fix(package): pin python-multipart to v0.0.12 due to breakage in dependent packages with later versions

- Use a hash of the last processed points instead of a `hasProcessed` flag to determine whether or not we should re-process a given set of points.
- Store point coords in state instead of pulling them out of the konva node positions. This makes moving a point a more explicit action in code.
- Add a `roundCoord` util to round the x and y values of a coordinate.
- Ensure we always re-process when $points changes.

invokeai/app/api/routers/model_manager.py

@@ -808,7 +808,11 @@ def get_is_installed(
     for model in installed_models:
         if model.source == starter_model.source:
             return True
-        if model.name == starter_model.name and model.base == starter_model.base and model.type == starter_model.type:
+        if (
+            (model.name == starter_model.name or model.name in starter_model.previous_names)
+            and model.base == starter_model.base
+            and model.type == starter_model.type
+        ):
             return True
     return False
 

invokeai/app/invocations/fields.py

@@ -133,7 +133,6 @@ class FieldDescriptions:
     clip_embed_model = "CLIP Embed loader"
     unet = "UNet (scheduler, LoRAs)"
     transformer = "Transformer"
-    mmditx = "MMDiTX"
     vae = "VAE"
     cond = "Conditioning tensor"
     controlnet_model = "ControlNet model to load"
@@ -141,7 +140,6 @@ class FieldDescriptions:
     lora_model = "LoRA model to load"
     main_model = "Main model (UNet, VAE, CLIP) to load"
     flux_model = "Flux model (Transformer) to load"
-    sd3_model = "SD3 model (MMDiTX) to load"
     sdxl_main_model = "SDXL Main model (UNet, VAE, CLIP1, CLIP2) to load"
     sdxl_refiner_model = "SDXL Refiner Main Modde (UNet, VAE, CLIP2) to load"
     onnx_main_model = "ONNX Main model (UNet, VAE, CLIP) to load"

invokeai/app/invocations/flux_model_loader.py

@@ -1,86 +0,0 @@
-from typing import Literal
-
-from invokeai.app.invocations.baseinvocation import (
-    BaseInvocation,
-    BaseInvocationOutput,
-    Classification,
-    invocation,
-    invocation_output,
-)
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
-from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, T5EncoderField, TransformerField, VAEField
-from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.flux.util import max_seq_lengths
-from invokeai.backend.model_manager.config import CheckpointConfigBase, SubModelType
-
-
-@invocation_output("flux_model_loader_output")
-class FluxModelLoaderOutput(BaseInvocationOutput):
-    """Flux base model loader output"""
-
-    transformer: TransformerField = OutputField(description=FieldDescriptions.transformer, title="Transformer")
-    clip: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP")
-    t5_encoder: T5EncoderField = OutputField(description=FieldDescriptions.t5_encoder, title="T5 Encoder")
-    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
-    max_seq_len: Literal[256, 512] = OutputField(
-        description="The max sequence length to used for the T5 encoder. (256 for schnell transformer, 512 for dev transformer)",
-        title="Max Seq Length",
-    )
-
-
-@invocation(
-    "flux_model_loader",
-    title="Flux Main Model",
-    tags=["model", "flux"],
-    category="model",
-    version="1.0.4",
-    classification=Classification.Prototype,
-)
-class FluxModelLoaderInvocation(BaseInvocation):
-    """Loads a flux base model, outputting its submodels."""
-
-    model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.flux_model,
-        ui_type=UIType.FluxMainModel,
-        input=Input.Direct,
-    )
-
-    t5_encoder_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.t5_encoder, ui_type=UIType.T5EncoderModel, input=Input.Direct, title="T5 Encoder"
-    )
-
-    clip_embed_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.clip_embed_model,
-        ui_type=UIType.CLIPEmbedModel,
-        input=Input.Direct,
-        title="CLIP Embed",
-    )
-
-    vae_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.vae_model, ui_type=UIType.FluxVAEModel, title="VAE"
-    )
-
-    def invoke(self, context: InvocationContext) -> FluxModelLoaderOutput:
-        for key in [self.model.key, self.t5_encoder_model.key, self.clip_embed_model.key, self.vae_model.key]:
-            if not context.models.exists(key):
-                raise ValueError(f"Unknown model: {key}")
-
-        transformer = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
-        vae = self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
-
-        tokenizer = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
-        clip_encoder = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
-
-        tokenizer2 = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
-        t5_encoder = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
-
-        transformer_config = context.models.get_config(transformer)
-        assert isinstance(transformer_config, CheckpointConfigBase)
-
-        return FluxModelLoaderOutput(
-            transformer=TransformerField(transformer=transformer, loras=[]),
-            clip=CLIPField(tokenizer=tokenizer, text_encoder=clip_encoder, loras=[], skipped_layers=0),
-            t5_encoder=T5EncoderField(tokenizer=tokenizer2, text_encoder=t5_encoder),
-            vae=VAEField(vae=vae),
-            max_seq_len=max_seq_lengths[transformer_config.config_path],
-        )

invokeai/app/invocations/model.py

@@ -1,5 +1,5 @@
 import copy
-from typing import List, Optional
+from typing import List, Literal, Optional
 
 from pydantic import BaseModel, Field
 
@@ -13,9 +13,11 @@ from invokeai.app.invocations.baseinvocation import (
 from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.shared.models import FreeUConfig
+from invokeai.backend.flux.util import max_seq_lengths
 from invokeai.backend.model_manager.config import (
     AnyModelConfig,
     BaseModelType,
+    CheckpointConfigBase,
     ModelType,
     SubModelType,
 )
@@ -137,6 +139,78 @@ class ModelIdentifierInvocation(BaseInvocation):
         return ModelIdentifierOutput(model=self.model)
 
