise tensor used to noise the init_latents.
+        """
+        assert init_latents.dim() == inpaint_mask.dim() == noise.dim() == 4
+        assert init_latents.shape[-2:] == inpaint_mask.shape[-2:] == noise.shape[-2:]
+
+        self._init_latents = init_latents
+        self._inpaint_mask = inpaint_mask
+        self._noise = noise
+
+    def _apply_mask_gradient_adjustment(self, t_prev: float) -> torch.Tensor:
+        """Applies inpaint mask gradient adjustment and returns the inpaint mask to be used at the current timestep."""
+        # As we progress through the denoising process, we promote gradient regions of the mask to have a full weight of
+        # 1.0. This helps to produce more coherent seams around the inpainted region. We experimented with a (small)
+        # number of promotion strategies (e.g. gradual promotion based on timestep), but found that a simple cutoff
+        # threshold worked well.
+        # We use a small epsilon to avoid any potential issues with floating point precision.
+        eps = 1e-4
+        mask_gradient_t_cutoff = 0.5
+        if t_prev > mask_gradient_t_cutoff:
+            # Early in the denoising process, use the inpaint mask as-is.
+            return self._inpaint_mask
+        else:
+            # After the cut-off, promote all non-zero mask values to 1.0.
+            mask = self._inpaint_mask.where(self._inpaint_mask <= (0.0 + eps), 1.0)
+
+        return mask
+
+    def merge_intermediate_latents_with_init_latents(
+        self, intermediate_latents: torch.Tensor, t_prev: float
+    ) -> torch.Tensor:
+        """Merge the intermediate latents with the initial latents for the current timestep using the inpaint mask. I.e.
+        update the intermediate latents to keep the regions that are not being inpainted on the correct noise
+        trajectory.
+
+        This function should be called after each denoising step.
+        """
+
+        mask = self._apply_mask_gradient_adjustment(t_prev)
+
+        # Noise the init latents for the current timestep.
+        noised_init_latents = self._noise * t_prev + (1.0 - t_prev) * self._init_latents
+
+        # Merge the intermediate latents with the noised_init_latents using the inpaint_mask.
+        return intermediate_latents * mask + noised_init_latents * (1.0 - mask)

invokeai/frontend/web/public/locales/en.json

@@ -174,7 +174,8 @@
         "placeholderSelectAModel": "Select a model",
         "reset": "Reset",
         "none": "None",
-        "new": "New"
+        "new": "New",
+        "generating": "Generating"
     },
     "hrf": {
         "hrf": "High Resolution Fix",
@@ -704,6 +705,8 @@
         "baseModel": "Base Model",
         "cancel": "Cancel",
         "clipEmbed": "CLIP Embed",
+        "clipLEmbed": "CLIP-L Embed",
+        "clipGEmbed": "CLIP-G Embed",
         "config": "Config",
         "convert": "Convert",
         "convertingModelBegin": "Converting Model. Please wait.",
@@ -997,7 +1000,7 @@
         "controlNetControlMode": "Control Mode",
         "copyImage": "Copy Image",
         "denoisingStrength": "Denoising Strength",
-        "noRasterLayers": "No Raster Layers",
+        "disabledNoRasterContent": "Disabled (No Raster Content)",
         "downloadImage": "Download Image",
         "general": "General",
         "guidance": "Guidance",
@@ -1137,6 +1140,7 @@
         "resetWebUI": "Reset Web UI",
         "resetWebUIDesc1": "Resetting the web UI only resets the browser's local cache of your images and remembered settings. It does not delete any images from disk.",
         "resetWebUIDesc2": "If images aren't showing up in the gallery or something else isn't working, please try resetting before submitting an issue on GitHub.",
+        "showDetailedInvocationProgress": "Show Progress Details",
         "showProgressInViewer": "Show Progress Images in Viewer",
         "ui": "User Interface",
         "clearIntermediatesDisabled": "Queue must be empty to clear intermediates",
@@ -1671,7 +1675,7 @@
         "clearCaches": "Clear Caches",
         "recalculateRects": "Recalculate Rects",
         "clipToBbox": "Clip Strokes to Bbox",
-        "outputOnlyMaskedRegions": "Output Only Masked Regions",
+        "outputOnlyMaskedRegions": "Output Only Generated Regions",
         "addLayer": "Add Layer",
         "duplicate": "Duplicate",
         "moveToFront": "Move to Front",
@@ -1787,7 +1791,7 @@
         },
         "ipAdapterMethod": {
             "ipAdapterMethod": "IP Adapter Method",
-            "full": "Full",
+            "full": "Style and Composition",
             "style": "Style Only",
             "composition": "Composition Only"
         },
@@ -1999,7 +2003,9 @@
         "upscaleModelDesc": "Upscale (image to image) model",
         "missingUpscaleInitialImage": "Missing initial image for upscaling",
         "missingUpscaleModel": "Missing upscale model",
-        "missingTileControlNetModel": "No valid tile ControlNet models installed"
+        "missingTileControlNetModel": "No valid tile ControlNet models installed",
+        "incompatibleBaseModel": "Unsupported main model architecture for upscaling",
+        "incompatibleBaseModelDesc": "Upscaling is supported for SD1.5 and SDXL architecture models only. Change the main model to enable upscaling."
     },
     "stylePresets": {
         "active": "Active",

invokeai/frontend/web/src/app/store/middleware/listenerMiddleware/listeners/enqueueRequestedLinear.ts

@@ -8,6 +8,7 @@ import { $canvasManager } from 'features/controlLayers/store/ephemeral';
 import { prepareLinearUIBatch } from 'features/nodes/util/graph/buildLinearBatchConfig';
 import { buildFLUXGraph } from 'features/nodes/util/graph/generation/buildFLUXGraph';
 import { buildSD1Graph } from 'features/nodes/util/graph/generation/buildSD1Graph';
+import { buildSD3Graph } from 'features/nodes/util/graph/generation/buildSD3Graph';
 import { buildSDXLGraph } from 'features/nodes/util/graph/generation/buildSDXLGraph';
 import type { Graph } from 'features/nodes/util/graph/generation/Graph';
 import { toast } from 'features/toast/toast';
@@ -34,8 +35,8 @@ export const addEnqueueRequestedLinear = (startAppListening: AppStartListening)
       let buildGraphResult: Result<
         {
           g: Graph;
-          noise: Invocation<'noise' | 'flux_denoise'>;
-          posCond: Invocation<'compel' | 'sdxl_compel_prompt' | 'flux_text_encoder'>;
+          noise: Invocation<'noise' | 'flux_denoise' | 'sd3_denoise'>;
+          posCond: Invocation<'compel' | 'sdxl_compel_prompt' | 'flux_text_encoder' | 'sd3_text_encoder'>;
         },
         Error
       >;
@@ -51,6 +52,9 @@ export const addEnqueueRequestedLinear = (startAppListening: AppStartListening)
         case `sd-2`:
           buildGraphResult = await withResultAsync(() => buildSD1Graph(state, manager));
           break;
+        case `sd-3`:
+          buildGraphResult = await withResultAsync(() => buildSD3Graph(state, manager));
+          break;
         case `flux`:
           buildGraphResult = await withResultAsync(() => buildFLUXGraph(state, manager));
           break;

invokeai/frontend/web/src/app/store/middleware/listenerMiddleware/listeners/imageUploaded.ts

@@ -41,29 +41,33 @@ export const addImageUploadedFulfilledListener = (startAppListening: AppStartLis
 
       log.debug({ imageDTO }, 'Image uploaded');
 
+      if (action.meta.arg.originalArgs.silent || imageDTO.is_intermediate) {
+        // When a "silent" upload is requested, or the image is intermediate, we can skip all post-upload actions,
+        // like toasts and switching the gallery view
+        return;
+      }
+
       const boardId = imageDTO.board_id ?? 'none';
 
-      if (action.meta.arg.originalArgs.withToast) {
-        const DEFAULT_UPLOADED_TOAST = {
-          id: 'IMAGE_UPLOADED',
-          title: t('toast.imageUploaded'),
-          status: 'success',
-        } as const;
+      const DEFAULT_UPLOADED_TOAST = {
+        id: 'IMAGE_UPLOADED',
+        title: t('toast.imageUploaded'),
+        status: 'success',
+      } as const;
 
-        // default action - just upload and alert user
-        if (lastUploadedToastTimeout !== null) {
-          window.clearTimeout(lastUploadedToastTimeout);
-        }
-        const toastApi = toast({
-          ...DEFAULT_UPLOADED_TOAST,
-          title: DEFAULT_UPLOADED_TOAST.title,
-          description: getUploadedToastDescription(boardId, state),
-          duration: null, // we will close the toast manually
-        });
-        lastUploadedToastTimeout = window.setTimeout(() => {
-          toastApi.close();
-        }, 3000);
+      // default action - just upload and alert user
+      if (lastUploadedToastTimeout !== null) {
+        window.clearTimeout(lastUploadedToastTimeout);
       }
+      const toastApi = toast({
+        ...DEFAULT_UPLOADED_TOAST,
+        title: DEFAULT_UPLOADED_TOAST.title,
+        description: getUploadedToastDescription(boardId, state),
+        duration: null, // we will close the toast manually
+      });
+      lastUploadedToastTimeout = window.setTimeout(() => {
+        toastApi.close();
+      }, 3000);
 
       /**
        * We only want to change the board and view if this is the first upload of a batch, else we end up hijacking

invokeai/frontend/web/src/app/types/invokeai.ts

@@ -25,7 +25,8 @@ export type AppFeature =
   | 'invocationCache'
   | 'bulkDownload'
   | 'starterModels'
-  | 'hfToken';
+  | 'hfToken'
+  | 'invocationProgressAlert';
 
 /**
  * A disable-able Stable Diffusion feature

invokeai/frontend/web/src/common/hooks/useImageUploadButton.tsx

@@ -61,6 +61,11 @@ export const useImageUploadButton = ({ onUpload, isDisabled, allowMultiple }: Us
           log.warn('Multiple files dropped but only one allowed');
           return;
         }
+        if (files.length === 0) {
+          // Should never happen
+          log.warn('No files dropped');
+          return;
+        }
         const file = files[0];
         assert(file !== undefined); // should never happen
         const imageDTO = await uploadImage({
@@ -68,18 +73,20 @@ export const useImageUploadButton = ({ onUpload, isDisabled, allowMultiple }: Us
           image_category: 'user',
           is_intermediate: false,
           board_id: autoAddBoardId === 'none' ? undefined : autoAddBoardId,
+          silent: true,
         }).unwrap();
         if (onUpload) {
           onUpload(imageDTO);
         }
       } else {
-        //
         const imageDTOs = await uploadImages(
-          files.map((file) => ({
+          files.map((file, i) => ({
             file,
             image_category: 'user',
             is_intermediate: false,
             board_id: autoAddBoardId === 'none' ? undefined : autoAddBoardId,
+            silent: false,
+            isFirstUploadOfBatch: i === 0,
           }))
         );
         if (onUpload) {

invokeai/frontend/web/src/common/hooks/useIsReadyToEnqueue.ts

@@ -119,11 +119,20 @@ const createSelector = (
             reasons.push({ content: i18n.t('upscaling.exceedsMaxSize') });
           }
         }
-        if (!upscale.upscaleModel) {
-          reasons.push({ content: i18n.t('upscaling.missingUpscaleModel') });
-        }
-        if (!upscale.tileControlnetModel) {
-          reasons.push({ content: i18n.t('upscaling.missingTileControlNetModel') });
+        if (model && !['sd-1', 'sdxl'].includes(model.base)) {
+          // When we are using an upsupported model, do not add the other warnings
+          reasons.push({ content: i18n.t('upscaling.incompatibleBaseModel') });
+        } else {
+          // Using a compatible model, add all warnings
+          if (!model) {
+            reasons.push({ content: i18n.t('parameters.invoke.noModelSelected') });
+          }
+          if (!upscale.upscaleModel) {
+            reasons.push({ content: i18n.t('upscaling.missingUpscaleModel') });
+          }
+          if (!upscale.tileControlnetModel) {
+            reasons.push({ content: i18n.t('upscaling.missingTileControlNetModel') });
+          }
         }
       } else {
         if (canvasIsFiltering) {

invokeai/frontend/web/src/features/controlLayers/components/CanvasAddEntityButtons.tsx

@@ -9,7 +9,7 @@ import {
   useAddRegionalGuidance,
   useAddRegionalReferenceImage,
 } from 'features/controlLayers/hooks/addLayerHooks';
-import { selectIsFLUX } from 'features/controlLayers/store/paramsSlice';
+import { selectIsFLUX, selectIsSD3 } from 'features/controlLayers/store/paramsSlice';
 import { memo } from 'react';
 import { useTranslation } from 'react-i18next';
 import { PiPlusBold } from 'react-icons/pi';
@@ -23,6 +23,7 @@ export const CanvasAddEntityButtons = memo(() => {
   const addGlobalReferenceImage = useAddGlobalReferenceImage();
   const addRegionalReferenceImage = useAddRegionalReferenceImage();
   const isFLUX = useAppSelector(selectIsFLUX);
+  const isSD3 = useAppSelector(selectIsSD3);
 
   return (
     <Flex w="full" h="full" justifyContent="center" gap={4}>
@@ -36,6 +37,7 @@ export const CanvasAddEntityButtons = memo(() => {
               justifyContent="flex-start"
               leftIcon={<PiPlusBold />}
               onClick={addGlobalReferenceImage}
+              isDisabled={isSD3}
             >
               {t('controlLayers.globalReferenceImage')}
             </Button>
@@ -61,7 +63,7 @@ export const CanvasAddEntityButtons = memo(() => {
               justifyContent="flex-start"
               leftIcon={<PiPlusBold />}
               onClick={addRegionalGuidance}
-              isDisabled={isFLUX}
+              isDisabled={isFLUX || isSD3}
             >
               {t('controlLayers.regionalGuidance')}
             </Button>
@@ -73,7 +75,7 @@ export const CanvasAddEntityButtons = memo(() => {
               justifyContent="flex-start"
               leftIcon={<PiPlusBold />}
               onClick={addRegionalReferenceImage}
-              isDisabled={isFLUX}
+              isDisabled={isFLUX || isSD3}
             >
               {t('controlLayers.regionalReferenceImage')}
             </Button>
@@ -88,6 +90,7 @@ export const CanvasAddEntityButtons = memo(() => {
               justifyContent="flex-start"
               leftIcon={<PiPlusBold />}
               onClick={addControlLayer}
+              isDisabled={isSD3}
             >
               {t('controlLayers.controlLayer')}
             </Button>

invokeai/frontend/web/src/features/controlLayers/components/CanvasAlerts/CanvasAlertsInvocationProgress.tsx

@@ -0,0 +1,45 @@
+import { Alert, AlertDescription, AlertIcon, AlertTitle } from '@invoke-ai/ui-library';
+import { useStore } from '@nanostores/react';
+import { useAppSelector } from 'app/store/storeHooks';
+import { useFeatureStatus } from 'features/system/hooks/useFeatureStatus';
+import { selectSystemShouldShowInvocationProgressDetail } from 'features/system/store/systemSlice';
+import { memo } from 'react';
+import { useTranslation } from 'react-i18next';
+import { $invocationProgressMessage } from 'services/events/stores';
+
+const CanvasAlertsInvocationProgressContent = memo(() => {
+  const { t } = useTranslation();
+  const invocationProgressMessage = useStore($invocationProgressMessage);
+
+  if (!invocationProgressMessage) {
+    return null;
+  }
+
+  return (
+    <Alert status="loading" borderRadius="base" fontSize="sm" shadow="md" w="fit-content">
+      <AlertIcon />
+      <AlertTitle>{t('common.generating')}</AlertTitle>
+      <AlertDescription>{invocationProgressMessage}</AlertDescription>
+    </Alert>
+  );
+});
+CanvasAlertsInvocationProgressContent.displayName = 'CanvasAlertsInvocationProgressContent';
+
+export const CanvasAlertsInvocationProgress = memo(() => {
+  const isProgressMessageAlertEnabled = useFeatureStatus('invocationProgressAlert');
+  const shouldShowInvocationProgressDetail = useAppSelector(selectSystemShouldShowInvocationProgressDetail);
+
+  // The alert is disabled at the system level
+  if (!isProgressMessageAlertEnabled) {
+    return null;
+  }
+
+  // The alert is disabled at the user level
+  if (!shouldShowInvocationProgressDetail) {
+    return null;
+  }
+
+  return <CanvasAlertsInvocationProgressContent />;
+});
+
+CanvasAlertsInvocationProgress.displayName = 'CanvasAlertsInvocationProgress';

invokeai/frontend/web/src/features/controlLayers/components/CanvasEntityList/EntityListGlobalActionBarAddLayerMenu.tsx