 
+@invocation_output("flux_model_loader_output")
+class FluxModelLoaderOutput(BaseInvocationOutput):
+    """Flux base model loader output"""
+
+    transformer: TransformerField = OutputField(description=FieldDescriptions.transformer, title="Transformer")
+    clip: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP")
+    t5_encoder: T5EncoderField = OutputField(description=FieldDescriptions.t5_encoder, title="T5 Encoder")
+    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
+    max_seq_len: Literal[256, 512] = OutputField(
+        description="The max sequence length to used for the T5 encoder. (256 for schnell transformer, 512 for dev transformer)",
+        title="Max Seq Length",
+    )
+
+
+@invocation(
+    "flux_model_loader",
+    title="Flux Main Model",
+    tags=["model", "flux"],
+    category="model",
+    version="1.0.4",
+    classification=Classification.Prototype,
+)
+class FluxModelLoaderInvocation(BaseInvocation):
+    """Loads a flux base model, outputting its submodels."""
+
+    model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.flux_model,
+        ui_type=UIType.FluxMainModel,
+        input=Input.Direct,
+    )
+
+    t5_encoder_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.t5_encoder, ui_type=UIType.T5EncoderModel, input=Input.Direct, title="T5 Encoder"
+    )
+
+    clip_embed_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.clip_embed_model,
+        ui_type=UIType.CLIPEmbedModel,
+        input=Input.Direct,
+        title="CLIP Embed",
+    )
+
+    vae_model: ModelIdentifierField = InputField(
+        description=FieldDescriptions.vae_model, ui_type=UIType.FluxVAEModel, title="VAE"
+    )
+
+    def invoke(self, context: InvocationContext) -> FluxModelLoaderOutput:
+        for key in [self.model.key, self.t5_encoder_model.key, self.clip_embed_model.key, self.vae_model.key]:
+            if not context.models.exists(key):
+                raise ValueError(f"Unknown model: {key}")
+
+        transformer = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
+        vae = self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
+
+        tokenizer = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
+        clip_encoder = self.clip_embed_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
+
+        tokenizer2 = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
+        t5_encoder = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
+
+        transformer_config = context.models.get_config(transformer)
+        assert isinstance(transformer_config, CheckpointConfigBase)
+
+        return FluxModelLoaderOutput(
+            transformer=TransformerField(transformer=transformer, loras=[]),
+            clip=CLIPField(tokenizer=tokenizer, text_encoder=clip_encoder, loras=[], skipped_layers=0),
+            t5_encoder=T5EncoderField(tokenizer=tokenizer2, text_encoder=t5_encoder),
+            vae=VAEField(vae=vae),
+            max_seq_len=max_seq_lengths[transformer_config.config_path],
+        )
+
+
 @invocation(
     "main_model_loader",
     title="Main Model",

invokeai/app/invocations/sd3_model_loader.py

@@ -1,102 +0,0 @@
-from invokeai.app.invocations.baseinvocation import (
-    BaseInvocation,
-    BaseInvocationOutput,
-    Classification,
-    invocation,
-    invocation_output,
-)
-from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
-from invokeai.app.invocations.model import CLIPField, ModelIdentifierField, T5EncoderField, TransformerField, VAEField
-from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.model_manager.config import CheckpointConfigBase, SubModelType
-
-
-@invocation_output("sd3_model_loader_output")
-class Sd3ModelLoaderOutput(BaseInvocationOutput):
-    """SD3 base model loader output."""
-
-    mmditx: TransformerField = OutputField(description=FieldDescriptions.mmditx, title="MMDiTX")
-    clip_l: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP L")
-    clip_g: CLIPField = OutputField(description=FieldDescriptions.clip, title="CLIP G")
-    t5_encoder: T5EncoderField = OutputField(description=FieldDescriptions.t5_encoder, title="T5 Encoder")
-    vae: VAEField = OutputField(description=FieldDescriptions.vae, title="VAE")
-
-
-@invocation(
-    "sd3_model_loader",
-    title="SD3 Main Model",
-    tags=["model", "sd3"],
-    category="model",
-    version="1.0.0",
-    classification=Classification.Prototype,
-)
-class Sd3ModelLoaderInvocation(BaseInvocation):
-    """Loads a SD3 base model, outputting its submodels."""
-
-    # TODO(ryand): Create a UIType.Sd3MainModelField to use here.
-    model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.sd3_model,
-        ui_type=UIType.MainModel,
-        input=Input.Direct,
-    )
-
-    # TODO(ryand): Make the text encoders optional.
-    # Note: The text encoders are optional for SD3. The model was trained with dropout, so any can be left out at
-    # inference time. Typically, only the T5 encoder is omitted, since it is the largest by far.
-    t5_encoder_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.t5_encoder, ui_type=UIType.T5EncoderModel, input=Input.Direct, title="T5 Encoder"
-    )
-
-    clip_l_embed_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.clip_embed_model,
-        ui_type=UIType.CLIPEmbedModel,
-        input=Input.Direct,
-        title="CLIP L Embed",
-    )
-
-    clip_g_embed_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.clip_embed_model,
-        ui_type=UIType.CLIPEmbedModel,
-        input=Input.Direct,
-        title="CLIP G Embed",
-    )
-
-    # TODO(ryand): Create a UIType.Sd3VaModelField to use here.
-    vae_model: ModelIdentifierField = InputField(
-        description=FieldDescriptions.vae_model, ui_type=UIType.VAEModel, title="VAE"
-    )
-
-    def invoke(self, context: InvocationContext) -> Sd3ModelLoaderOutput:
-        for key in [
-            self.model.key,
-            self.t5_encoder_model.key,
-            self.clip_l_embed_model.key,
-            self.clip_g_embed_model.key,
-            self.vae_model.key,
-        ]:
-            if not context.models.exists(key):
-                raise ValueError(f"Unknown model: {key}")
-
-        # TODO(ryand): Figure out the sub-model types for SD3.
-        mmditx = self.model.model_copy(update={"submodel_type": SubModelType.Transformer})
-        vae = self.vae_model.model_copy(update={"submodel_type": SubModelType.VAE})
-
-        tokenizer_l = self.clip_l_embed_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
-        clip_encoder_l = self.clip_l_embed_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
-
-        tokenizer_g = self.clip_g_embed_model.model_copy(update={"submodel_type": SubModelType.Tokenizer})
-        clip_encoder_g = self.clip_g_embed_model.model_copy(update={"submodel_type": SubModelType.TextEncoder})
-
-        tokenizer_t5 = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.Tokenizer2})
-        t5_encoder = self.t5_encoder_model.model_copy(update={"submodel_type": SubModelType.TextEncoder2})
-
-        transformer_config = context.models.get_config(mmditx)
-        assert isinstance(transformer_config, CheckpointConfigBase)
-
-        return Sd3ModelLoaderOutput(
-            mmditx=TransformerField(transformer=mmditx, loras=[]),
-            clip_l=CLIPField(tokenizer=tokenizer_l, text_encoder=clip_encoder_l, loras=[], skipped_layers=0),
-            clip_g=CLIPField(tokenizer=tokenizer_g, text_encoder=clip_encoder_g, loras=[], skipped_layers=0),
-            t5_encoder=T5EncoderField(tokenizer=tokenizer_t5, text_encoder=t5_encoder),
-            vae=VAEField(vae=vae),
-        )

invokeai/backend/model_manager/config.py

@@ -53,8 +53,6 @@ class BaseModelType(str, Enum):
     Any = "any"
     StableDiffusion1 = "sd-1"
     StableDiffusion2 = "sd-2"
-    # TODO(ryand): Should this just be StableDiffusion3?
-    StableDiffusion35 = "sd-3.5"
     StableDiffusionXL = "sdxl"
     StableDiffusionXLRefiner = "sdxl-refiner"
     Flux = "flux"

invokeai/backend/model_manager/load/model_loaders/sd3.py

@@ -1,55 +0,0 @@
-from pathlib import Path
-from typing import Optional
-
-from invokeai.backend.model_manager.config import (
-    AnyModel,
-    AnyModelConfig,
-    BaseModelType,
-    CheckpointConfigBase,
-    MainCheckpointConfig,
-    ModelFormat,
-    ModelType,
-    SubModelType,
-)
-from invokeai.backend.model_manager.load.load_default import ModelLoader
-from invokeai.backend.model_manager.load.model_loader_registry import ModelLoaderRegistry
-
-
-@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion35, type=ModelType.Main, format=ModelFormat.Checkpoint)
-class FluxCheckpointModel(ModelLoader):
-    """Class to load main models."""
-
-    def _load_model(
-        self,
-        config: AnyModelConfig,
-        submodel_type: Optional[SubModelType] = None,
-    ) -> AnyModel:
-        if not isinstance(config, CheckpointConfigBase):
-            raise ValueError("Only CheckpointConfigBase models are currently supported here.")
-
-        match submodel_type:
-            case SubModelType.Transformer:
-                return self._load_from_singlefile(config)
-
-        raise ValueError(
-            f"Only Transformer submodels are currently supported. Received: {submodel_type.value if submodel_type else 'None'}"
-        )
-
-    def _load_from_singlefile(
-        self,
-        config: AnyModelConfig,
-    ) -> AnyModel:
-        assert isinstance(config, MainCheckpointConfig)
-        model_path = Path(config.path)
-
-        # model = Flux(params[config.config_path])
-        # sd = load_file(model_path)
-        # if "model.diffusion_model.double_blocks.0.img_attn.norm.key_norm.scale" in sd:
-        #     sd = convert_bundle_to_flux_transformer_checkpoint(sd)
-        # new_sd_size = sum([ten.nelement() * torch.bfloat16.itemsize for ten in sd.values()])
-        # self._ram_cache.make_room(new_sd_size)
-        # for k in sd.keys():
-        #     # We need to cast to bfloat16 due to it being the only currently supported dtype for inference
-        #     sd[k] = sd[k].to(torch.bfloat16)
-        # model.load_state_dict(sd, assign=True)
-        return model

invokeai/backend/model_manager/probe.py

@@ -37,7 +37,6 @@ from invokeai.backend.model_manager.config import (
 from invokeai.backend.model_manager.util.model_util import lora_token_vector_length, read_checkpoint_meta
 from invokeai.backend.quantization.gguf.ggml_tensor import GGMLTensor
 from invokeai.backend.quantization.gguf.loaders import gguf_sd_loader
-from invokeai.backend.sd3.sd3_state_dict_utils import is_sd3_checkpoint
 from invokeai.backend.spandrel_image_to_image_model import SpandrelImageToImageModel
 from invokeai.backend.util.silence_warnings import SilenceWarnings
 