@@ -9,7 +9,7 @@ import {
   useAddRegionalReferenceImage,
 } from 'features/controlLayers/hooks/addLayerHooks';
 import { useCanvasIsBusy } from 'features/controlLayers/hooks/useCanvasIsBusy';
-import { selectIsFLUX } from 'features/controlLayers/store/paramsSlice';
+import { selectIsFLUX, selectIsSD3 } from 'features/controlLayers/store/paramsSlice';
 import { memo } from 'react';
 import { useTranslation } from 'react-i18next';
 import { PiPlusBold } from 'react-icons/pi';
@@ -24,6 +24,7 @@ export const EntityListGlobalActionBarAddLayerMenu = memo(() => {
   const addRasterLayer = useAddRasterLayer();
   const addControlLayer = useAddControlLayer();
   const isFLUX = useAppSelector(selectIsFLUX);
+  const isSD3 = useAppSelector(selectIsSD3);
 
   return (
     <Menu>
@@ -40,7 +41,7 @@ export const EntityListGlobalActionBarAddLayerMenu = memo(() => {
       />
       <MenuList>
         <MenuGroup title={t('controlLayers.global')}>
-          <MenuItem icon={<PiPlusBold />} onClick={addGlobalReferenceImage}>
+          <MenuItem icon={<PiPlusBold />} onClick={addGlobalReferenceImage} isDisabled={isSD3}>
             {t('controlLayers.globalReferenceImage')}
           </MenuItem>
         </MenuGroup>
@@ -48,15 +49,15 @@ export const EntityListGlobalActionBarAddLayerMenu = memo(() => {
           <MenuItem icon={<PiPlusBold />} onClick={addInpaintMask}>
             {t('controlLayers.inpaintMask')}
           </MenuItem>
-          <MenuItem icon={<PiPlusBold />} onClick={addRegionalGuidance} isDisabled={isFLUX}>
+          <MenuItem icon={<PiPlusBold />} onClick={addRegionalGuidance} isDisabled={isFLUX || isSD3}>
             {t('controlLayers.regionalGuidance')}
           </MenuItem>
-          <MenuItem icon={<PiPlusBold />} onClick={addRegionalReferenceImage} isDisabled={isFLUX}>
+          <MenuItem icon={<PiPlusBold />} onClick={addRegionalReferenceImage} isDisabled={isFLUX || isSD3}>
             {t('controlLayers.regionalReferenceImage')}
           </MenuItem>
         </MenuGroup>
         <MenuGroup title={t('controlLayers.layer_other')}>
-          <MenuItem icon={<PiPlusBold />} onClick={addControlLayer}>
+          <MenuItem icon={<PiPlusBold />} onClick={addControlLayer} isDisabled={isSD3}>
             {t('controlLayers.controlLayer')}
           </MenuItem>
           <MenuItem icon={<PiPlusBold />} onClick={addRasterLayer}>

invokeai/frontend/web/src/features/controlLayers/components/CanvasEntityList/useCanvasEntityListDnd.ts

@@ -4,6 +4,7 @@ import { attachClosestEdge, extractClosestEdge } from '@atlaskit/pragmatic-drag-
 import type { CanvasEntityIdentifier } from 'features/controlLayers/store/types';
 import { singleCanvasEntityDndSource } from 'features/dnd/dnd';
 import { type DndListTargetState, idle } from 'features/dnd/types';
+import { firefoxDndFix } from 'features/dnd/util';
 import type { RefObject } from 'react';
 import { useEffect, useState } from 'react';
 
@@ -17,6 +18,7 @@ export const useCanvasEntityListDnd = (ref: RefObject<HTMLElement>, entityIdenti
       return;
     }
     return combine(
+      firefoxDndFix(element),
       draggable({
         element,
         getInitialData() {

invokeai/frontend/web/src/features/controlLayers/components/CanvasMainPanelContent.tsx

@@ -21,6 +21,8 @@ import { GatedImageViewer } from 'features/gallery/components/ImageViewer/ImageV
 import { memo, useCallback, useRef } from 'react';
 import { PiDotsThreeOutlineVerticalFill } from 'react-icons/pi';
 
+import { CanvasAlertsInvocationProgress } from './CanvasAlerts/CanvasAlertsInvocationProgress';
+
 const MenuContent = () => {
   return (
     <CanvasManagerProviderGate>
@@ -84,6 +86,7 @@ export const CanvasMainPanelContent = memo(() => {
                 <CanvasAlertsSelectedEntityStatus />
                 <CanvasAlertsPreserveMask />
                 <CanvasAlertsSendingToGallery />
+                <CanvasAlertsInvocationProgress />
               </Flex>
               <Flex position="absolute" top={1} insetInlineEnd={1}>
                 <Menu>

invokeai/frontend/web/src/features/controlLayers/components/ParamDenoisingStrength.tsx

@@ -17,12 +17,15 @@ import { selectImg2imgStrengthConfig } from 'features/system/store/configSlice';
 import { memo, useCallback } from 'react';
 import { useTranslation } from 'react-i18next';
 
-const selectIsEnabled = createSelector(selectActiveRasterLayerEntities, (entities) => entities.length > 0);
+const selectHasRasterLayersWithContent = createSelector(
+  selectActiveRasterLayerEntities,
+  (entities) => entities.length > 0
+);
 
 export const ParamDenoisingStrength = memo(() => {
   const img2imgStrength = useAppSelector(selectImg2imgStrength);
   const dispatch = useAppDispatch();
-  const isEnabled = useAppSelector(selectIsEnabled);
+  const hasRasterLayersWithContent = useAppSelector(selectHasRasterLayersWithContent);
 
   const onChange = useCallback(
     (v: number) => {
@@ -37,16 +40,16 @@ export const ParamDenoisingStrength = memo(() => {
   const [invokeBlue300] = useToken('colors', ['invokeBlue.300']);
 
   return (
-    <FormControl isDisabled={!isEnabled} p={1} justifyContent="space-between" h={8}>
+    <FormControl isDisabled={!hasRasterLayersWithContent} p={1} justifyContent="space-between" h={8}>
       <Flex gap={3} alignItems="center">
         <InformationalPopover feature="paramDenoisingStrength">
           <FormLabel mr={0}>{`${t('parameters.denoisingStrength')}`}</FormLabel>
         </InformationalPopover>
-        {isEnabled && (
+        {hasRasterLayersWithContent && (
           <WavyLine amplitude={img2imgStrength * 10} stroke={invokeBlue300} strokeWidth={1} width={40} height={14} />
         )}
       </Flex>
-      {isEnabled ? (
+      {hasRasterLayersWithContent ? (
         <>
           <CompositeSlider
             step={config.coarseStep}
@@ -70,9 +73,7 @@ export const ParamDenoisingStrength = memo(() => {
         </>
       ) : (
         <Flex alignItems="center">
-          <Badge opacity="0.6">
-            {t('common.disabled')} - {t('parameters.noRasterLayers')}
-          </Badge>
+          <Badge opacity="0.6">{t('parameters.disabledNoRasterContent')}</Badge>
         </Flex>
       )}
     </FormControl>

invokeai/frontend/web/src/features/controlLayers/components/StagingArea/StagingAreaToolbarSaveSelectedToGalleryButton.tsx

@@ -1,17 +1,19 @@
 import { IconButton } from '@invoke-ai/ui-library';
 import { useAppSelector } from 'app/store/storeHooks';
+import { withResultAsync } from 'common/util/result';
 import { selectSelectedImage } from 'features/controlLayers/store/canvasStagingAreaSlice';
 import { selectAutoAddBoardId } from 'features/gallery/store/gallerySelectors';
+import { toast } from 'features/toast/toast';
 import { memo, useCallback } from 'react';
 import { useTranslation } from 'react-i18next';
 import { PiFloppyDiskBold } from 'react-icons/pi';
-import { useAddImagesToBoardMutation, useChangeImageIsIntermediateMutation } from 'services/api/endpoints/images';
+import { uploadImage } from 'services/api/endpoints/images';
+
+const TOAST_ID = 'SAVE_STAGING_AREA_IMAGE_TO_GALLERY';
 
 export const StagingAreaToolbarSaveSelectedToGalleryButton = memo(() => {
   const autoAddBoardId = useAppSelector(selectAutoAddBoardId);
   const selectedImage = useAppSelector(selectSelectedImage);
-  const [addImageToBoard] = useAddImagesToBoardMutation();
-  const [changeIsImageIntermediate] = useChangeImageIsIntermediateMutation();
 
   const { t } = useTranslation();
 
@@ -19,21 +21,42 @@ export const StagingAreaToolbarSaveSelectedToGalleryButton = memo(() => {
     if (!selectedImage) {
       return;
     }
-    if (autoAddBoardId !== 'none') {
-      await addImageToBoard({ imageDTOs: [selectedImage.imageDTO], board_id: autoAddBoardId }).unwrap();
-      // The changeIsImageIntermediate request will use the board_id on this specific imageDTO object, so we need to
-      // update it before making the request - else the optimistic board updates will get out of whack.
-      changeIsImageIntermediate({
-        imageDTO: { ...selectedImage.imageDTO, board_id: autoAddBoardId },
+
+    // To save the image to gallery, we will download it and re-upload it. This allows the user to delete the image
+    // the gallery without borking the canvas, which may need this image to exist.
+    const result = await withResultAsync(async () => {
+      // Download the image
+      const res = await fetch(selectedImage.imageDTO.image_url);
+      const blob = await res.blob();
+      // Create a new file with the same name, which we will upload
+      const file = new File([blob], `copy_of_${selectedImage.imageDTO.image_name}`, { type: 'image/png' });
+
+      await uploadImage({
+        file,
+        // Image should show up in the Images tab
+        image_category: 'general',
         is_intermediate: false,
+        // TODO(psyche): Maybe this should just save to the currently-selected board?
+        board_id: autoAddBoardId === 'none' ? undefined : autoAddBoardId,
+        // We will do our own toast - opt out of the default handling
+        silent: true,
+      });
+    });
+
+    if (result.isOk()) {
+      toast({
+        id: TOAST_ID,
+        title: t('controlLayers.savedToGalleryOk'),
+        status: 'success',
       });
     } else {
-      changeIsImageIntermediate({
-        imageDTO: selectedImage.imageDTO,
-        is_intermediate: false,
+      toast({
+        id: TOAST_ID,
+        title: t('controlLayers.savedToGalleryError'),
+        status: 'error',
       });
     }
-  }, [addImageToBoard, autoAddBoardId, changeIsImageIntermediate, selectedImage]);
+  }, [autoAddBoardId, selectedImage, t]);
 
   return (
     <IconButton
@@ -42,7 +65,7 @@ export const StagingAreaToolbarSaveSelectedToGalleryButton = memo(() => {
       icon={<PiFloppyDiskBold />}
       onClick={saveSelectedImageToGallery}
       colorScheme="invokeBlue"
-      isDisabled={!selectedImage || !selectedImage.imageDTO.is_intermediate}
+      isDisabled={!selectedImage}
     />
   );
 });

invokeai/frontend/web/src/features/controlLayers/hooks/saveCanvasHooks.ts

@@ -25,6 +25,8 @@ import type {
   RegionalGuidanceReferenceImageState,
 } from 'features/controlLayers/store/types';
 import { imageDTOToImageObject, imageDTOToImageWithDims, initialControlNet } from 'features/controlLayers/store/util';
+import { selectAutoAddBoardId } from 'features/gallery/store/gallerySelectors';
+import type { BoardId } from 'features/gallery/store/types';
 import { toast } from 'features/toast/toast';
 import { useCallback, useMemo } from 'react';
 import { useTranslation } from 'react-i18next';
@@ -70,6 +72,11 @@ const useSaveCanvas = ({ region, saveToGallery, toastOk, toastError, onSave, wit
       metadata = selectCanvasMetadata(store.getState());
     }
 
+    let boardId: BoardId | undefined = undefined;
+    if (saveToGallery) {
+      boardId = selectAutoAddBoardId(store.getState());
+    }
+
     const result = await withResultAsync(() => {
       const rasterAdapters = canvasManager.compositor.getVisibleAdaptersOfType('raster_layer');
       return canvasManager.compositor.getCompositeImageDTO(
@@ -78,6 +85,8 @@ const useSaveCanvas = ({ region, saveToGallery, toastOk, toastError, onSave, wit
         {
           is_intermediate: !saveToGallery,
           metadata,
+          board_id: boardId,
+          silent: true,
         },
         undefined,
         true // force upload the image to ensure it gets added to the gallery
@@ -222,8 +231,8 @@ export const useNewRasterLayerFromBbox = () => {
       toastError: t('controlLayers.newRasterLayerError'),
     };
   }, [dispatch, t]);
-  const newRasterLayerFromBbox = useSaveCanvas(arg);
-  return newRasterLayerFromBbox;
+  const func = useSaveCanvas(arg);
+  return func;
 };
 
 export const useNewControlLayerFromBbox = () => {

invokeai/frontend/web/src/features/controlLayers/konva/CanvasCompositorModule.ts

@@ -28,19 +28,17 @@ import type {
 } from 'features/controlLayers/store/types';
 import { getEntityIdentifier } from 'features/controlLayers/store/types';
 import { imageDTOToImageObject } from 'features/controlLayers/store/util';
-import { selectAutoAddBoardId } from 'features/gallery/store/gallerySelectors';
 import { toast } from 'features/toast/toast';
 import { t } from 'i18next';
 import { atom, computed } from 'nanostores';
 import type { Logger } from 'roarr';
 import { serializeError } from 'serialize-error';
-import type { UploadImageArg } from 'services/api/endpoints/images';
 import { getImageDTOSafe, uploadImage } from 'services/api/endpoints/images';
-import type { ImageDTO } from 'services/api/types';
+import type { ImageDTO, UploadImageArg } from 'services/api/types';
 import stableHash from 'stable-hash';
 import type { Equals } from 'tsafe';
 import { assert } from 'tsafe';
-import type { JsonObject } from 'type-fest';
+import type { JsonObject, SetOptional } from 'type-fest';
 
 type CompositingOptions = {
   /**
@@ -259,7 +257,7 @@ export class CanvasCompositorModule extends CanvasModuleBase {
   getCompositeImageDTO = async (
     adapters: CanvasEntityAdapter[],
     rect: Rect,
-    uploadOptions: Pick<UploadImageArg, 'is_intermediate' | 'metadata'>,
+    uploadOptions: SetOptional<Omit<UploadImageArg, 'file'>, 'image_category'>,
     compositingOptions?: CompositingOptions,
     forceUpload?: boolean
   ): Promise<ImageDTO> => {
@@ -299,10 +297,7 @@ export class CanvasCompositorModule extends CanvasModuleBase {
       uploadImage({
         file: new File([blob], 'canvas-composite.png', { type: 'image/png' }),
         image_category: 'general',
-        is_intermediate: uploadOptions.is_intermediate,
-        board_id: uploadOptions.is_intermediate ? undefined : selectAutoAddBoardId(this.manager.store.getState()),
-        metadata: uploadOptions.metadata,
-        withToast: false,
+        ...uploadOptions,
       })
     );
     this.$isUploading.set(false);

invokeai/frontend/web/src/features/controlLayers/konva/CanvasEntity/CanvasEntityObjectRenderer.ts

@@ -493,7 +493,7 @@ export class CanvasEntityObjectRenderer extends CanvasModuleBase {
       file: new File([blob], `${this.id}_rasterized.png`, { type: 'image/png' }),
       image_category: 'other',
       is_intermediate: true,
-      withToast: false,
+      silent: true,
     });
     const imageObject = imageDTOToImageObject(imageDTO);
     if (replaceObjects) {

invokeai/frontend/web/src/features/controlLayers/konva/CanvasTool/CanvasBboxToolModule.ts

@@ -1,13 +1,8 @@
-import {
-  roundDownToMultiple,
-  roundToMultiple,
-  roundToMultipleMin,
-  roundUpToMultiple,
-} from 'common/util/roundDownToMultiple';
+import { roundToMultiple, roundToMultipleMin } from 'common/util/roundDownToMultiple';
 import type { CanvasManager } from 'features/controlLayers/konva/CanvasManager';
 import { CanvasModuleBase } from 'features/controlLayers/konva/CanvasModuleBase';
 import type { CanvasToolModule } from 'features/controlLayers/konva/CanvasTool/CanvasToolModule';
-import { getKonvaNodeDebugAttrs, getPrefixedId } from 'features/controlLayers/konva/util';
+import { fitRectToGrid, getKonvaNodeDebugAttrs, getPrefixedId } from 'features/controlLayers/konva/util';
 import { selectBboxOverlay } from 'features/controlLayers/store/canvasSettingsSlice';
 import { selectBbox } from 'features/controlLayers/store/selectors';
 import type { Coordinate, Rect } from 'features/controlLayers/store/types';
@@ -398,18 +393,12 @@ export class CanvasBboxToolModule extends CanvasModuleBase {
     }
 