@@ -121,7 +120,6 @@ class ModelProbe(object):
         "T2IAdapter": ModelType.T2IAdapter,
         "CLIPModel": ModelType.CLIPEmbed,
         "CLIPTextModel": ModelType.CLIPEmbed,
-        "CLIPTextModelWithProjection": ModelType.CLIPEmbed,
         "T5EncoderModel": ModelType.T5Encoder,
         "FluxControlNetModel": ModelType.ControlNet,
     }
@@ -243,11 +241,6 @@ class ModelProbe(object):
         for key in [str(k) for k in ckpt.keys()]:
             if key.startswith(
                 (
-                    # The following prefixes appear when multiple models have been bundled together in a single file (I
-                    # believe the format originated in ComfyUI).
-                    # first_stage_model = VAE
-                    # cond_stage_model = Text Encoder
-                    # model.diffusion_model = UNet / Transformer
                     "cond_stage_model.",
                     "first_stage_model.",
                     "model.diffusion_model.",
@@ -404,9 +397,6 @@ class ModelProbe(object):
                     #   is used rather than attempting to support flux with separate model types and format
                     #   If changed in the future, please fix me
                     config_file = "flux-schnell"
-            elif base_type == BaseModelType.StableDiffusion35:
-                # TODO(ryand): Think about what to do here.
-                config_file = "sd3.5-large"
             else:
                 config_file = LEGACY_CONFIGS[base_type][variant_type]
                 if isinstance(config_file, dict):  # need another tier for sd-2.x models
@@ -472,8 +462,9 @@ MODEL_NAME_TO_PREPROCESSOR = {
     "normal": "normalbae_image_processor",
     "sketch": "pidi_image_processor",
     "scribble": "lineart_image_processor",
-    "lineart": "lineart_image_processor",
+    "lineart anime": "lineart_anime_image_processor",
     "lineart_anime": "lineart_anime_image_processor",
+    "lineart": "lineart_image_processor",
     "softedge": "hed_image_processor",
     "hed": "hed_image_processor",
     "shuffle": "content_shuffle_image_processor",
@@ -526,7 +517,7 @@ class CheckpointProbeBase(ProbeBase):
     def get_variant_type(self) -> ModelVariantType:
         model_type = ModelProbe.get_model_type_from_checkpoint(self.model_path, self.checkpoint)
         base_type = self.get_base_type()
-        if model_type != ModelType.Main or base_type in (BaseModelType.Flux, BaseModelType.StableDiffusion35):
+        if model_type != ModelType.Main or base_type == BaseModelType.Flux:
             return ModelVariantType.Normal
         state_dict = self.checkpoint.get("state_dict") or self.checkpoint
         in_channels = state_dict["model.diffusion_model.input_blocks.0.0.weight"].shape[1]
@@ -551,10 +542,6 @@ class PipelineCheckpointProbe(CheckpointProbeBase):
             or "model.diffusion_model.double_blocks.0.img_attn.norm.key_norm.scale" in state_dict
         ):
             return BaseModelType.Flux
-
-        if is_sd3_checkpoint(state_dict):
-            return BaseModelType.StableDiffusion35
-
         key_name = "model.diffusion_model.input_blocks.2.1.transformer_blocks.0.attn2.to_k.weight"
         if key_name in state_dict and state_dict[key_name].shape[-1] == 768:
             return BaseModelType.StableDiffusion1

invokeai/backend/model_manager/starter_models.py

@@ -13,6 +13,9 @@ class StarterModelWithoutDependencies(BaseModel):
     type: ModelType
     format: Optional[ModelFormat] = None
     is_installed: bool = False
+    # allows us to track what models a user has installed across name changes within starter models
+    # if you update a starter model name, please add the old one to this list for that starter model
+    previous_names: list[str] = []
 
 
 class StarterModel(StarterModelWithoutDependencies):
@@ -243,44 +246,49 @@ easy_neg_sd1 = StarterModel(
 # endregion
 # region IP Adapter
 ip_adapter_sd1 = StarterModel(
-    name="IP Adapter",
+    name="Standard Reference (IP Adapter)",
     base=BaseModelType.StableDiffusion1,
     source="https://huggingface.co/InvokeAI/ip_adapter_sd15/resolve/main/ip-adapter_sd15.safetensors",
-    description="IP-Adapter for SD 1.5 models",
+    description="References images with a more generalized/looser degree of precision.",
     type=ModelType.IPAdapter,
     dependencies=[ip_adapter_sd_image_encoder],
+    previous_names=["IP Adapter"],
 )
 ip_adapter_plus_sd1 = StarterModel(
-    name="IP Adapter Plus",
+    name="Precise Reference (IP Adapter Plus)",
     base=BaseModelType.StableDiffusion1,
     source="https://huggingface.co/InvokeAI/ip_adapter_plus_sd15/resolve/main/ip-adapter-plus_sd15.safetensors",
-    description="Refined IP-Adapter for SD 1.5 models",
+    description="References images with a higher degree of precision.",
     type=ModelType.IPAdapter,
     dependencies=[ip_adapter_sd_image_encoder],
+    previous_names=["IP Adapter Plus"],
 )
 ip_adapter_plus_face_sd1 = StarterModel(
-    name="IP Adapter Plus Face",
+    name="Face Reference (IP Adapter Plus Face)",
     base=BaseModelType.StableDiffusion1,
     source="https://huggingface.co/InvokeAI/ip_adapter_plus_face_sd15/resolve/main/ip-adapter-plus-face_sd15.safetensors",
-    description="Refined IP-Adapter for SD 1.5 models, adapted for faces",
+    description="References images with a higher degree of precision, adapted for faces",
     type=ModelType.IPAdapter,
     dependencies=[ip_adapter_sd_image_encoder],
+    previous_names=["IP Adapter Plus Face"],
 )
 ip_adapter_sdxl = StarterModel(
-    name="IP Adapter SDXL",
+    name="Standard Reference (IP Adapter ViT-H)",
     base=BaseModelType.StableDiffusionXL,
     source="https://huggingface.co/InvokeAI/ip_adapter_sdxl_vit_h/resolve/main/ip-adapter_sdxl_vit-h.safetensors",
-    description="IP-Adapter for SDXL models",
+    description="References images with a higher degree of precision.",
     type=ModelType.IPAdapter,
     dependencies=[ip_adapter_sdxl_image_encoder],
+    previous_names=["IP Adapter SDXL"],
 )
 ip_adapter_flux = StarterModel(
-    name="XLabs FLUX IP-Adapter",
+    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",
-    description="FLUX IP-Adapter",
+    description="References images with a more generalized/looser degree of precision.",
     type=ModelType.IPAdapter,
     dependencies=[clip_vit_l_image_encoder],
+    previous_names=["XLabs FLUX IP-Adapter"],
 )
 # endregion
 # region ControlNet
@@ -299,157 +307,162 @@ qr_code_cnet_sdxl = StarterModel(
     type=ModelType.ControlNet,
 )
 canny_sd1 = StarterModel(
-    name="canny",
+    name="Hard Edge Detection (canny)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_canny",
-    description="ControlNet weights trained on sd-1.5 with canny conditioning.",
+    description="Uses detected edges in the image to control composition.",
     type=ModelType.ControlNet,
+    previous_names=["canny"],
 )
 inpaint_cnet_sd1 = StarterModel(
-    name="inpaint",
+    name="Inpainting",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_inpaint",
     description="ControlNet weights trained on sd-1.5 with canny conditioning, inpaint version",
     type=ModelType.ControlNet,
+    previous_names=["inpaint"],
 )
 mlsd_sd1 = StarterModel(
-    name="mlsd",
+    name="Line Drawing (mlsd)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_mlsd",
-    description="ControlNet weights trained on sd-1.5 with canny conditioning, MLSD version",
+    description="Uses straight line detection for controlling the generation.",
     type=ModelType.ControlNet,
+    previous_names=["mlsd"],
 )
 depth_sd1 = StarterModel(
-    name="depth",
+    name="Depth Map",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11f1p_sd15_depth",
-    description="ControlNet weights trained on sd-1.5 with depth conditioning",
+    description="Uses depth information in the image to control the depth in the generation.",
     type=ModelType.ControlNet,
+    previous_names=["depth"],
 )
 normal_bae_sd1 = StarterModel(
-    name="normal_bae",
+    name="Lighting Detection (Normals)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_normalbae",
-    description="ControlNet weights trained on sd-1.5 with normalbae image conditioning",
+    description="Uses detected lighting information to guide the lighting of the composition.",
     type=ModelType.ControlNet,
+    previous_names=["normal_bae"],
 )
 seg_sd1 = StarterModel(
-    name="seg",
+    name="Segmentation Map",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_seg",
-    description="ControlNet weights trained on sd-1.5 with seg image conditioning",
+    description="Uses segmentation maps to guide the structure of the composition.",
     type=ModelType.ControlNet,
+    previous_names=["seg"],
 )
 lineart_sd1 = StarterModel(
-    name="lineart",
+    name="Lineart",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_lineart",