     // Determine the bbox size that fits within the visible rect. The bbox must be at least 64px in width and height,
-    // and its width and height must be multiples of 8px.
+    // and its width and height must be multiples of the bbox grid size.
     const gridSize = this.manager.stateApi.getBboxGridSize();
 
-    // To be conservative, we will round up the x and y to the nearest grid size, and round down the width and height.
-    // This ensures the bbox is never _larger_ than the visible rect. If the bbox is larger than the visible, we
-    // will always trigger the outpainting workflow, which is not what the user wants.
-    const x = roundUpToMultiple(visibleRect.x, gridSize);
-    const y = roundUpToMultiple(visibleRect.y, gridSize);
-    const width = roundDownToMultiple(visibleRect.width, gridSize);
-    const height = roundDownToMultiple(visibleRect.height, gridSize);
+    const rect = fitRectToGrid(visibleRect, gridSize);
 
-    this.manager.stateApi.setGenerationBbox({ x, y, width, height });
+    this.manager.stateApi.setGenerationBbox(rect);
   };
 
   /**

invokeai/frontend/web/src/features/controlLayers/konva/util.test.ts

@@ -1,4 +1,6 @@
-import { getPrefixedId, getRectUnion } from 'features/controlLayers/konva/util';
+import { roundUpToMultiple } from 'common/util/roundDownToMultiple';
+import { fitRectToGrid, getPrefixedId, getRectUnion } from 'features/controlLayers/konva/util';
+import type { Rect } from 'features/controlLayers/store/types';
 import { describe, expect, it } from 'vitest';
 
 describe('util', () => {
@@ -44,4 +46,74 @@ describe('util', () => {
       expect(union).toEqual({ x: 0, y: 0, width: 0, height: 0 });
     });
   });
+
+  describe('fitRectToGrid', () => {
+    it('should fit rect within grid without exceeding bounds', () => {
+      const rect: Rect = { x: 0, y: 0, width: 1047, height: 1758 };
+      const gridSize = 50;
+      const result = fitRectToGrid(rect, gridSize);
+
+      expect(result.x).toBe(roundUpToMultiple(rect.x, gridSize));
+      expect(result.y).toBe(roundUpToMultiple(rect.y, gridSize));
+      expect(result.width).toBeLessThanOrEqual(rect.width);
+      expect(result.height).toBeLessThanOrEqual(rect.height);
+      expect(result.width % gridSize).toBe(0);
+      expect(result.height % gridSize).toBe(0);
+    });
+
+    it('should handle small rect within grid bounds', () => {
+      const rect: Rect = { x: 20, y: 30, width: 80, height: 90 };
+      const gridSize = 25;
+      const result = fitRectToGrid(rect, gridSize);
+
+      expect(result.x).toBe(25);
+      expect(result.y).toBe(50);
+      expect(result.width % gridSize).toBe(0);
+      expect(result.height % gridSize).toBe(0);
+      expect(result.width).toBeLessThanOrEqual(rect.width);
+      expect(result.height).toBeLessThanOrEqual(rect.height);
+    });
+
+    it('should handle rect starting outside of grid alignment', () => {
+      const rect: Rect = { x: 13, y: 27, width: 94, height: 112 };
+      const gridSize = 20;
+      const result = fitRectToGrid(rect, gridSize);
+
+      expect(result.x).toBe(20);
+      expect(result.y).toBe(40);
+      expect(result.width % gridSize).toBe(0);
+      expect(result.height % gridSize).toBe(0);
+      expect(result.width).toBeLessThanOrEqual(rect.width);
+      expect(result.height).toBeLessThanOrEqual(rect.height);
+    });
+
+    it('should return the same rect if already aligned to grid', () => {
+      const rect: Rect = { x: 100, y: 100, width: 200, height: 300 };
+      const gridSize = 50;
+      const result = fitRectToGrid(rect, gridSize);
+
+      expect(result).toEqual(rect);
+    });
+
+    it('should handle large grid sizes relative to rect dimensions', () => {
+      const rect: Rect = { x: 250, y: 300, width: 400, height: 500 };
+      const gridSize = 100;
+      const result = fitRectToGrid(rect, gridSize);
+
+      expect(result.x).toBe(300);
+      expect(result.y).toBe(300);
+      expect(result.width % gridSize).toBe(0);
+      expect(result.height % gridSize).toBe(0);
+      expect(result.width).toBeLessThanOrEqual(rect.width);
+      expect(result.height).toBeLessThanOrEqual(rect.height);
+    });
+
+    it('should handle rect with zero width and height', () => {
+      const rect: Rect = { x: 40, y: 60, width: 100, height: 200 };
+      const gridSize = 20;
+      const result = fitRectToGrid(rect, gridSize);
+
+      expect(result).toEqual({ x: 40, y: 60, width: 100, height: 200 });
+    });
+  });
 });

invokeai/frontend/web/src/features/controlLayers/konva/util.ts

@@ -1,5 +1,6 @@
 import type { Selector, Store } from '@reduxjs/toolkit';
 import { $authToken } from 'app/store/nanostores/authToken';
+import { roundDownToMultiple, roundUpToMultiple } from 'common/util/roundDownToMultiple';
 import type {
   CanvasEntityIdentifier,
   CanvasObjectState,
@@ -560,6 +561,33 @@ export const getRectIntersection = (...rects: Rect[]): Rect => {
   return rect || getEmptyRect();
 };
 
+/**
+ * Fits a rect to the nearest multiple of the grid size, rounding down. The returned rect will be smaller than or equal
+ * to the input rect, and will be aligned to the grid.
+ *
+ * In other words, shrink the rect inwards on each size until it fits within the visible rect and aligns to the grid.
+ *
+ * @param rect The rect to fit
+ * @param gridSize The size of the grid
+ * @returns The fitted rect
+ */
+export const fitRectToGrid = (rect: Rect, gridSize: number): Rect => {
+  // Rounding x and y up effectively shrinks the left and top edges of the rect, and rounding width and height down
+  // effectively shrinks the right and bottom edges.
+  const x = roundUpToMultiple(rect.x, gridSize);
+  const y = roundUpToMultiple(rect.y, gridSize);
+
+  // Because we've just shifted the rect's x and y, we need to adjust the width and height by the same amount before
+  // we round those values down.
+  const offsetX = x - rect.x;
+  const offsetY = y - rect.y;
+
+  const width = roundDownToMultiple(rect.width - offsetX, gridSize);
+  const height = roundDownToMultiple(rect.height - offsetY, gridSize);
+
+  return { x, y, width, height };
+};
+
 /**
  * Asserts that the value is never reached. Used for exhaustive checks in switch statements or conditional logic to ensure
  * that all possible values are handled.

invokeai/frontend/web/src/features/controlLayers/store/canvasSettingsSlice.ts

@@ -90,7 +90,7 @@ const initialState: CanvasSettingsState = {
   invertScrollForToolWidth: false,
   color: { r: 31, g: 160, b: 224, a: 1 }, // invokeBlue.500
   sendToCanvas: false,
-  outputOnlyMaskedRegions: false,
+  outputOnlyMaskedRegions: true,
   autoProcess: true,
   snapToGrid: true,
   showProgressOnCanvas: true,

invokeai/frontend/web/src/features/controlLayers/store/paramsSlice.ts

@@ -9,6 +9,8 @@ import type {
   ParameterCFGRescaleMultiplier,
   ParameterCFGScale,
   ParameterCLIPEmbedModel,
+  ParameterCLIPGEmbedModel,
+  ParameterCLIPLEmbedModel,
   ParameterGuidance,
   ParameterMaskBlurMethod,
   ParameterModel,
@@ -71,6 +73,8 @@ export type ParamsState = {
   refinerStart: number;
   t5EncoderModel: ParameterT5EncoderModel | null;
   clipEmbedModel: ParameterCLIPEmbedModel | null;
+  clipLEmbedModel: ParameterCLIPLEmbedModel | null;
+  clipGEmbedModel: ParameterCLIPGEmbedModel | null;
 };
 
 const initialState: ParamsState = {
@@ -115,6 +119,8 @@ const initialState: ParamsState = {
   refinerStart: 0.8,
   t5EncoderModel: null,
   clipEmbedModel: null,
+  clipLEmbedModel: null,
+  clipGEmbedModel: null,
 };
 
 export const paramsSlice = createSlice({
@@ -192,6 +198,12 @@ export const paramsSlice = createSlice({
     clipEmbedModelSelected: (state, action: PayloadAction<ParameterCLIPEmbedModel | null>) => {
       state.clipEmbedModel = action.payload;
     },
+    clipLEmbedModelSelected: (state, action: PayloadAction<ParameterCLIPLEmbedModel | null>) => {
+      state.clipLEmbedModel = action.payload;
+    },
+    clipGEmbedModelSelected: (state, action: PayloadAction<ParameterCLIPGEmbedModel | null>) => {
+      state.clipGEmbedModel = action.payload;
+    },
     vaePrecisionChanged: (state, action: PayloadAction<ParameterPrecision>) => {
       state.vaePrecision = action.payload;
     },
@@ -305,6 +317,8 @@ export const {
   vaePrecisionChanged,
   t5EncoderModelSelected,
   clipEmbedModelSelected,
+  clipLEmbedModelSelected,
+  clipGEmbedModelSelected,
   setClipSkip,
   shouldUseCpuNoiseChanged,
   positivePromptChanged,
@@ -341,6 +355,7 @@ export const createParamsSelector = <T>(selector: Selector<ParamsState, T>) =>
 export const selectBase = createParamsSelector((params) => params.model?.base);
 export const selectIsSDXL = createParamsSelector((params) => params.model?.base === 'sdxl');
 export const selectIsFLUX = createParamsSelector((params) => params.model?.base === 'flux');
+export const selectIsSD3 = createParamsSelector((params) => params.model?.base === 'sd-3');
 
 export const selectModel = createParamsSelector((params) => params.model);
 export const selectModelKey = createParamsSelector((params) => params.model?.key);
@@ -349,6 +364,10 @@ export const selectFLUXVAE = createParamsSelector((params) => params.fluxVAE);
 export const selectVAEKey = createParamsSelector((params) => params.vae?.key);
 export const selectT5EncoderModel = createParamsSelector((params) => params.t5EncoderModel);
 export const selectCLIPEmbedModel = createParamsSelector((params) => params.clipEmbedModel);
+export const selectCLIPLEmbedModel = createParamsSelector((params) => params.clipLEmbedModel);
+
+export const selectCLIPGEmbedModel = createParamsSelector((params) => params.clipGEmbedModel);
+
 export const selectCFGScale = createParamsSelector((params) => params.cfgScale);
 export const selectGuidance = createParamsSelector((params) => params.guidance);
 export const selectSteps = createParamsSelector((params) => params.steps);

invokeai/frontend/web/src/features/dnd/DndImage.tsx

@@ -1,3 +1,4 @@
+import { combine } from '@atlaskit/pragmatic-drag-and-drop/combine';
 import { draggable } from '@atlaskit/pragmatic-drag-and-drop/element/adapter';
 import type { ImageProps, SystemStyleObject } from '@invoke-ai/ui-library';
 import { Image } from '@invoke-ai/ui-library';
@@ -5,6 +6,7 @@ import { useAppStore } from 'app/store/nanostores/store';
 import { singleImageDndSource } from 'features/dnd/dnd';
 import type { DndDragPreviewSingleImageState } from 'features/dnd/DndDragPreviewSingleImage';
 import { createSingleImageDragPreview, setSingleImageDragPreview } from 'features/dnd/DndDragPreviewSingleImage';
+import { firefoxDndFix } from 'features/dnd/util';
 import { useImageContextMenu } from 'features/gallery/components/ImageContextMenu/ImageContextMenu';
 import { memo, useEffect, useState } from 'react';
 import type { ImageDTO } from 'services/api/types';
@@ -35,25 +37,28 @@ export const DndImage = memo(({ imageDTO, asThumbnail, ...rest }: Props) => {
     if (!element) {
       return;
     }
-    return draggable({
-      element,
-      getInitialData: () => singleImageDndSource.getData({ imageDTO }, imageDTO.image_name),
-      onDragStart: () => {
-        setIsDragging(true);
-      },
-      onDrop: () => {
-        setIsDragging(false);
-      },
-      onGenerateDragPreview: (args) => {
-        if (singleImageDndSource.typeGuard(args.source.data)) {
-          setSingleImageDragPreview({
-            singleImageDndData: args.source.data,
-            onGenerateDragPreviewArgs: args,
-            setDragPreviewState,
-          });
-        }
-      },
-    });
+    return combine(
+      firefoxDndFix(element),
+      draggable({
+        element,
+        getInitialData: () => singleImageDndSource.getData({ imageDTO }, imageDTO.image_name),
+        onDragStart: () => {
+          setIsDragging(true);
+        },
+        onDrop: () => {
+          setIsDragging(false);
+        },
+        onGenerateDragPreview: (args) => {
+          if (singleImageDndSource.typeGuard(args.source.data)) {
+            setSingleImageDragPreview({
+              singleImageDndData: args.source.data,
+              onGenerateDragPreviewArgs: args,
+              setDragPreviewState,
+            });
+          }
+        },
+      })
+    );
   }, [imageDTO, element, store]);
 
   useImageContextMenu(imageDTO, element);

invokeai/frontend/web/src/features/dnd/FullscreenDropzone.tsx

@@ -11,10 +11,11 @@ import type { DndTargetState } from 'features/dnd/types';
 import { selectAutoAddBoardId } from 'features/gallery/store/gallerySelectors';
 import { selectMaxImageUploadCount } from 'features/system/store/configSlice';
 import { toast } from 'features/toast/toast';
-import { memo, useEffect, useRef, useState } from 'react';
+import { memo, useCallback, useEffect, useMemo, useRef, useState } from 'react';
 import { useTranslation } from 'react-i18next';
-import { type UploadImageArg, uploadImages } from 'services/api/endpoints/images';
+import { uploadImages } from 'services/api/endpoints/images';
 import { useBoardName } from 'services/api/hooks/useBoardName';
+import type { UploadImageArg } from 'services/api/types';
 import { z } from 'zod';
 
 const ACCEPTED_IMAGE_TYPES = ['image/png', 'image/jpg', 'image/jpeg'];
@@ -71,13 +72,47 @@ export const FullscreenDropzone = memo(() => {
   const maxImageUploadCount = useAppSelector(selectMaxImageUploadCount);
   const [dndState, setDndState] = useState<DndTargetState>('idle');
 