-    description="ControlNet weights trained on sd-1.5 with lineart image conditioning",
+    description="Uses lineart detection to guide the lighting of the composition.",
     type=ModelType.ControlNet,
+    previous_names=["lineart"],
 )
 lineart_anime_sd1 = StarterModel(
-    name="lineart_anime",
+    name="Lineart Anime",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15s2_lineart_anime",
-    description="ControlNet weights trained on sd-1.5 with anime image conditioning",
+    description="Uses anime lineart detection to guide the lighting of the composition.",
     type=ModelType.ControlNet,
+    previous_names=["lineart_anime"],
 )
 openpose_sd1 = StarterModel(
-    name="openpose",
+    name="Pose Detection (openpose)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_openpose",
-    description="ControlNet weights trained on sd-1.5 with openpose image conditioning",
+    description="Uses pose information to control the pose of human characters in the generation.",
     type=ModelType.ControlNet,
+    previous_names=["openpose"],
 )
 scribble_sd1 = StarterModel(
-    name="scribble",
+    name="Contour Detection (scribble)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_scribble",
-    description="ControlNet weights trained on sd-1.5 with scribble image conditioning",
+    description="Uses edges, contours, or line art in the image to control composition.",
     type=ModelType.ControlNet,
+    previous_names=["scribble"],
 )
 softedge_sd1 = StarterModel(
-    name="softedge",
+    name="Soft Edge Detection (softedge)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11p_sd15_softedge",
-    description="ControlNet weights trained on sd-1.5 with soft edge conditioning",
+    description="Uses a soft edge detection map to control composition.",
     type=ModelType.ControlNet,
+    previous_names=["softedge"],
 )
 shuffle_sd1 = StarterModel(
-    name="shuffle",
+    name="Remix (shuffle)",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11e_sd15_shuffle",
     description="ControlNet weights trained on sd-1.5 with shuffle image conditioning",
     type=ModelType.ControlNet,
+    previous_names=["shuffle"],
 )
 tile_sd1 = StarterModel(
-    name="tile",
+    name="Tile",
     base=BaseModelType.StableDiffusion1,
     source="lllyasviel/control_v11f1e_sd15_tile",
-    description="ControlNet weights trained on sd-1.5 with tiled image conditioning",
-    type=ModelType.ControlNet,
-)
-ip2p_sd1 = StarterModel(
-    name="ip2p",
-    base=BaseModelType.StableDiffusion1,
-    source="lllyasviel/control_v11e_sd15_ip2p",
-    description="ControlNet weights trained on sd-1.5 with ip2p conditioning.",
+    description="Uses image data to replicate exact colors/structure in the resulting generation.",
     type=ModelType.ControlNet,
+    previous_names=["tile"],
 )
 canny_sdxl = StarterModel(
-    name="canny-sdxl",
+    name="Hard Edge Detection (canny)",
     base=BaseModelType.StableDiffusionXL,
     source="xinsir/controlNet-canny-sdxl-1.0",
-    description="ControlNet weights trained on sdxl-1.0 with canny conditioning, by Xinsir.",
+    description="Uses detected edges in the image to control composition.",
     type=ModelType.ControlNet,
+    previous_names=["canny-sdxl"],
 )
 depth_sdxl = StarterModel(
-    name="depth-sdxl",
+    name="Depth Map",
     base=BaseModelType.StableDiffusionXL,
     source="diffusers/controlNet-depth-sdxl-1.0",
-    description="ControlNet weights trained on sdxl-1.0 with depth conditioning.",
+    description="Uses depth information in the image to control the depth in the generation.",
     type=ModelType.ControlNet,
+    previous_names=["depth-sdxl"],
 )
 softedge_sdxl = StarterModel(
-    name="softedge-dexined-sdxl",
+    name="Soft Edge Detection (softedge)",
     base=BaseModelType.StableDiffusionXL,
     source="SargeZT/controlNet-sd-xl-1.0-softedge-dexined",
-    description="ControlNet weights trained on sdxl-1.0 with dexined soft edge preprocessing.",
-    type=ModelType.ControlNet,
-)
-depth_zoe_16_sdxl = StarterModel(
-    name="depth-16bit-zoe-sdxl",
-    base=BaseModelType.StableDiffusionXL,
-    source="SargeZT/controlNet-sd-xl-1.0-depth-16bit-zoe",
-    description="ControlNet weights trained on sdxl-1.0 with Zoe's preprocessor (16 bits).",
-    type=ModelType.ControlNet,
-)
-depth_zoe_32_sdxl = StarterModel(
-    name="depth-zoe-sdxl",
-    base=BaseModelType.StableDiffusionXL,
-    source="diffusers/controlNet-zoe-depth-sdxl-1.0",
-    description="ControlNet weights trained on sdxl-1.0 with Zoe's preprocessor (32 bits).",
+    description="Uses a soft edge detection map to control composition.",
     type=ModelType.ControlNet,
+    previous_names=["softedge-dexined-sdxl"],
 )
 openpose_sdxl = StarterModel(
-    name="openpose-sdxl",
+    name="Pose Detection (openpose)",
     base=BaseModelType.StableDiffusionXL,
     source="xinsir/controlNet-openpose-sdxl-1.0",
-    description="ControlNet weights trained on sdxl-1.0 compatible with the DWPose processor by Xinsir.",
+    description="Uses pose information to control the pose of human characters in the generation.",
     type=ModelType.ControlNet,
+    previous_names=["openpose-sdxl", "controlnet-openpose-sdxl"],
 )
 scribble_sdxl = StarterModel(
-    name="scribble-sdxl",
+    name="Contour Detection (scribble)",
     base=BaseModelType.StableDiffusionXL,
     source="xinsir/controlNet-scribble-sdxl-1.0",
-    description="ControlNet weights trained on sdxl-1.0 compatible with various lineart processors and black/white sketches by Xinsir.",
+    description="Uses edges, contours, or line art in the image to control composition.",
     type=ModelType.ControlNet,
+    previous_names=["scribble-sdxl", "controlnet-scribble-sdxl"],
 )
 tile_sdxl = StarterModel(
-    name="tile-sdxl",
+    name="Tile",
     base=BaseModelType.StableDiffusionXL,
     source="xinsir/controlNet-tile-sdxl-1.0",
-    description="ControlNet weights trained on sdxl-1.0 with tiled image conditioning",
+    description="Uses image data to replicate exact colors/structure in the resulting generation.",
+    type=ModelType.ControlNet,
+    previous_names=["tile-sdxl"],
+)
+union_cnet_sdxl = StarterModel(
+    name="Multi-Guidance Detection (Union Pro)",
+    base=BaseModelType.StableDiffusionXL,
+    source="InvokeAI/Xinsir-SDXL_Controlnet_Union",
+    description="A unified ControlNet for SDXL model that supports 10+ control types",
     type=ModelType.ControlNet,
 )
 union_cnet_flux = StarterModel(
@@ -462,60 +475,52 @@ union_cnet_flux = StarterModel(
 # endregion
 # region T2I Adapter
 t2i_canny_sd1 = StarterModel(
-    name="canny-sd15",
+    name="Hard Edge Detection (canny)",
     base=BaseModelType.StableDiffusion1,
     source="TencentARC/t2iadapter_canny_sd15v2",
-    description="T2I Adapter weights trained on sd-1.5 with canny conditioning.",
+    description="Uses detected edges in the image to control composition",
     type=ModelType.T2IAdapter,
+    previous_names=["canny-sd15"],
 )
 t2i_sketch_sd1 = StarterModel(
-    name="sketch-sd15",
+    name="Sketch",
     base=BaseModelType.StableDiffusion1,
     source="TencentARC/t2iadapter_sketch_sd15v2",
-    description="T2I Adapter weights trained on sd-1.5 with sketch conditioning.",
+    description="Uses a sketch to control composition",
     type=ModelType.T2IAdapter,
+    previous_names=["sketch-sd15"],
 )
 t2i_depth_sd1 = StarterModel(
-    name="depth-sd15",
+    name="Depth Map",
     base=BaseModelType.StableDiffusion1,
     source="TencentARC/t2iadapter_depth_sd15v2",
-    description="T2I Adapter weights trained on sd-1.5 with depth conditioning.",
-    type=ModelType.T2IAdapter,
-)
-t2i_zoe_depth_sd1 = StarterModel(
-    name="zoedepth-sd15",
-    base=BaseModelType.StableDiffusion1,
-    source="TencentARC/t2iadapter_zoedepth_sd15v1",
-    description="T2I Adapter weights trained on sd-1.5 with zoe depth conditioning.",
+    description="Uses depth information in the image to control the depth in the generation.",
     type=ModelType.T2IAdapter,
+    previous_names=["depth-sd15"],
 )
 t2i_canny_sdxl = StarterModel(
-    name="canny-sdxl",
+    name="Hard Edge Detection (canny)",
     base=BaseModelType.StableDiffusionXL,
     source="TencentARC/t2i-adapter-canny-sdxl-1.0",
-    description="T2I Adapter weights trained on sdxl-1.0 with canny conditioning.",
-    type=ModelType.T2IAdapter,
-)
-t2i_zoe_depth_sdxl = StarterModel(
-    name="zoedepth-sdxl",
-    base=BaseModelType.StableDiffusionXL,
-    source="TencentARC/t2i-adapter-depth-zoe-sdxl-1.0",
-    description="T2I Adapter weights trained on sdxl-1.0 with zoe depth conditioning.",
+    description="Uses detected edges in the image to control composition",
     type=ModelType.T2IAdapter,
+    previous_names=["canny-sdxl"],
 )
 t2i_lineart_sdxl = StarterModel(
-    name="lineart-sdxl",
+    name="Lineart",
     base=BaseModelType.StableDiffusionXL,
     source="TencentARC/t2i-adapter-lineart-sdxl-1.0",
-    description="T2I Adapter weights trained on sdxl-1.0 with lineart conditioning.",
+    description="Uses lineart detection to guide the lighting of the composition.",
     type=ModelType.T2IAdapter,
+    previous_names=["lineart-sdxl"],
 )
 t2i_sketch_sdxl = StarterModel(
-    name="sketch-sdxl",
+    name="Sketch",
     base=BaseModelType.StableDiffusionXL,
     source="TencentARC/t2i-adapter-sketch-sdxl-1.0",
-    description="T2I Adapter weights trained on sdxl-1.0 with sketch conditioning.",
+    description="Uses a sketch to control composition",
     type=ModelType.T2IAdapter,
+    previous_names=["sketch-sdxl"],
 )
 # endregion
 # region SpandrelImageToImage
@@ -600,22 +605,18 @@ STARTER_MODELS: list[StarterModel] = [
     softedge_sd1,
     shuffle_sd1,
     tile_sd1,
-    ip2p_sd1,
     canny_sdxl,
     depth_sdxl,
     softedge_sdxl,
-    depth_zoe_16_sdxl,
-    depth_zoe_32_sdxl,
     openpose_sdxl,
     scribble_sdxl,
     tile_sdxl,
+    union_cnet_sdxl,
     union_cnet_flux,
     t2i_canny_sd1,
     t2i_sketch_sd1,
     t2i_depth_sd1,
-    t2i_zoe_depth_sd1,
     t2i_canny_sdxl,
-    t2i_zoe_depth_sdxl,
     t2i_lineart_sdxl,
     t2i_sketch_sdxl,
     realesrgan_x4,
@@ -646,7 +647,6 @@ sd1_bundle: list[StarterModel] = [
     softedge_sd1,
     shuffle_sd1,
     tile_sd1,
-    ip2p_sd1,
     swinir,
 ]
 