+  const uploadFilesSchema = useMemo(() => getFilesSchema(maxImageUploadCount), [maxImageUploadCount]);
+
+  const validateAndUploadFiles = useCallback(
+    (files: File[]) => {
+      const { getState } = getStore();
+      const parseResult = uploadFilesSchema.safeParse(files);
+
+      if (!parseResult.success) {
+        const description =
+          maxImageUploadCount === undefined
+            ? t('toast.uploadFailedInvalidUploadDesc')
+            : t('toast.uploadFailedInvalidUploadDesc_withCount', { count: maxImageUploadCount });
+
+        toast({
+          id: 'UPLOAD_FAILED',
+          title: t('toast.uploadFailed'),
+          description,
+          status: 'error',
+        });
+        return;
+      }
+      const autoAddBoardId = selectAutoAddBoardId(getState());
+
+      const uploadArgs: UploadImageArg[] = files.map((file, i) => ({
+        file,
+        image_category: 'user',
+        is_intermediate: false,
+        board_id: autoAddBoardId === 'none' ? undefined : autoAddBoardId,
+        isFirstUploadOfBatch: i === 0,
+      }));
+
+      uploadImages(uploadArgs);
+    },
+    [maxImageUploadCount, t, uploadFilesSchema]
+  );
+
   useEffect(() => {
     const element = ref.current;
     if (!element) {
       return;
     }
-    const { getState } = getStore();
-    const uploadFilesSchema = getFilesSchema(maxImageUploadCount);
 
     return combine(
       dropTargetForExternal({
@@ -85,32 +120,7 @@ export const FullscreenDropzone = memo(() => {
         canDrop: containsFiles,
         onDrop: ({ source }) => {
           const files = getFiles({ source });
-          const parseResult = uploadFilesSchema.safeParse(files);
-
-          if (!parseResult.success) {
-            const description =
-              maxImageUploadCount === undefined
-                ? t('toast.uploadFailedInvalidUploadDesc')
-                : t('toast.uploadFailedInvalidUploadDesc_withCount', { count: maxImageUploadCount });
-
-            toast({
-              id: 'UPLOAD_FAILED',
-              title: t('toast.uploadFailed'),
-              description,
-              status: 'error',
-            });
-            return;
-          }
-          const autoAddBoardId = selectAutoAddBoardId(getState());
-
-          const uploadArgs: UploadImageArg[] = files.map((file) => ({
-            file,
-            image_category: 'user',
-            is_intermediate: false,
-            board_id: autoAddBoardId === 'none' ? undefined : autoAddBoardId,
-          }));
-
-          uploadImages(uploadArgs);
+          validateAndUploadFiles(files);
         },
         onDragEnter: () => {
           setDndState('over');
@@ -131,7 +141,27 @@ export const FullscreenDropzone = memo(() => {
         },
       })
     );
-  }, [maxImageUploadCount, t]);
+  }, [validateAndUploadFiles]);
+
+  useEffect(() => {
+    const controller = new AbortController();
+
+    document.addEventListener(
+      'paste',
+      (e) => {
+        if (!e.clipboardData?.files) {
+          return;
+        }
+        const files = Array.from(e.clipboardData.files);
+        validateAndUploadFiles(files);
+      },
+      { signal: controller.signal }
+    );
+
+    return () => {
+      controller.abort();
+    };
+  }, [validateAndUploadFiles]);
 
   return (
     <Box ref={ref} data-dnd-state={dndState} sx={sx}>

invokeai/frontend/web/src/features/dnd/util.ts

@@ -1,6 +1,7 @@
 import type { GetOffsetFn } from '@atlaskit/pragmatic-drag-and-drop/dist/types/public-utils/element/custom-native-drag-preview/types';
 import type { Input } from '@atlaskit/pragmatic-drag-and-drop/types';
 import type { SystemStyleObject } from '@invoke-ai/ui-library';
+import { noop } from 'lodash-es';
 import type { CSSProperties } from 'react';
 
 /**
@@ -44,3 +45,67 @@ export function triggerPostMoveFlash(element: HTMLElement, backgroundColor: CSSP
     iterations: 1,
   });
 }
+
+/**
+ * Firefox has a bug where input or textarea elements with draggable parents do not allow selection of their text.
+ *
+ * This helper function implements a workaround by setting the draggable attribute to false when the mouse is over a
+ * input or textarea child of the draggable. It reverts the attribute on mouse out.
+ *
+ * The fix is only applied for Firefox, and should be used in every `pragmatic-drag-and-drop` `draggable`.
+ *
+ * See:
+ * - https://github.com/atlassian/pragmatic-drag-and-drop/issues/111
+ * - https://bugzilla.mozilla.org/show_bug.cgi?id=1853069
+ *
+ * @example
+ * ```tsx
+ * useEffect(() => {
+ *   const element = ref.current;
+ *   if (!element) {
+ *     return;
+ *   }
+ *   return combine(
+ *     firefoxDndFix(element),
+ *     // The rest of the draggable setup is the same
+ *     draggable({
+ *       element,
+ *       // ...
+ *     }),
+ *   );
+ *```
+ * @param element The draggable element
+ * @returns A cleanup function that removes the event listeners
+ */
+export const firefoxDndFix = (element: HTMLElement): (() => void) => {
+  if (!navigator.userAgent.includes('Firefox')) {
+    return noop;
+  }
+
+  const abortController = new AbortController();
+
+  element.addEventListener(
+    'mouseover',
+    (event) => {
+      if (event.target instanceof HTMLTextAreaElement || event.target instanceof HTMLInputElement) {
+        element.setAttribute('draggable', 'false');
+      }
+    },
+    { signal: abortController.signal }
+  );
+
+  element.addEventListener(
+    'mouseout',
+    (event) => {
+      if (event.target instanceof HTMLTextAreaElement || event.target instanceof HTMLInputElement) {
+        element.setAttribute('draggable', 'true');
+      }
+    },
+    { signal: abortController.signal }
+  );
+
+  return () => {
+    element.setAttribute('draggable', 'true');
+    abortController.abort();
+  };
+};

invokeai/frontend/web/src/features/gallery/components/ImageContextMenu/ImageMenuItemNewFromImageSubMenu.tsx

@@ -1,8 +1,10 @@
 import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
 import { useAppStore } from 'app/store/nanostores/store';
+import { useAppSelector } from 'app/store/storeHooks';
 import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
 import { NewLayerIcon } from 'features/controlLayers/components/common/icons';
 import { useCanvasIsBusy } from 'features/controlLayers/hooks/useCanvasIsBusy';
+import { selectIsFLUX, selectIsSD3 } from 'features/controlLayers/store/paramsSlice';
 import { useImageViewer } from 'features/gallery/components/ImageViewer/useImageViewer';
 import { useImageDTOContext } from 'features/gallery/contexts/ImageDTOContext';
 import { sentImageToCanvas } from 'features/gallery/store/actions';
@@ -20,6 +22,8 @@ export const ImageMenuItemNewFromImageSubMenu = memo(() => {
   const imageDTO = useImageDTOContext();
   const imageViewer = useImageViewer();
   const isBusy = useCanvasIsBusy();
+  const isFLUX = useAppSelector(selectIsFLUX);
+  const isSD3 = useAppSelector(selectIsSD3);
 
   const onClickNewCanvasWithRasterLayerFromImage = useCallback(() => {
     const { dispatch, getState } = store;
@@ -110,17 +114,25 @@ export const ImageMenuItemNewFromImageSubMenu = memo(() => {
           <MenuItem
             icon={<PiFileBold />}
             onClickCapture={onClickNewCanvasWithControlLayerFromImage}
-            isDisabled={isBusy}
+            isDisabled={isBusy || isSD3}
           >
             {t('controlLayers.canvasAsControlLayer')}
           </MenuItem>
           <MenuItem icon={<NewLayerIcon />} onClickCapture={onClickNewInpaintMaskFromImage} isDisabled={isBusy}>
             {t('controlLayers.inpaintMask')}
           </MenuItem>
-          <MenuItem icon={<NewLayerIcon />} onClickCapture={onClickNewRegionalGuidanceFromImage} isDisabled={isBusy}>
+          <MenuItem
+            icon={<NewLayerIcon />}
+            onClickCapture={onClickNewRegionalGuidanceFromImage}
+            isDisabled={isBusy || isFLUX || isSD3}
+          >
             {t('controlLayers.regionalGuidance')}
           </MenuItem>
-          <MenuItem icon={<NewLayerIcon />} onClickCapture={onClickNewControlLayerFromImage} isDisabled={isBusy}>
+          <MenuItem
+            icon={<NewLayerIcon />}
+            onClickCapture={onClickNewControlLayerFromImage}
+            isDisabled={isBusy || isSD3}
+          >
             {t('controlLayers.controlLayer')}
           </MenuItem>
           <MenuItem icon={<NewLayerIcon />} onClickCapture={onClickNewRasterLayerFromImage} isDisabled={isBusy}>

invokeai/frontend/web/src/features/gallery/components/ImageGrid/GalleryImage.tsx

@@ -12,6 +12,7 @@ import type { DndDragPreviewMultipleImageState } from 'features/dnd/DndDragPrevi
 import { createMultipleImageDragPreview, setMultipleImageDragPreview } from 'features/dnd/DndDragPreviewMultipleImage';
 import type { DndDragPreviewSingleImageState } from 'features/dnd/DndDragPreviewSingleImage';
 import { createSingleImageDragPreview, setSingleImageDragPreview } from 'features/dnd/DndDragPreviewSingleImage';
+import { firefoxDndFix } from 'features/dnd/util';
 import { useImageContextMenu } from 'features/gallery/components/ImageContextMenu/ImageContextMenu';
 import { GalleryImageHoverIcons } from 'features/gallery/components/ImageGrid/GalleryImageHoverIcons';
 import { getGalleryImageDataTestId } from 'features/gallery/components/ImageGrid/getGalleryImageDataTestId';
@@ -66,7 +67,7 @@ const galleryImageContainerSX = {
     },
     '&:hover::before': {
       boxShadow:
-        'inset 0px 0px 0px 2px var(--invoke-colors-invokeBlue-300), inset 0px 0px 0px 3px var(--invoke-colors-invokeBlue-800)',
+        'inset 0px 0px 0px 1px var(--invoke-colors-invokeBlue-300), inset 0px 0px 0px 2px var(--invoke-colors-invokeBlue-800)',
     },
     '&:hover[data-selected=true]::before': {
       boxShadow:
@@ -115,13 +116,17 @@ export const GalleryImage = memo(({ imageDTO }: Props) => {
       return;
     }
     return combine(
+      firefoxDndFix(element),
       draggable({
         element,
         getInitialData: () => {
           const { gallery } = store.getState();
           // When we have multiple images selected, and the dragged image is part of the selection, initiate a
           // multi-image drag.
-          if (gallery.selection.length > 1 && gallery.selection.includes(imageDTO)) {
+          if (
+            gallery.selection.length > 1 &&
+            gallery.selection.find(({ image_name }) => image_name === imageDTO.image_name) !== undefined
+          ) {
             return multipleImageDndSource.getData({
               imageDTOs: gallery.selection,
               boardId: gallery.selectedBoardId,

invokeai/frontend/web/src/features/gallery/components/ImageViewer/CurrentImagePreview.tsx

@@ -2,6 +2,7 @@ import { Box, Flex } from '@invoke-ai/ui-library';
 import { useStore } from '@nanostores/react';
 import { skipToken } from '@reduxjs/toolkit/query';
 import { useAppSelector } from 'app/store/storeHooks';
+import { CanvasAlertsInvocationProgress } from 'features/controlLayers/components/CanvasAlerts/CanvasAlertsInvocationProgress';
 import { CanvasAlertsSendingToCanvas } from 'features/controlLayers/components/CanvasAlerts/CanvasAlertsSendingTo';
 import { DndImage } from 'features/dnd/DndImage';
 import ImageMetadataViewer from 'features/gallery/components/ImageMetadataViewer/ImageMetadataViewer';
@@ -48,9 +49,18 @@ const CurrentImagePreview = () => {
       position="relative"
     >
       <ImageContent imageDTO={imageDTO} />
-      <Box position="absolute" top={0} insetInlineStart={0}>
+      <Flex
+        flexDir="column"
+        gap={2}
+        position="absolute"
+        top={0}
+        insetInlineStart={0}
+        pointerEvents="none"
+        alignItems="flex-start"
+      >
         <CanvasAlertsSendingToCanvas />
-      </Box>
+        <CanvasAlertsInvocationProgress />
+      </Flex>
       {shouldShowImageDetails && imageDTO && (
         <Box position="absolute" opacity={0.8} top={0} width="full" height="full" borderRadius="base">
           <ImageMetadataViewer image={imageDTO} />

invokeai/frontend/web/src/features/gallery/components/ImageViewer/ImageViewer.tsx

@@ -44,7 +44,6 @@ export const ImageViewer = memo(({ closeButton }: Props) => {
       right={0}
       bottom={0}
       left={0}
-      rowGap={2}
       alignItems="center"
       justifyContent="center"
     >

invokeai/frontend/web/src/features/modelManagerV2/subpanels/ModelPanel/Fields/BaseModelSelect.tsx

@@ -10,6 +10,8 @@ import type { UpdateModelArg } from 'services/api/endpoints/models';
 const options: ComboboxOption[] = [
   { value: 'sd-1', label: MODEL_TYPE_MAP['sd-1'] },
   { value: 'sd-2', label: MODEL_TYPE_MAP['sd-2'] },
+  { value: 'sd-3', label: MODEL_TYPE_MAP['sd-3'] },
+  { value: 'flux', label: MODEL_TYPE_MAP['flux'] },
   { value: 'sdxl', label: MODEL_TYPE_MAP['sdxl'] },
   { value: 'sdxl-refiner', label: MODEL_TYPE_MAP['sdxl-refiner'] },
 ];

invokeai/frontend/web/src/features/nodes/components/sidePanel/workflow/useLinearViewFieldDnd.ts

@@ -4,6 +4,7 @@ import { attachClosestEdge, extractClosestEdge } from '@atlaskit/pragmatic-drag-
 import { singleWorkflowFieldDndSource } from 'features/dnd/dnd';
 import type { DndListTargetState } from 'features/dnd/types';
 import { idle } from 'features/dnd/types';
+import { firefoxDndFix } from 'features/dnd/util';
 import type { FieldIdentifier } from 'features/nodes/types/field';
 import type { RefObject } from 'react';
 import { useEffect, useState } from 'react';
@@ -18,6 +19,7 @@ export const useLinearViewFieldDnd = (ref: RefObject<HTMLElement>, fieldIdentifi
       return;
     }
     return combine(
+      firefoxDndFix(element),
       draggable({
         element,
         getInitialData() {

invokeai/frontend/web/src/features/nodes/util/graph/buildLinearBatchConfig.ts

@@ -9,8 +9,8 @@ export const prepareLinearUIBatch = (
   state: RootState,
   g: Graph,
   prepend: boolean,
-  noise: Invocation<'noise' | 'flux_denoise'>,
-  posCond: Invocation<'compel' | 'sdxl_compel_prompt' | 'flux_text_encoder'>,
+  noise: Invocation<'noise' | 'flux_denoise' | 'sd3_denoise'>,
+  posCond: Invocation<'compel' | 'sdxl_compel_prompt' | 'flux_text_encoder' | 'sd3_text_encoder'>,
   origin: 'canvas' | 'workflows' | 'upscaling',
   destination: 'canvas' | 'gallery'
 ): BatchConfig => {

invokeai/frontend/web/src/features/nodes/util/graph/generation/addImageToImage.ts

@@ -2,16 +2,18 @@ import type { CanvasManager } from 'features/controlLayers/konva/CanvasManager';
 import { getPrefixedId } from 'features/controlLayers/konva/util';
 import type { CanvasState, Dimensions } from 'features/controlLayers/store/types';
 import type { Graph } from 'features/nodes/util/graph/generation/Graph';
-import { addImageToLatents } from 'features/nodes/util/graph/graphBuilderUtils';
 import { isEqual } from 'lodash-es';
 import type { Invocation } from 'services/api/types';
 
 type AddImageToImageArg = {
   g: Graph;
   manager: CanvasManager;
-  l2i: Invocation<'l2i' | 'flux_vae_decode'>;
-  denoise: Invocation<'denoise_latents' | 'flux_denoise'>;
-  vaeSource: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'seamless' | 'vae_loader'>;
+  l2i: Invocation<'l2i' | 'flux_vae_decode' | 'sd3_l2i'>;
+  i2lNodeType: 'i2l' | 'flux_vae_encode' | 'sd3_i2l';
+  denoise: Invocation<'denoise_latents' | 'flux_denoise' | 'sd3_denoise'>;
+  vaeSource: Invocation<
+    'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'seamless' | 'vae_loader' | 'sd3_model_loader'
+  >;
   originalSize: Dimensions;
   scaledSize: Dimensions;
   bbox: CanvasState['bbox'];
@@ -23,6 +25,7 @@ export const addImageToImage = async ({
   g,
   manager,
   l2i,
+  i2lNodeType,
   denoise,
   vaeSource,
   originalSize,
@@ -30,10 +33,13 @@ export const addImageToImage = async ({
   bbox,
   denoising_start,
   fp32,
-}: AddImageToImageArg): Promise<Invocation<'img_resize' | 'l2i' | 'flux_vae_decode'>> => {
+}: AddImageToImageArg): Promise<Invocation<'img_resize' | 'l2i' | 'flux_vae_decode' | 'sd3_l2i'>> => {
   denoise.denoising_start = denoising_start;
   const adapters = manager.compositor.getVisibleAdaptersOfType('raster_layer');
-  const { image_name } = await manager.compositor.getCompositeImageDTO(adapters, bbox.rect, { is_intermediate: true });
+  const { image_name } = await manager.compositor.getCompositeImageDTO(adapters, bbox.rect, {
+    is_intermediate: true,
+    silent: true,
+  });
 
   if (!isEqual(scaledSize, originalSize)) {
     // Resize the initial image to the scaled size, denoise, then resize back to the original size
@@ -44,7 +50,12 @@ export const addImageToImage = async ({
       ...scaledSize,
     });
 
-    const i2l = addImageToLatents(g, l2i.type === 'flux_vae_decode', fp32);
+    const i2l = g.addNode({
+      id: i2lNodeType,
+      type: i2lNodeType,
+      image: image_name ? { image_name } : undefined,
+      ...(i2lNodeType === 'i2l' ? { fp32 } : {}),
+    });
 
     const resizeImageToOriginalSize = g.addNode({
       type: 'img_resize',
@@ -61,7 +72,12 @@ export const addImageToImage = async ({
     return resizeImageToOriginalSize;
   } else {
     // No need to resize, just decode
-    const i2l = addImageToLatents(g, l2i.type === 'flux_vae_decode', fp32, image_name);
+    const i2l = g.addNode({
+      id: i2lNodeType,
+      type: i2lNodeType,
+      image: image_name ? { image_name } : undefined,
+      ...(i2lNodeType === 'i2l' ? { fp32 } : {}),
+    });
     g.addEdge(vaeSource, 'vae', i2l, 'vae');
     g.addEdge(i2l, 'latents', denoise, 'latents');
     return l2i;

invokeai/frontend/web/src/features/nodes/util/graph/generation/addInpaint.ts

@@ -6,7 +6,6 @@ import { selectParamsSlice } from 'features/controlLayers/store/paramsSlice';
 import { selectCanvasSlice } from 'features/controlLayers/store/selectors';
 import type { Dimensions } from 'features/controlLayers/store/types';
 import type { Graph } from 'features/nodes/util/graph/generation/Graph';
-import { addImageToLatents } from 'features/nodes/util/graph/graphBuilderUtils';
 import { isEqual } from 'lodash-es';
 import type { Invocation } from 'services/api/types';
 
@@ -14,10 +13,13 @@ type AddInpaintArg = {
   state: RootState;
   g: Graph;
   manager: CanvasManager;
-  l2i: Invocation<'l2i' | 'flux_vae_decode'>;
-  denoise: Invocation<'denoise_latents' | 'flux_denoise'>;
-  vaeSource: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'seamless' | 'vae_loader'>;
-  modelLoader: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader'>;
+  l2i: Invocation<'l2i' | 'flux_vae_decode' | 'sd3_l2i'>;
+  i2lNodeType: 'i2l' | 'flux_vae_encode' | 'sd3_i2l';
+  denoise: Invocation<'denoise_latents' | 'flux_denoise' | 'sd3_denoise'>;
+  vaeSource: Invocation<
+    'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'seamless' | 'vae_loader' | 'sd3_model_loader'
+  >;
+  modelLoader: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'sd3_model_loader'>;
   originalSize: Dimensions;
   scaledSize: Dimensions;
   denoising_start: number;
@@ -29,6 +31,7 @@ export const addInpaint = async ({
   g,
   manager,
   l2i,
+  i2lNodeType,
   denoise,
   vaeSource,
   modelLoader,
@@ -48,16 +51,23 @@ export const addInpaint = async ({
   const rasterAdapters = manager.compositor.getVisibleAdaptersOfType('raster_layer');
   const initialImage = await manager.compositor.getCompositeImageDTO(rasterAdapters, bbox.rect, {
     is_intermediate: true,
+    silent: true,
   });
 
   const inpaintMaskAdapters = manager.compositor.getVisibleAdaptersOfType('inpaint_mask');
   const maskImage = await manager.compositor.getCompositeImageDTO(inpaintMaskAdapters, bbox.rect, {
     is_intermediate: true,
+    silent: true,
   });
 
   if (!isEqual(scaledSize, originalSize)) {
     // Scale before processing requires some resizing
-    const i2l = addImageToLatents(g, modelLoader.type === 'flux_model_loader', fp32, initialImage.image_name);
+    const i2l = g.addNode({
+      id: i2lNodeType,
+      type: i2lNodeType,
+      image: initialImage.image_name ? { image_name: initialImage.image_name } : undefined,
+      ...(i2lNodeType === 'i2l' ? { fp32 } : {}),
+    });
 
     const resizeImageToScaledSize = g.addNode({
       type: 'img_resize',
@@ -102,7 +112,7 @@ export const addInpaint = async ({
     g.addEdge(vaeSource, 'vae', i2l, 'vae');
 
     g.addEdge(vaeSource, 'vae', createGradientMask, 'vae');
-    if (modelLoader.type !== 'flux_model_loader') {
+    if (modelLoader.type !== 'flux_model_loader' && modelLoader.type !== 'sd3_model_loader') {
       g.addEdge(modelLoader, 'unet', createGradientMask, 'unet');
     }
     g.addEdge(resizeImageToScaledSize, 'image', createGradientMask, 'image');
@@ -126,7 +136,12 @@ export const addInpaint = async ({
     return resizeOutput;
   } else {
     // No scale before processing, much simpler
-    const i2l = addImageToLatents(g, modelLoader.type === 'flux_model_loader', fp32, initialImage.image_name);
+    const i2l = g.addNode({
+      id: i2lNodeType,
+      type: i2lNodeType,
+      image: initialImage.image_name ? { image_name: initialImage.image_name } : undefined,
+      ...(i2lNodeType === 'i2l' ? { fp32 } : {}),
+    });
 
     const alphaToMask = g.addNode({
       id: getPrefixedId('alpha_to_mask'),
@@ -153,7 +168,7 @@ export const addInpaint = async ({
     g.addEdge(i2l, 'latents', denoise, 'latents');
     g.addEdge(vaeSource, 'vae', i2l, 'vae');
     g.addEdge(vaeSource, 'vae', createGradientMask, 'vae');
-    if (modelLoader.type !== 'flux_model_loader') {
+    if (modelLoader.type !== 'flux_model_loader' && modelLoader.type !== 'sd3_model_loader') {
       g.addEdge(modelLoader, 'unet', createGradientMask, 'unet');
     }
     g.addEdge(createGradientMask, 'denoise_mask', denoise, 'denoise_mask');

invokeai/frontend/web/src/features/nodes/util/graph/generation/addNSFWChecker.ts

@@ -11,7 +11,7 @@ import type { Invocation } from 'services/api/types';
 export const addNSFWChecker = (
   g: Graph,
   imageOutput: Invocation<
-    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode'
+    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode' | 'sd3_l2i'
   >
 ): Invocation<'img_nsfw'> => {
   const nsfw = g.addNode({

invokeai/frontend/web/src/features/nodes/util/graph/generation/addOutpaint.ts

@@ -6,7 +6,7 @@ import { selectParamsSlice } from 'features/controlLayers/store/paramsSlice';
 import { selectCanvasSlice } from 'features/controlLayers/store/selectors';
 import type { Dimensions } from 'features/controlLayers/store/types';
 import type { Graph } from 'features/nodes/util/graph/generation/Graph';
-import { addImageToLatents, getInfill } from 'features/nodes/util/graph/graphBuilderUtils';
+import { getInfill } from 'features/nodes/util/graph/graphBuilderUtils';
 import { isEqual } from 'lodash-es';
 import type { Invocation } from 'services/api/types';
 
@@ -14,10 +14,13 @@ type AddOutpaintArg = {
   state: RootState;
   g: Graph;
   manager: CanvasManager;
-  l2i: Invocation<'l2i' | 'flux_vae_decode'>;
-  denoise: Invocation<'denoise_latents' | 'flux_denoise'>;
-  vaeSource: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'seamless' | 'vae_loader'>;
-  modelLoader: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader'>;
+  l2i: Invocation<'l2i' | 'flux_vae_decode' | 'sd3_l2i'>;
+  i2lNodeType: 'i2l' | 'flux_vae_encode' | 'sd3_i2l';
+  denoise: Invocation<'denoise_latents' | 'flux_denoise' | 'sd3_denoise'>;
+  vaeSource: Invocation<
+    'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'seamless' | 'vae_loader' | 'sd3_model_loader'
+  >;
+  modelLoader: Invocation<'main_model_loader' | 'sdxl_model_loader' | 'flux_model_loader' | 'sd3_model_loader'>;
   originalSize: Dimensions;
   scaledSize: Dimensions;
   denoising_start: number;
@@ -29,6 +32,7 @@ export const addOutpaint = async ({
   g,
   manager,
   l2i,
+  i2lNodeType,
   denoise,
   vaeSource,
   modelLoader,
@@ -48,11 +52,13 @@ export const addOutpaint = async ({
   const rasterAdapters = manager.compositor.getVisibleAdaptersOfType('raster_layer');
   const initialImage = await manager.compositor.getCompositeImageDTO(rasterAdapters, bbox.rect, {
     is_intermediate: true,
+    silent: true,
   });
 
   const inpaintMaskAdapters = manager.compositor.getVisibleAdaptersOfType('inpaint_mask');
   const maskImage = await manager.compositor.getCompositeImageDTO(inpaintMaskAdapters, bbox.rect, {
     is_intermediate: true,
+    silent: true,
   });
 
   const infill = getInfill(g, params);
@@ -108,14 +114,19 @@ export const addOutpaint = async ({
     g.addEdge(infill, 'image', createGradientMask, 'image');
     g.addEdge(resizeInputMaskToScaledSize, 'image', createGradientMask, 'mask');
     g.addEdge(vaeSource, 'vae', createGradientMask, 'vae');
-    if (modelLoader.type !== 'flux_model_loader') {
+    if (modelLoader.type !== 'flux_model_loader' && modelLoader.type !== 'sd3_model_loader') {
       g.addEdge(modelLoader, 'unet', createGradientMask, 'unet');
     }
 
     g.addEdge(createGradientMask, 'denoise_mask', denoise, 'denoise_mask');
 
     // Decode infilled image and connect to denoise
-    const i2l = addImageToLatents(g, modelLoader.type === 'flux_model_loader', fp32);
+    const i2l = g.addNode({
+      id: i2lNodeType,
+      type: i2lNodeType,
+      ...(i2lNodeType === 'i2l' ? { fp32 } : {}),
+    });
+
     g.addEdge(infill, 'image', i2l, 'image');
     g.addEdge(vaeSource, 'vae', i2l, 'vae');
     g.addEdge(i2l, 'latents', denoise, 'latents');
@@ -150,7 +161,11 @@ export const addOutpaint = async ({
   } else {
     infill.image = { image_name: initialImage.image_name };
     // No scale before processing, much simpler
-    const i2l = addImageToLatents(g, modelLoader.type === 'flux_model_loader', fp32);
+    const i2l = g.addNode({
+      id: i2lNodeType,
+      type: i2lNodeType,
+      ...(i2lNodeType === 'i2l' ? { fp32 } : {}),
+    });
     const maskAlphaToMask = g.addNode({
       id: getPrefixedId('mask_alpha_to_mask'),
       type: 'tomask',
@@ -187,7 +202,7 @@ export const addOutpaint = async ({
     g.addEdge(i2l, 'latents', denoise, 'latents');
     g.addEdge(vaeSource, 'vae', i2l, 'vae');
     g.addEdge(vaeSource, 'vae', createGradientMask, 'vae');
-    if (modelLoader.type !== 'flux_model_loader') {
+    if (modelLoader.type !== 'flux_model_loader' && modelLoader.type !== 'sd3_model_loader') {
       g.addEdge(modelLoader, 'unet', createGradientMask, 'unet');
     }
 

invokeai/frontend/web/src/features/nodes/util/graph/generation/addTextToImage.ts

@@ -6,7 +6,7 @@ import type { Invocation } from 'services/api/types';
 
 type AddTextToImageArg = {
   g: Graph;
-  l2i: Invocation<'l2i' | 'flux_vae_decode'>;
+  l2i: Invocation<'l2i' | 'flux_vae_decode' | 'sd3_l2i'>;
   originalSize: Dimensions;
   scaledSize: Dimensions;
 };
@@ -16,7 +16,7 @@ export const addTextToImage = ({
   l2i,
   originalSize,
   scaledSize,
-}: AddTextToImageArg): Invocation<'img_resize' | 'l2i' | 'flux_vae_decode'> => {
+}: AddTextToImageArg): Invocation<'img_resize' | 'l2i' | 'flux_vae_decode' | 'sd3_l2i'> => {
   if (!isEqual(scaledSize, originalSize)) {
     // We need to resize the output image back to the original size
     const resizeImageToOriginalSize = g.addNode({

invokeai/frontend/web/src/features/nodes/util/graph/generation/addWatermarker.ts

@@ -11,7 +11,7 @@ import type { Invocation } from 'services/api/types';
 export const addWatermarker = (
   g: Graph,
   imageOutput: Invocation<
-    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode'
+    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode' | 'sd3_l2i'
   >
 ): Invocation<'img_watermark'> => {
   const watermark = g.addNode({

invokeai/frontend/web/src/features/nodes/util/graph/generation/buildFLUXGraph.ts

@@ -139,7 +139,7 @@ export const buildFLUXGraph = async (
   }
 
   let canvasOutput: Invocation<
-    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode'
+    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode' | 'sd3_l2i'
   > = l2i;
 
   if (generationMode === 'txt2img') {
@@ -149,6 +149,7 @@ export const buildFLUXGraph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'flux_vae_encode',
       denoise,
       vaeSource: modelLoader,
       originalSize,
@@ -163,6 +164,7 @@ export const buildFLUXGraph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'flux_vae_encode',
       denoise,
       vaeSource: modelLoader,
       modelLoader,
@@ -177,6 +179,7 @@ export const buildFLUXGraph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'flux_vae_encode',
       denoise,
       vaeSource: modelLoader,
       modelLoader,

invokeai/frontend/web/src/features/nodes/util/graph/generation/buildSD1Graph.ts

@@ -170,7 +170,7 @@ export const buildSD1Graph = async (
   const denoising_start = 1 - params.img2imgStrength;
 
   let canvasOutput: Invocation<
-    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode'
+    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode' | 'sd3_l2i'
   > = l2i;
 
   if (generationMode === 'txt2img') {
@@ -180,6 +180,7 @@ export const buildSD1Graph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'i2l',
       denoise,
       vaeSource,
       originalSize,
@@ -194,6 +195,7 @@ export const buildSD1Graph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'i2l',
       denoise,
       vaeSource,
       modelLoader,
@@ -208,6 +210,7 @@ export const buildSD1Graph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'i2l',
       denoise,
       vaeSource,
       modelLoader,

invokeai/frontend/web/src/features/nodes/util/graph/generation/buildSD3Graph.ts

@@ -0,0 +1,216 @@
+import { logger } from 'app/logging/logger';
+import type { RootState } from 'app/store/store';
+import type { CanvasManager } from 'features/controlLayers/konva/CanvasManager';
+import { getPrefixedId } from 'features/controlLayers/konva/util';
+import { selectCanvasSettingsSlice } from 'features/controlLayers/store/canvasSettingsSlice';
+import { selectParamsSlice } from 'features/controlLayers/store/paramsSlice';
+import { selectCanvasMetadata, selectCanvasSlice } from 'features/controlLayers/store/selectors';
+import { fetchModelConfigWithTypeGuard } from 'features/metadata/util/modelFetchingHelpers';
+import { addImageToImage } from 'features/nodes/util/graph/generation/addImageToImage';
+import { addInpaint } from 'features/nodes/util/graph/generation/addInpaint';
+import { addNSFWChecker } from 'features/nodes/util/graph/generation/addNSFWChecker';
+import { addOutpaint } from 'features/nodes/util/graph/generation/addOutpaint';
+import { addTextToImage } from 'features/nodes/util/graph/generation/addTextToImage';
+import { addWatermarker } from 'features/nodes/util/graph/generation/addWatermarker';
+import { Graph } from 'features/nodes/util/graph/generation/Graph';
+import {
+  CANVAS_OUTPUT_PREFIX,
+  getBoardField,
+  getPresetModifiedPrompts,
+  getSizes,
+} from 'features/nodes/util/graph/graphBuilderUtils';
+import type { Invocation } from 'services/api/types';
+import { isNonRefinerMainModelConfig } from 'services/api/types';
+import type { Equals } from 'tsafe';
+import { assert } from 'tsafe';
+
+const log = logger('system');
+
+export const buildSD3Graph = async (
+  state: RootState,
+  manager: CanvasManager
+): Promise<{ g: Graph; noise: Invocation<'sd3_denoise'>; posCond: Invocation<'sd3_text_encoder'> }> => {
+  const generationMode = await manager.compositor.getGenerationMode();
+  log.debug({ generationMode }, 'Building SD3 graph');
+
+  const params = selectParamsSlice(state);
+  const canvasSettings = selectCanvasSettingsSlice(state);
+  const canvas = selectCanvasSlice(state);
+
+  const { bbox } = canvas;
+
+  const {
+    model,
+    cfgScale: cfg_scale,
+    seed,
+    steps,
+    vae,
+    t5EncoderModel,
+    clipLEmbedModel,
+    clipGEmbedModel,
+    optimizedDenoisingEnabled,
+    img2imgStrength,
+  } = params;
+
+  assert(model, 'No model found in state');
+
+  const { originalSize, scaledSize } = getSizes(bbox);
+  const { positivePrompt, negativePrompt } = getPresetModifiedPrompts(state);
+
+  const g = new Graph(getPrefixedId('sd3_graph'));
+  const modelLoader = g.addNode({
+    type: 'sd3_model_loader',
+    id: getPrefixedId('sd3_model_loader'),
+    model,
+    t5_encoder_model: t5EncoderModel,
+    clip_l_model: clipLEmbedModel,
+    clip_g_model: clipGEmbedModel,
+    vae_model: vae,
+  });
+  const posCond = g.addNode({
+    type: 'sd3_text_encoder',
+    id: getPrefixedId('pos_cond'),
+    prompt: positivePrompt,
+  });
+
+  const negCond = g.addNode({
+    type: 'sd3_text_encoder',
+    id: getPrefixedId('neg_cond'),
+    prompt: negativePrompt,
+  });
+
+  const denoise = g.addNode({
+    type: 'sd3_denoise',
+    id: getPrefixedId('sd3_denoise'),
+    cfg_scale,
+    steps,
+    denoising_start: 0,
+    denoising_end: 1,
+    width: scaledSize.width,
+    height: scaledSize.height,
+  });
+  const l2i = g.addNode({
+    type: 'sd3_l2i',
+    id: getPrefixedId('l2i'),
+  });
+
+  g.addEdge(modelLoader, 'transformer', denoise, 'transformer');
+  g.addEdge(modelLoader, 'clip_l', posCond, 'clip_l');
+  g.addEdge(modelLoader, 'clip_l', negCond, 'clip_l');
+  g.addEdge(modelLoader, 'clip_g', posCond, 'clip_g');
+  g.addEdge(modelLoader, 'clip_g', negCond, 'clip_g');
+  g.addEdge(modelLoader, 't5_encoder', posCond, 't5_encoder');
+  g.addEdge(modelLoader, 't5_encoder', negCond, 't5_encoder');
+
+  g.addEdge(posCond, 'conditioning', denoise, 'positive_conditioning');
+  g.addEdge(negCond, 'conditioning', denoise, 'negative_conditioning');
+
+  g.addEdge(denoise, 'latents', l2i, 'latents');
+
+  const modelConfig = await fetchModelConfigWithTypeGuard(model.key, isNonRefinerMainModelConfig);
+  assert(modelConfig.base === 'sd-3');
+
+  g.upsertMetadata({
+    generation_mode: 'sd3_txt2img',
+    cfg_scale,
+    width: originalSize.width,
+    height: originalSize.height,
+    positive_prompt: positivePrompt,
+    negative_prompt: negativePrompt,
+    model: Graph.getModelMetadataField(modelConfig),
+    seed,
+    steps,
+    vae: vae ?? undefined,
+  });
+  g.addEdge(modelLoader, 'vae', l2i, 'vae');
+
+  let denoising_start: number;
+  if (optimizedDenoisingEnabled) {
+    // We rescale the img2imgStrength (with exponent 0.2) to effectively use the entire range [0, 1] and make the scale
+    // more user-friendly for SD3.5. Without this, most of the 'change' is concentrated in the high denoise strength
+    // range (>0.9).
+    denoising_start = 1 - img2imgStrength ** 0.2;
+  } else {
+    denoising_start = 1 - img2imgStrength;
+  }
+
+  let canvasOutput: Invocation<
+    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode' | 'sd3_l2i'
+  > = l2i;
+
+  if (generationMode === 'txt2img') {
+    canvasOutput = addTextToImage({ g, l2i, originalSize, scaledSize });
+  } else if (generationMode === 'img2img') {
+    canvasOutput = await addImageToImage({
+      g,
+      manager,
+      l2i,
+      i2lNodeType: 'sd3_i2l',
+      denoise,
+      vaeSource: modelLoader,
+      originalSize,
+      scaledSize,
+      bbox,
+      denoising_start,
+      fp32: false,
+    });
+  } else if (generationMode === 'inpaint') {
+    canvasOutput = await addInpaint({
+      state,
+      g,
+      manager,
+      l2i,
+      i2lNodeType: 'sd3_i2l',
+      denoise,
+      vaeSource: modelLoader,
+      modelLoader,
+      originalSize,
+      scaledSize,
+      denoising_start,
+      fp32: false,
+    });
+  } else if (generationMode === 'outpaint') {
+    canvasOutput = await addOutpaint({
+      state,
+      g,
+      manager,
+      l2i,
+      i2lNodeType: 'sd3_i2l',
+      denoise,
+      vaeSource: modelLoader,
+      modelLoader,
+      originalSize,
+      scaledSize,
+      denoising_start,
+      fp32: false,
+    });
+  } else {
+    assert<Equals<typeof generationMode, never>>(false);
+  }
+
+  if (state.system.shouldUseNSFWChecker) {
+    canvasOutput = addNSFWChecker(g, canvasOutput);
+  }
+
+  if (state.system.shouldUseWatermarker) {
+    canvasOutput = addWatermarker(g, canvasOutput);
+  }
+
+  // This image will be staged, should not be saved to the gallery or added to a board.
+  const is_intermediate = canvasSettings.sendToCanvas;
+  const board = canvasSettings.sendToCanvas ? undefined : getBoardField(state);
+
+  if (!canvasSettings.sendToCanvas) {
+    g.upsertMetadata(selectCanvasMetadata(state));
+  }
+
+  g.updateNode(canvasOutput, {
+    id: getPrefixedId(CANVAS_OUTPUT_PREFIX),
+    is_intermediate,
+    use_cache: false,
+    board,
+  });
+
+  g.setMetadataReceivingNode(canvasOutput);
+  return { g, noise: denoise, posCond };
+};

invokeai/frontend/web/src/features/nodes/util/graph/generation/buildSDXLGraph.ts

@@ -175,7 +175,7 @@ export const buildSDXLGraph = async (
     : 1 - params.img2imgStrength;
 
   let canvasOutput: Invocation<
-    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode'
+    'l2i' | 'img_nsfw' | 'img_watermark' | 'img_resize' | 'canvas_v2_mask_and_crop' | 'flux_vae_decode' | 'sd3_l2i'
   > = l2i;
 
   if (generationMode === 'txt2img') {
@@ -185,6 +185,7 @@ export const buildSDXLGraph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'i2l',
       denoise,
       vaeSource,
       originalSize,
@@ -199,6 +200,7 @@ export const buildSDXLGraph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'i2l',
       denoise,
       vaeSource,
       modelLoader,
@@ -213,6 +215,7 @@ export const buildSDXLGraph = async (
       g,
       manager,
       l2i,
+      i2lNodeType: 'i2l',
       denoise,
       vaeSource,
       modelLoader,

invokeai/frontend/web/src/features/nodes/util/graph/graphBuilderUtils.ts

@@ -118,16 +118,4 @@ export const getInfill = (
   assert(false, 'Unknown infill method');
 };
 
-export const addImageToLatents = (g: Graph, isFlux: boolean, fp32: boolean, image_name?: string) => {
-  if (isFlux) {
-    return g.addNode({
-      id: 'flux_vae_encode',
-      type: 'flux_vae_encode',
-      image: image_name ? { image_name } : undefined,
-    });
-  } else {
-    return g.addNode({ id: 'i2l', type: 'i2l', fp32, image: image_name ? { image_name } : undefined });
-  }
-};
-
 export const CANVAS_OUTPUT_PREFIX = 'canvas_output';

invokeai/frontend/web/src/features/parameters/components/Advanced/ParamCLIPGEmbedModelSelect.tsx

@@ -0,0 +1,42 @@
+import { Combobox, FormControl, FormLabel } from '@invoke-ai/ui-library';
+import { useAppDispatch, useAppSelector } from 'app/store/storeHooks';
+import { useModelCombobox } from 'common/hooks/useModelCombobox';
+import { clipGEmbedModelSelected, selectCLIPGEmbedModel } from 'features/controlLayers/store/paramsSlice';
+import { zModelIdentifierField } from 'features/nodes/types/common';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { useCLIPEmbedModels } from 'services/api/hooks/modelsByType';
+import type { CLIPGEmbedModelConfig } from 'services/api/types';
+import { isCLIPGEmbedModelConfig } from 'services/api/types';
+
+const ParamCLIPEmbedModelSelect = () => {
+  const dispatch = useAppDispatch();
+  const { t } = useTranslation();
+  const clipEmbedModel = useAppSelector(selectCLIPGEmbedModel);
+  const [modelConfigs, { isLoading }] = useCLIPEmbedModels();
+
+  const _onChange = useCallback(
+    (clipEmbedModel: CLIPGEmbedModelConfig | null) => {
+      if (clipEmbedModel) {
+        dispatch(clipGEmbedModelSelected(zModelIdentifierField.parse(clipEmbedModel)));
+      }
+    },
+    [dispatch]
+  );
+
+  const { options, value, onChange, noOptionsMessage } = useModelCombobox({
+    modelConfigs: modelConfigs.filter((config) => isCLIPGEmbedModelConfig(config)),
+    onChange: _onChange,
+    selectedModel: clipEmbedModel,
+    isLoading,
+  });
+
+  return (
+    <FormControl isDisabled={!options.length} isInvalid={!options.length} minW={0} flexGrow={1}>
+      <FormLabel m={0}>{t('modelManager.clipGEmbed')}</FormLabel>
+      <Combobox value={value} options={options} onChange={onChange} noOptionsMessage={noOptionsMessage} />
+    </FormControl>
+  );
+};
+
+export default memo(ParamCLIPEmbedModelSelect);

invokeai/frontend/web/src/features/parameters/components/Advanced/ParamCLIPLEmbedModelSelect.tsx

@@ -0,0 +1,42 @@
+import { Combobox, FormControl, FormLabel } from '@invoke-ai/ui-library';
+import { useAppDispatch, useAppSelector } from 'app/store/storeHooks';
+import { useModelCombobox } from 'common/hooks/useModelCombobox';
+import { clipLEmbedModelSelected, selectCLIPLEmbedModel } from 'features/controlLayers/store/paramsSlice';
+import { zModelIdentifierField } from 'features/nodes/types/common';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { useCLIPEmbedModels } from 'services/api/hooks/modelsByType';
+import type { CLIPLEmbedModelConfig } from 'services/api/types';
+import { isCLIPLEmbedModelConfig } from 'services/api/types';
+
+const ParamCLIPEmbedModelSelect = () => {
+  const dispatch = useAppDispatch();
+  const { t } = useTranslation();
+  const clipEmbedModel = useAppSelector(selectCLIPLEmbedModel);
+  const [modelConfigs, { isLoading }] = useCLIPEmbedModels();
+
+  const _onChange = useCallback(
+    (clipEmbedModel: CLIPLEmbedModelConfig | null) => {
+      if (clipEmbedModel) {
+        dispatch(clipLEmbedModelSelected(zModelIdentifierField.parse(clipEmbedModel)));
+      }
+    },
+    [dispatch]
+  );
+
+  const { options, value, onChange, noOptionsMessage } = useModelCombobox({
+    modelConfigs: modelConfigs.filter((config) => isCLIPLEmbedModelConfig(config)),
+    onChange: _onChange,
+    selectedModel: clipEmbedModel,
+    isLoading,
+  });
+
+  return (
+    <FormControl isDisabled={!options.length} isInvalid={!options.length} minW={0} flexGrow={1}>
+      <FormLabel m={0}>{t('modelManager.clipLEmbed')}</FormLabel>
+      <Combobox value={value} options={options} onChange={onChange} noOptionsMessage={noOptionsMessage} />
+    </FormControl>
+  );
+};
+
+export default memo(ParamCLIPEmbedModelSelect);

invokeai/frontend/web/src/features/parameters/components/MainModel/ParamMainModelSelect.tsx

@@ -9,7 +9,7 @@ import { selectActiveTab } from 'features/ui/store/uiSelectors';
 import { memo, useCallback, useMemo } from 'react';
 import { useTranslation } from 'react-i18next';
 import { MdMoneyOff } from 'react-icons/md';
-import { useNonSD3MainModels } from 'services/api/hooks/modelsByType';
+import { useMainModels } from 'services/api/hooks/modelsByType';
 import { type AnyModelConfig, isCheckpointMainModelConfig, type MainModelConfig } from 'services/api/types';
 
 const ParamMainModelSelect = () => {
@@ -18,7 +18,7 @@ const ParamMainModelSelect = () => {
   const activeTabName = useAppSelector(selectActiveTab);
   const selectedModelKey = useAppSelector(selectModelKey);
   // const selectedModel = useAppSelector(selectModel);
-  const [modelConfigs, { isLoading }] = useNonSD3MainModels();
+  const [modelConfigs, { isLoading }] = useMainModels();
 
   const selectedModel = useMemo(() => {
     if (!modelConfigs) {

invokeai/frontend/web/src/features/parameters/types/constants.ts

@@ -7,9 +7,10 @@ export const MODEL_TYPE_MAP = {
   any: 'Any',
   'sd-1': 'Stable Diffusion 1.x',
   'sd-2': 'Stable Diffusion 2.x',
+  'sd-3': 'Stable Diffusion 3.x',
   sdxl: 'Stable Diffusion XL',
   'sdxl-refiner': 'Stable Diffusion XL Refiner',
-  flux: 'Flux',
+  flux: 'FLUX',
 };
 
 /**

invokeai/frontend/web/src/features/parameters/types/parameterSchemas.ts

@@ -123,6 +123,16 @@ export const zParameterCLIPEmbedModel = zModelIdentifierField;
 export type ParameterCLIPEmbedModel = z.infer<typeof zParameterCLIPEmbedModel>;
 // #endregion
 
+// #region CLIP embed Model
+export const zParameterCLIPLEmbedModel = zModelIdentifierField;
+export type ParameterCLIPLEmbedModel = z.infer<typeof zParameterCLIPLEmbedModel>;
+// #endregion
+
+// #region CLIP embed Model
+export const zParameterCLIPGEmbedModel = zModelIdentifierField;
+export type ParameterCLIPGEmbedModel = z.infer<typeof zParameterCLIPGEmbedModel>;
+// #endregion
+
 // #region LoRA Model
 const zParameterLoRAModel = zModelIdentifierField;
 export type ParameterLoRAModel = z.infer<typeof zParameterLoRAModel>;

invokeai/frontend/web/src/features/settingsAccordions/components/AdvancedSettingsAccordion/AdvancedSettingsAccordion.tsx

@@ -3,9 +3,11 @@ import { Flex, FormControlGroup, StandaloneAccordion } from '@invoke-ai/ui-libra
 import { skipToken } from '@reduxjs/toolkit/query';
 import { createMemoizedSelector } from 'app/store/createMemoizedSelector';
 import { useAppSelector } from 'app/store/storeHooks';
-import { selectIsFLUX, selectParamsSlice, selectVAEKey } from 'features/controlLayers/store/paramsSlice';
+import { selectIsFLUX, selectIsSD3, selectParamsSlice, selectVAEKey } from 'features/controlLayers/store/paramsSlice';
 import ParamCFGRescaleMultiplier from 'features/parameters/components/Advanced/ParamCFGRescaleMultiplier';
 import ParamCLIPEmbedModelSelect from 'features/parameters/components/Advanced/ParamCLIPEmbedModelSelect';
+import ParamCLIPGEmbedModelSelect from 'features/parameters/components/Advanced/ParamCLIPGEmbedModelSelect';
+import ParamCLIPLEmbedModelSelect from 'features/parameters/components/Advanced/ParamCLIPLEmbedModelSelect';
 import ParamClipSkip from 'features/parameters/components/Advanced/ParamClipSkip';
 import ParamT5EncoderModelSelect from 'features/parameters/components/Advanced/ParamT5EncoderModelSelect';
 import ParamSeamlessXAxis from 'features/parameters/components/Seamless/ParamSeamlessXAxis';
@@ -35,6 +37,7 @@ export const AdvancedSettingsAccordion = memo(() => {
   const { currentData: vaeConfig } = useGetModelConfigQuery(vaeKey ?? skipToken);
   const activeTabName = useAppSelector(selectActiveTab);
   const isFLUX = useAppSelector(selectIsFLUX);
+  const isSD3 = useAppSelector(selectIsSD3);
 
   const selectBadges = useMemo(
     () =>
@@ -88,7 +91,7 @@ export const AdvancedSettingsAccordion = memo(() => {
       <Flex gap={4} alignItems="center" p={4} flexDir="column" data-testid="advanced-settings-accordion">
         <Flex gap={4} w="full">
           {isFLUX ? <ParamFLUXVAEModelSelect /> : <ParamVAEModelSelect />}
-          {!isFLUX && <ParamVAEPrecision />}
+          {!isFLUX && !isSD3 && <ParamVAEPrecision />}
         </Flex>
         {activeTabName === 'upscaling' ? (
           <Flex gap={4} alignItems="center">
@@ -98,7 +101,7 @@ export const AdvancedSettingsAccordion = memo(() => {
           </Flex>
         ) : (
           <>
-            {!isFLUX && (
+            {!isFLUX && !isSD3 && (
               <>
                 <FormControlGroup formLabelProps={formLabelProps}>
                   <ParamClipSkip />
@@ -118,6 +121,13 @@ export const AdvancedSettingsAccordion = memo(() => {
                 <ParamCLIPEmbedModelSelect />
               </FormControlGroup>
             )}
+            {isSD3 && (
+              <FormControlGroup>
+                <ParamT5EncoderModelSelect />
+                <ParamCLIPLEmbedModelSelect />
+                <ParamCLIPGEmbedModelSelect />
+              </FormControlGroup>
+            )}
           </>
         )}
       </Flex>

invokeai/frontend/web/src/features/settingsAccordions/components/GenerationSettingsAccordion/GenerationSettingsAccordion.tsx

@@ -4,7 +4,7 @@ import { EMPTY_ARRAY } from 'app/store/constants';
 import { createMemoizedSelector } from 'app/store/createMemoizedSelector';
 import { useAppSelector } from 'app/store/storeHooks';
 import { selectLoRAsSlice } from 'features/controlLayers/store/lorasSlice';
-import { selectIsFLUX } from 'features/controlLayers/store/paramsSlice';
+import { selectIsFLUX, selectIsSD3 } from 'features/controlLayers/store/paramsSlice';
 import { LoRAList } from 'features/lora/components/LoRAList';
 import LoRASelect from 'features/lora/components/LoRASelect';
 import ParamCFGScale from 'features/parameters/components/Core/ParamCFGScale';
@@ -30,6 +30,7 @@ export const GenerationSettingsAccordion = memo(() => {
   const modelConfig = useSelectedModelConfig();
   const activeTabName = useAppSelector(selectActiveTab);
   const isFLUX = useAppSelector(selectIsFLUX);
+  const isSD3 = useAppSelector(selectIsSD3);
   const selectBadges = useMemo(
     () =>
       createMemoizedSelector(selectLoRAsSlice, (loras) => {
@@ -74,7 +75,7 @@ export const GenerationSettingsAccordion = memo(() => {
         <Expander label={t('accordions.advanced.options')} isOpen={isOpenExpander} onToggle={onToggleExpander}>
           <Flex gap={4} flexDir="column" pb={4}>
             <FormControlGroup formLabelProps={formLabelProps}>
-              {!isFLUX && <ParamScheduler />}
+              {!isFLUX && !isSD3 && <ParamScheduler />}
               <ParamSteps />
               {isFLUX ? <ParamGuidance /> : <ParamCFGScale />}
             </FormControlGroup>

invokeai/frontend/web/src/features/settingsAccordions/components/ImageSettingsAccordion/ImageSettingsAccordion.tsx

@@ -3,7 +3,7 @@ import { Expander, Flex, FormControlGroup, StandaloneAccordion } from '@invoke-a
 import { EMPTY_ARRAY } from 'app/store/constants';
 import { createMemoizedSelector } from 'app/store/createMemoizedSelector';
 import { useAppSelector } from 'app/store/storeHooks';
-import { selectIsFLUX, selectParamsSlice } from 'features/controlLayers/store/paramsSlice';
+import { selectIsFLUX, selectIsSD3, selectParamsSlice } from 'features/controlLayers/store/paramsSlice';
 import { selectCanvasSlice, selectScaleMethod } from 'features/controlLayers/store/selectors';
 import { ParamOptimizedDenoisingToggle } from 'features/parameters/components/Advanced/ParamOptimizedDenoisingToggle';
 import BboxScaledHeight from 'features/parameters/components/Bbox/BboxScaledHeight';
@@ -60,6 +60,7 @@ export const ImageSettingsAccordion = memo(() => {
     defaultIsOpen: false,
   });
   const isFLUX = useAppSelector(selectIsFLUX);
+  const isSD3 = useAppSelector(selectIsSD3);
 
   return (
     <StandaloneAccordion
@@ -77,7 +78,7 @@ export const ImageSettingsAccordion = memo(() => {
         </Flex>
         <Expander label={t('accordions.advanced.options')} isOpen={isOpenExpander} onToggle={onToggleExpander}>
           <Flex gap={4} pb={4} flexDir="column">
-            {isFLUX && <ParamOptimizedDenoisingToggle />}
+            {(isFLUX || isSD3) && <ParamOptimizedDenoisingToggle />}
             <BboxScaleMethod />
             {scaleMethod !== 'none' && (
               <FormControlGroup formLabelProps={scalingLabelProps}>

invokeai/frontend/web/src/features/settingsAccordions/components/UpscaleSettingsAccordion/UpscaleWarning.tsx

@@ -34,8 +34,15 @@ export const UpscaleWarning = () => {
     dispatch(tileControlnetModelChanged(validModel || null));
   }, [model?.base, modelConfigs, dispatch]);
 
+  const isBaseModelCompatible = useMemo(() => {
+    return model && ['sd-1', 'sdxl'].includes(model.base);
+  }, [model]);
+
   const modelWarnings = useMemo(() => {
     const _warnings: string[] = [];
+    if (!isBaseModelCompatible) {
+      return _warnings;
+    }
     if (!model) {
       _warnings.push(t('upscaling.mainModelDesc'));
     }
@@ -46,7 +53,7 @@ export const UpscaleWarning = () => {
       _warnings.push(t('upscaling.upscaleModelDesc'));
     }
     return _warnings;
-  }, [model, tileControlnetModel, upscaleModel, t]);
+  }, [isBaseModelCompatible, model, tileControlnetModel, upscaleModel, t]);
 
   const otherWarnings = useMemo(() => {
     const _warnings: string[] = [];
@@ -58,22 +65,25 @@ export const UpscaleWarning = () => {
     return _warnings;
   }, [isTooLargeToUpscale, t, maxUpscaleDimension]);
 
+  const allWarnings = useMemo(() => [...modelWarnings, ...otherWarnings], [modelWarnings, otherWarnings]);
+
   const handleGoToModelManager = useCallback(() => {
     dispatch(setActiveTab('models'));
     $installModelsTab.set(3);
   }, [dispatch]);
 
-  if (modelWarnings.length && isModelsTabDisabled) {
+  if (isBaseModelCompatible && modelWarnings.length > 0 && isModelsTabDisabled) {
     return null;
   }
 
-  if ((!modelWarnings.length && !otherWarnings.length) || isLoading) {
+  if ((isBaseModelCompatible && allWarnings.length === 0) || isLoading) {
     return null;
   }
 
   return (
     <Flex bg="error.500" borderRadius="base" padding={4} direction="column" fontSize="sm" gap={2}>
-      {!!modelWarnings.length && (
+      {!isBaseModelCompatible && <Text>{t('upscaling.incompatibleBaseModelDesc')}</Text>}
+      {modelWarnings.length > 0 && (
         <Text>
           <Trans
             i18nKey="upscaling.missingModelsWarning"
@@ -85,11 +95,13 @@ export const UpscaleWarning = () => {
           />
         </Text>
       )}
-      <UnorderedList>
-        {[...modelWarnings, ...otherWarnings].map((warning) => (
-          <ListItem key={warning}>{warning}</ListItem>
-        ))}
-      </UnorderedList>
+      {allWarnings.length > 0 && (
+        <UnorderedList>
+          {allWarnings.map((warning) => (
+            <ListItem key={warning}>{warning}</ListItem>
+          ))}
+        </UnorderedList>
+      )}
     </Flex>
   );
 };

invokeai/frontend/web/src/features/system/components/SettingsModal/SettingsModal.tsx

@@ -26,17 +26,20 @@ import { SettingsDeveloperLogLevel } from 'features/system/components/SettingsMo
 import { SettingsDeveloperLogNamespaces } from 'features/system/components/SettingsModal/SettingsDeveloperLogNamespaces';
 import { useClearIntermediates } from 'features/system/components/SettingsModal/useClearIntermediates';
 import { StickyScrollable } from 'features/system/components/StickyScrollable';
+import { useFeatureStatus } from 'features/system/hooks/useFeatureStatus';
 import {
   selectSystemShouldAntialiasProgressImage,
   selectSystemShouldConfirmOnDelete,
   selectSystemShouldConfirmOnNewSession,
   selectSystemShouldEnableInformationalPopovers,
   selectSystemShouldEnableModelDescriptions,
+  selectSystemShouldShowInvocationProgressDetail,
   selectSystemShouldUseNSFWChecker,
   selectSystemShouldUseWatermarker,
   setShouldConfirmOnDelete,
   setShouldEnableInformationalPopovers,
   setShouldEnableModelDescriptions,
+  setShouldShowInvocationProgressDetail,
   shouldAntialiasProgressImageChanged,
   shouldConfirmOnNewSessionToggled,
   shouldUseNSFWCheckerChanged,
@@ -103,6 +106,8 @@ const SettingsModal = ({ config = defaultConfig, children }: SettingsModalProps)
   const shouldEnableInformationalPopovers = useAppSelector(selectSystemShouldEnableInformationalPopovers);
   const shouldEnableModelDescriptions = useAppSelector(selectSystemShouldEnableModelDescriptions);
   const shouldConfirmOnNewSession = useAppSelector(selectSystemShouldConfirmOnNewSession);
+  const shouldShowInvocationProgressDetail = useAppSelector(selectSystemShouldShowInvocationProgressDetail);
+  const isInvocationProgressAlertEnabled = useFeatureStatus('invocationProgressAlert');
   const onToggleConfirmOnNewSession = useCallback(() => {
     dispatch(shouldConfirmOnNewSessionToggled());
   }, [dispatch]);
@@ -170,6 +175,13 @@ const SettingsModal = ({ config = defaultConfig, children }: SettingsModalProps)
     [dispatch]
   );
 
+  const handleChangeShouldShowInvocationProgressDetail = useCallback(
+    (e: ChangeEvent<HTMLInputElement>) => {
+      dispatch(setShouldShowInvocationProgressDetail(e.target.checked));
+    },
+    [dispatch]
+  );
+
   return (
     <>
       {cloneElement(children, {
@@ -221,6 +233,15 @@ const SettingsModal = ({ config = defaultConfig, children }: SettingsModalProps)
                         onChange={handleChangeShouldAntialiasProgressImage}
                       />
                     </FormControl>
+                    {isInvocationProgressAlertEnabled && (
+                      <FormControl>
+                        <FormLabel>{t('settings.showDetailedInvocationProgress')}</FormLabel>
+                        <Switch
+                          isChecked={shouldShowInvocationProgressDetail}
+                          onChange={handleChangeShouldShowInvocationProgressDetail}
+                        />
+                      </FormControl>
+                    )}
                     <FormControl>
                       <InformationalPopover feature="noiseUseCPU" inPortal={false}>
                         <FormLabel>{t('parameters.useCpuNoise')}</FormLabel>

invokeai/frontend/web/src/features/system/store/systemSlice.ts

@@ -21,6 +21,7 @@ const initialSystemState: SystemState = {
   logIsEnabled: true,
   logLevel: 'debug',
   logNamespaces: [...zLogNamespace.options],
+  shouldShowInvocationProgressDetail: false,
 };
 
 export const systemSlice = createSlice({
@@ -64,6 +65,9 @@ export const systemSlice = createSlice({
     shouldConfirmOnNewSessionToggled(state) {
       state.shouldConfirmOnNewSession = !state.shouldConfirmOnNewSession;
     },
+    setShouldShowInvocationProgressDetail(state, action: PayloadAction<boolean>) {
+      state.shouldShowInvocationProgressDetail = action.payload;
+    },
   },
 });
 
@@ -79,6 +83,7 @@ export const {
   setShouldEnableInformationalPopovers,
   setShouldEnableModelDescriptions,
   shouldConfirmOnNewSessionToggled,
+  setShouldShowInvocationProgressDetail,
 } = systemSlice.actions;
 
 /* eslint-disable-next-line @typescript-eslint/no-explicit-any */
@@ -117,3 +122,6 @@ export const selectSystemShouldEnableModelDescriptions = createSystemSelector(
   (system) => system.shouldEnableModelDescriptions
 );
 export const selectSystemShouldConfirmOnNewSession = createSystemSelector((system) => system.shouldConfirmOnNewSession);
+export const selectSystemShouldShowInvocationProgressDetail = createSystemSelector(
+  (system) => system.shouldShowInvocationProgressDetail
+);

invokeai/frontend/web/src/features/system/store/types.ts

@@ -41,4 +41,5 @@ export interface SystemState {
   logIsEnabled: boolean;
   logLevel: LogLevel;
   logNamespaces: LogNamespace[];
+  shouldShowInvocationProgressDetail: boolean;
 }

invokeai/frontend/web/src/services/api/endpoints/images.ts

@@ -6,10 +6,10 @@ import type { components, paths } from 'services/api/schema';
 import type {
   DeleteBoardResult,
   GraphAndWorkflowResponse,
-  ImageCategory,
   ImageDTO,
   ListImagesArgs,
   ListImagesResponse,
+  UploadImageArg,
 } from 'services/api/types';
 import { getCategories, getListImagesUrl } from 'services/api/util';
 import type { JsonObject } from 'type-fest';
@@ -260,20 +260,7 @@ export const imagesApi = api.injectEndpoints({
         return [];
       },
     }),
-    uploadImage: build.mutation<
-      ImageDTO,
-      {
-        file: File;
-        image_category: ImageCategory;
-        is_intermediate: boolean;
-        session_id?: string;
-        board_id?: string;
-        crop_visible?: boolean;
-        metadata?: JsonObject;
-        isFirstUploadOfBatch?: boolean;
-        withToast?: boolean;
-      }
-    >({
+    uploadImage: build.mutation<ImageDTO, UploadImageArg>({
       query: ({ file, image_category, is_intermediate, session_id, board_id, crop_visible, metadata }) => {
         const formData = new FormData();
         formData.append('file', file);
@@ -558,7 +545,6 @@ export const {
   useClearIntermediatesMutation,
   useAddImagesToBoardMutation,
   useRemoveImagesFromBoardMutation,
-  useChangeImageIsIntermediateMutation,
   useDeleteBoardAndImagesMutation,
   useDeleteBoardMutation,
   useStarImagesMutation,
@@ -622,79 +608,17 @@ export const getImageMetadata = (
   return req.unwrap();
 };
 
-export type UploadImageArg = {
-  file: File;
-  image_category: ImageCategory;
-  is_intermediate: boolean;
-  session_id?: string;
-  board_id?: string;
-  crop_visible?: boolean;
-  metadata?: JsonObject;
-  withToast?: boolean;
-};
-
 export const uploadImage = (arg: UploadImageArg): Promise<ImageDTO> => {
-  const {
-    file,
-    image_category,
-    is_intermediate,
-    crop_visible = false,
-    board_id,
-    metadata,
-    session_id,
-    withToast = true,
-  } = arg;
-
   const { dispatch } = getStore();
-
-  const req = dispatch(
-    imagesApi.endpoints.uploadImage.initiate(
-      {
-        file,
-        image_category,
-        is_intermediate,
-        crop_visible,
-        board_id,
-        metadata,
-        session_id,
-        withToast,
-      },
-      { track: false }
-    )
-  );
+  const req = dispatch(imagesApi.endpoints.uploadImage.initiate(arg, { track: false }));
   return req.unwrap();
 };
 
 export const uploadImages = async (args: UploadImageArg[]): Promise<ImageDTO[]> => {
   const { dispatch } = getStore();
   const results = await Promise.allSettled(
-    args.map((arg, i) => {
-      const {
-        file,
-        image_category,
-        is_intermediate,
-        crop_visible = false,
-        board_id,
-        metadata,
-        session_id,
-        withToast = true,
-      } = arg;
-      const req = dispatch(
-        imagesApi.endpoints.uploadImage.initiate(
-          {
-            file,
-            image_category,
-            is_intermediate,
-            crop_visible,
-            board_id,
-            metadata,
-            session_id,
-            isFirstUploadOfBatch: i === 0,
-            withToast,
-          },
-          { track: false }
-        )
-      );
+    args.map((arg) => {
+      const req = dispatch(imagesApi.endpoints.uploadImage.initiate(arg, { track: false }));
       return req.unwrap();
     })
   );

invokeai/frontend/web/src/services/api/hooks/modelsByType.ts

@@ -18,7 +18,6 @@ import {
   isIPAdapterModelConfig,
   isLoRAModelConfig,
   isNonRefinerMainModelConfig,
-  isNonSD3MainModelModelConfig,
   isNonSDXLMainModelConfig,
   isRefinerMainModelModelConfig,
   isSD3MainModelModelConfig,
@@ -53,7 +52,6 @@ const buildModelsHook =
   };
 
 export const useMainModels = buildModelsHook(isNonRefinerMainModelConfig);
-export const useNonSD3MainModels = buildModelsHook(isNonSD3MainModelModelConfig);
 export const useNonSDXLMainModels = buildModelsHook(isNonSDXLMainModelConfig);
 export const useRefinerModels = buildModelsHook(isRefinerMainModelModelConfig);
 export const useFluxModels = buildModelsHook(isFluxMainModelModelConfig);

invokeai/frontend/web/src/services/api/types.ts

@@ -1,5 +1,5 @@
 import type { components, paths } from 'services/api/schema';
-import type { SetRequired } from 'type-fest';
+import type { JsonObject, SetRequired } from 'type-fest';
 
 export type S = components['schemas'];
 
@@ -223,10 +223,6 @@ export const isSD3MainModelModelConfig = (config: AnyModelConfig): config is Mai
   return config.type === 'main' && config.base === 'sd-3';
 };
 
-export const isNonSD3MainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
-  return config.type === 'main' && config.base !== 'sd-3' && config.base !== 'sdxl-refiner';
-};
-
 export const isFluxMainModelModelConfig = (config: AnyModelConfig): config is MainModelConfig => {
   return config.type === 'main' && config.base === 'flux';
 };
@@ -291,3 +287,42 @@ export type SetHFTokenResponse = NonNullable<
 export type SetHFTokenArg = NonNullable<
   paths['/api/v2/models/hf_login']['post']['requestBody']['content']['application/json']
 >;
+
+export type UploadImageArg = {
+  /**
+   * The file object to upload
+   */
+  file: File;
+  /**
+   * THe category of image to upload
+   */
+  image_category: ImageCategory;
+  /**
+   * Whether the uploaded image is an intermediate image (intermediate images are not shown int he gallery)
+   */
+  is_intermediate: boolean;
+  /**
+   * The session with which to associate the uploaded image
+   */
+  session_id?: string;
+  /**
+   * The board id to add the image to
+   */
+  board_id?: string;
+  /**
+   * Whether or not to crop the image to its bounding box before saving
+   */
+  crop_visible?: boolean;
+  /**
+   * Metadata to embed in the image when saving it
+   */
+  metadata?: JsonObject;
+  /**
+   * Whether this upload should be "silent" (no toast on upload, no changing of gallery view)
+   */
+  silent?: boolean;
+  /**
+   * Whether this is the first upload of a batch (used when displaying user feedback with toasts - ignored if the upload is silent)
+   */
+  isFirstUploadOfBatch?: boolean;
+};

invokeai/frontend/web/src/services/events/stores.ts

@@ -1,3 +1,4 @@
+import { round } from 'lodash-es';
 import { atom, computed, map } from 'nanostores';
 import type { S } from 'services/api/types';
 import type { AppSocket } from 'services/events/types';
@@ -10,3 +11,16 @@ export const $lastProgressEvent = atom<S['InvocationProgressEvent'] | null>(null
 export const $progressImage = computed($lastProgressEvent, (val) => val?.image ?? null);
 export const $hasProgressImage = computed($lastProgressEvent, (val) => Boolean(val?.image));
 export const $isProgressFromCanvas = computed($lastProgressEvent, (val) => val?.destination === 'canvas');
+export const $invocationProgressMessage = computed($lastProgressEvent, (val) => {
+  if (!val) {
+    return null;
+  }
+  if (val.destination !== 'canvas') {
+    return null;
+  }
+  let message = val.message;
+  if (val.percentage) {
+    message += ` (${round(val.percentage * 100)}%)`;
+  }
+  return message;
+});

invokeai/version/invokeai_version.py

@@ -1 +1 @@
-__version__ = "5.4.1rc2"
+__version__ = "5.4.1"

tests/app/services/model_load/test_load_api.py

@@ -6,7 +6,11 @@ from diffusers import AutoencoderTiny
 
 from invokeai.app.services.invocation_services import InvocationServices
 from invokeai.app.services.model_manager import ModelManagerServiceBase
-from invokeai.app.services.shared.invocation_context import InvocationContext, build_invocation_context
+from invokeai.app.services.shared.invocation_context import (
+    InvocationContext,
+    InvocationContextData,
+    build_invocation_context,
+)
 from invokeai.backend.model_manager.load.load_base import LoadedModelWithoutConfig
 from tests.backend.model_manager.model_manager_fixtures import *  # noqa F403
 
@@ -19,7 +23,7 @@ def mock_context(
     mock_services.model_manager = mm2_model_manager
     return build_invocation_context(
         services=mock_services,
-        data=None,  # type: ignore
+        data=InvocationContextData(queue_item=None, invocation=None, source_invocation_id=None),  # type: ignore
         is_canceled=None,  # type: ignore
     )
 

tests/backend/flux/ip_adapter/xlabs_flux_ip_adapter_state_dict.py

@@ -1,6 +1,6 @@
 # State dict keys and shapes for an XLabs FLUX IP-Adapter model. Intended to be used for unit tests.
 # These keys were extracted from:
-# https://huggingface.co/XLabs-AI/flux-ip-adapter/blob/ad16be50d78a07ea83d8c4bde44ff9753235182e/flux-ip-adapter.safetensors
+# https://huggingface.co/XLabs-AI/flux-ip-adapter/resolve/main/ip_adapter.safetensors
 xlabs_sd_shapes = {
     "double_blocks.0.processor.ip_adapter_double_stream_k_proj.bias": [3072],
     "double_blocks.0.processor.ip_adapter_double_stream_k_proj.weight": [3072, 4096],

tests/backend/lora/conversions/lora_state_dicts/flux_lora_diffusers_no_proj_mlp_format.py

@@ -0,0 +1,688 @@
+# A sample state dict in the Diffusers FLUX LoRA format without .proj_mlp layers.
+# This format was added in response to https://github.com/invoke-ai/InvokeAI/issues/7129
+state_dict_keys = {
+    "transformer.single_transformer_blocks.0.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.0.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.0.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.0.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.0.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.0.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.1.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.1.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.1.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.1.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.1.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.1.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.10.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.10.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.10.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.10.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.10.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.10.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.11.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.11.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.11.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.11.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.11.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.11.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.12.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.12.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.12.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.12.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.12.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.12.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.13.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.13.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.13.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.13.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.13.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.13.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.14.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.14.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.14.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.14.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.14.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.14.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.15.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.15.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.15.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.15.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.15.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.15.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.16.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.16.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.16.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.16.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.16.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.16.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.17.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.17.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.17.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.17.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.17.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.17.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.18.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.18.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.18.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.18.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.18.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.18.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.19.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.19.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.19.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.19.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.19.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.19.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.2.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.2.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.2.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.2.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.2.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.2.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.20.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.20.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.20.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.20.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.20.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.20.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.21.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.21.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.21.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.21.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.21.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.21.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.22.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.22.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.22.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.22.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.22.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.22.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.23.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.23.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.23.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.23.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.23.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.23.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.24.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.24.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.24.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.24.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.24.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.24.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.25.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.25.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.25.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.25.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.25.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.25.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.26.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.26.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.26.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.26.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.26.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.26.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.27.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.27.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.27.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.27.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.27.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.27.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.28.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.28.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.28.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.28.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.28.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.28.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.29.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.29.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.29.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.29.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.29.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.29.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.3.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.3.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.3.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.3.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.3.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.3.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.30.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.30.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.30.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.30.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.30.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.30.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.31.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.31.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.31.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.31.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.31.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.31.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.32.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.32.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.32.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.32.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.32.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.32.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.33.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.33.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.33.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.33.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.33.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.33.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.34.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.34.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.34.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.34.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.34.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.34.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.35.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.35.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.35.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.35.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.35.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.35.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.36.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.36.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.36.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.36.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.36.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.36.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.37.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.37.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.37.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.37.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.37.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.37.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.4.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.4.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.4.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.4.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.4.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.4.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.5.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.5.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.5.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.5.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.5.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.5.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.6.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.6.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.6.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.6.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.6.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.6.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.7.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.7.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.7.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.7.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.7.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.7.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.8.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.8.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.8.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.8.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.8.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.8.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.9.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.9.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.9.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.9.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.single_transformer_blocks.9.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.single_transformer_blocks.9.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.0.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.0.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.0.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.0.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.0.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.0.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.1.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.1.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.1.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.1.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.1.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.1.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.10.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.10.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.10.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.10.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.10.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.10.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.11.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.11.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.11.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.11.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.11.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.11.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.12.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.12.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.12.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.12.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.12.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.12.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.13.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.13.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.13.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.13.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.13.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.13.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.14.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.14.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.14.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.14.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.14.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.14.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.15.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.15.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.15.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.15.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.15.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.15.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.16.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.16.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.16.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.16.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.16.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.16.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.17.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.17.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.17.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.17.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.17.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.17.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.18.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.18.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.18.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.18.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.18.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.18.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.2.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.2.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.2.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.2.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.2.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.2.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.3.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.3.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.3.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.3.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.3.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.3.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.4.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.4.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.4.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.4.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.4.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.4.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.5.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.5.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.5.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.5.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.5.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.5.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.6.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.6.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.6.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.6.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.6.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.6.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.7.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.7.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.7.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.7.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.7.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.7.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.8.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.8.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.8.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.8.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.8.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.8.ff_context.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.add_k_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.add_k_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.add_q_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.add_q_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.add_v_proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.add_v_proj.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.to_add_out.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.to_add_out.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.to_k.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.to_k.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.to_out.0.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.to_out.0.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.to_q.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.to_q.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.attn.to_v.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.attn.to_v.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.ff.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.ff.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.9.ff.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.9.ff.net.2.lora_B.weight": [3072, 16],
+    "transformer.transformer_blocks.9.ff_context.net.0.proj.lora_A.weight": [16, 3072],
+    "transformer.transformer_blocks.9.ff_context.net.0.proj.lora_B.weight": [12288, 16],
+    "transformer.transformer_blocks.9.ff_context.net.2.lora_A.weight": [16, 12288],
+    "transformer.transformer_blocks.9.ff_context.net.2.lora_B.weight": [3072, 16],
+}

tests/backend/lora/conversions/lora_state_dicts/utils.py

@@ -1,8 +1,8 @@
 import torch
 
 
-def keys_to_mock_state_dict(keys: list[str]) -> dict[str, torch.Tensor]:
+def keys_to_mock_state_dict(keys: dict[str, list[int]]) -> dict[str, torch.Tensor]:
     state_dict: dict[str, torch.Tensor] = {}
-    for k in keys:
-        state_dict[k] = torch.empty(1)
+    for k, shape in keys.items():
+        state_dict[k] = torch.empty(shape)
     return state_dict

tests/backend/lora/conversions/test_flux_diffusers_lora_conversion_utils.py

@@ -9,22 +9,26 @@ from invokeai.backend.lora.conversions.flux_lora_constants import FLUX_LORA_TRAN
 from tests.backend.lora.conversions.lora_state_dicts.flux_lora_diffusers_format import (
     state_dict_keys as flux_diffusers_state_dict_keys,
 )
+from tests.backend.lora.conversions.lora_state_dicts.flux_lora_diffusers_no_proj_mlp_format import (
+    state_dict_keys as flux_diffusers_no_proj_mlp_state_dict_keys,
+)
 from tests.backend.lora.conversions.lora_state_dicts.flux_lora_kohya_format import (
     state_dict_keys as flux_kohya_state_dict_keys,
 )
 from tests.backend.lora.conversions.lora_state_dicts.utils import keys_to_mock_state_dict
 
 
-def test_is_state_dict_likely_in_flux_diffusers_format_true():
+@pytest.mark.parametrize("sd_keys", [flux_diffusers_state_dict_keys, flux_diffusers_no_proj_mlp_state_dict_keys])
+def test_is_state_dict_likely_in_flux_diffusers_format_true(sd_keys: dict[str, list[int]]):
     """Test that is_state_dict_likely_in_flux_diffusers_format() can identify a state dict in the Diffusers FLUX LoRA format."""
     # Construct a state dict that is in the Diffusers FLUX LoRA format.
-    state_dict = keys_to_mock_state_dict(flux_diffusers_state_dict_keys)
+    state_dict = keys_to_mock_state_dict(sd_keys)
 
     assert is_state_dict_likely_in_flux_diffusers_format(state_dict)
 
 
 def test_is_state_dict_likely_in_flux_diffusers_format_false():
-    """Test that is_state_dict_likely_in_flux_diffusers_format() returns False for a state dict that is not in the Kohya
+    """Test that is_state_dict_likely_in_flux_diffusers_format() returns False for a state dict that is in the Kohya
     FLUX LoRA format.
     """
     # Construct a state dict that is not in the Kohya FLUX LoRA format.
@@ -33,16 +37,17 @@ def test_is_state_dict_likely_in_flux_diffusers_format_false():
     assert not is_state_dict_likely_in_flux_diffusers_format(state_dict)
 
 
-def test_lora_model_from_flux_diffusers_state_dict():
+@pytest.mark.parametrize("sd_keys", [flux_diffusers_state_dict_keys, flux_diffusers_no_proj_mlp_state_dict_keys])
+def test_lora_model_from_flux_diffusers_state_dict(sd_keys: dict[str, list[int]]):
     """Test that lora_model_from_flux_diffusers_state_dict() can load a state dict in the Diffusers FLUX LoRA format."""
     # Construct a state dict that is in the Diffusers FLUX LoRA format.
-    state_dict = keys_to_mock_state_dict(flux_diffusers_state_dict_keys)
+    state_dict = keys_to_mock_state_dict(sd_keys)
     # Load the state dict into a LoRAModelRaw object.
     model = lora_model_from_flux_diffusers_state_dict(state_dict, alpha=8.0)
 
     # Check that the model has the correct number of LoRA layers.
     expected_lora_layers: set[str] = set()
-    for k in flux_diffusers_state_dict_keys:
+    for k in sd_keys:
         k = k.replace("lora_A.weight", "")
         k = k.replace("lora_B.weight", "")
         expected_lora_layers.add(k)

tests/backend/lora/conversions/test_flux_kohya_lora_conversion_utils.py

@@ -23,7 +23,7 @@ from tests.backend.lora.conversions.lora_state_dicts.utils import keys_to_mock_s
 
 
 @pytest.mark.parametrize("sd_keys", [flux_kohya_state_dict_keys, flux_kohya_te1_state_dict_keys])
-def test_is_state_dict_likely_in_flux_kohya_format_true(sd_keys: list[str]):
+def test_is_state_dict_likely_in_flux_kohya_format_true(sd_keys: dict[str, list[int]]):
     """Test that is_state_dict_likely_in_flux_kohya_format() can identify a state dict in the Kohya FLUX LoRA format."""
     # Construct a state dict that is in the Kohya FLUX LoRA format.
     state_dict = keys_to_mock_state_dict(sd_keys)
@@ -83,7 +83,7 @@ def test_convert_flux_transformer_kohya_state_dict_to_invoke_format_error():
 
 
 @pytest.mark.parametrize("sd_keys", [flux_kohya_state_dict_keys, flux_kohya_te1_state_dict_keys])
-def test_lora_model_from_flux_kohya_state_dict(sd_keys: list[str]):
+def test_lora_model_from_flux_kohya_state_dict(sd_keys: dict[str, list[int]]):
     """Test that a LoRAModelRaw can be created from a state dict in the Kohya FLUX LoRA format."""
     # Construct a state dict that is in the Kohya FLUX LoRA format.
     state_dict = keys_to_mock_state_dict(sd_keys)

tests/test_nodes.py

@@ -9,6 +9,8 @@ from invokeai.app.invocations.baseinvocation import (
 from invokeai.app.invocations.fields import InputField, OutputField
 from invokeai.app.invocations.image import ImageField
 from invokeai.app.services.events.events_common import EventBase
+from invokeai.app.services.session_processor.session_processor_common import ProgressImage
+from invokeai.app.services.session_queue.session_queue_common import SessionQueueItem
 from invokeai.app.services.shared.invocation_context import InvocationContext
 
 
@@ -133,6 +135,16 @@ class TestEventService(EventServiceBase):
         self.events.append(event)
         pass
 
+    def emit_invocation_progress(
+        self,
+        queue_item: "SessionQueueItem",
+        invocation: "BaseInvocation",
+        message: str,
+        percentage: float | None = None,
+        image: "ProgressImage | None" = None,
+    ) -> None:
+        pass
+
 
 def wait_until(condition: Callable[[], bool], timeout: int = 10, interval: float = 0.1) -> None:
     import time