@@ -657,8 +657,6 @@ sdxl_bundle: list[StarterModel] = [
     canny_sdxl,
     depth_sdxl,
     softedge_sdxl,
-    depth_zoe_16_sdxl,
-    depth_zoe_32_sdxl,
     openpose_sdxl,
     scribble_sdxl,
     tile_sdxl,

invokeai/backend/sd3/mmditx.py

@@ -1,891 +0,0 @@
-# This file was originally copied from:
-# https://github.com/Stability-AI/sd3.5/blob/19bf11c4e1e37324c5aa5a61f010d4127848a09c/mmditx.py
-
-
-### This file contains impls for MM-DiT, the core model component of SD3
-
-import math
-from typing import Dict, List, Optional
-
-import numpy as np
-import torch
-from einops import rearrange, repeat
-
-from invokeai.backend.sd3.other_impls import Mlp, attention
-
-
-class PatchEmbed(torch.nn.Module):
-    """2D Image to Patch Embedding"""
-
-    def __init__(
-        self,
-        img_size: Optional[int] = 224,
-        patch_size: int = 16,
-        in_chans: int = 3,
-        embed_dim: int = 768,
-        flatten: bool = True,
-        bias: bool = True,
-        strict_img_size: bool = True,
-        dynamic_img_pad: bool = False,
-        dtype: torch.dtype | None = None,
-        device: torch.device | None = None,
-    ):
-        super().__init__()
-        self.patch_size = (patch_size, patch_size)
-        if img_size is not None:
-            self.img_size = (img_size, img_size)
-            self.grid_size = tuple([s // p for s, p in zip(self.img_size, self.patch_size, strict=False)])
-            self.num_patches = self.grid_size[0] * self.grid_size[1]
-        else:
-            self.img_size = None
-            self.grid_size = None
-            self.num_patches = None
-
-        # flatten spatial dim and transpose to channels last, kept for bwd compat
-        self.flatten = flatten
-        self.strict_img_size = strict_img_size
-        self.dynamic_img_pad = dynamic_img_pad
-
-        self.proj = torch.nn.Conv2d(
-            in_chans,
-            embed_dim,
-            kernel_size=patch_size,
-            stride=patch_size,
-            bias=bias,
-            dtype=dtype,
-            device=device,
-        )
-
-    def forward(self, x: torch.Tensor) -> torch.Tensor:
-        x = self.proj(x)
-        if self.flatten:
-            x = x.flatten(2).transpose(1, 2)  # NCHW -> NLC
-        return x
-
-
-def modulate(x: torch.Tensor, shift: torch.Tensor | None, scale: torch.Tensor) -> torch.Tensor:
-    if shift is None:
-        shift = torch.zeros_like(scale)
-    return x * (1 + scale.unsqueeze(1)) + shift.unsqueeze(1)
-
-
-#################################################################################
-#                   Sine/Cosine Positional Embedding Functions                  #
-#################################################################################
-
-
-def get_2d_sincos_pos_embed(
-    embed_dim: int,
-    grid_size: int,
-    cls_token: bool = False,
-    extra_tokens: int = 0,
-    scaling_factor: Optional[float] = None,
-    offset: Optional[float] = None,
-):
-    """
-    grid_size: int of the grid height and width
-    return:
-    pos_embed: [grid_size*grid_size, embed_dim] or [1+grid_size*grid_size, embed_dim] (w/ or w/o cls_token)
-    """
-    grid_h = np.arange(grid_size, dtype=np.float32)
-    grid_w = np.arange(grid_size, dtype=np.float32)
-    grid = np.meshgrid(grid_w, grid_h)  # here w goes first
-    grid = np.stack(grid, axis=0)
-    if scaling_factor is not None:
-        grid = grid / scaling_factor
-    if offset is not None:
-        grid = grid - offset
-    grid = grid.reshape([2, 1, grid_size, grid_size])
-    pos_embed = get_2d_sincos_pos_embed_from_grid(embed_dim, grid)
-    if cls_token and extra_tokens > 0:
-        pos_embed = np.concatenate([np.zeros([extra_tokens, embed_dim]), pos_embed], axis=0)
-    return pos_embed
-
-
-def get_2d_sincos_pos_embed_from_grid(embed_dim: int, grid):
-    assert embed_dim % 2 == 0
-    # use half of dimensions to encode grid_h
-    emb_h = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[0])  # (H*W, D/2)
-    emb_w = get_1d_sincos_pos_embed_from_grid(embed_dim // 2, grid[1])  # (H*W, D/2)
-    emb = np.concatenate([emb_h, emb_w], axis=1)  # (H*W, D)
-    return emb
-
-
-def get_1d_sincos_pos_embed_from_grid(embed_dim: int, pos):
-    """
-    embed_dim: output dimension for each position
-    pos: a list of positions to be encoded: size (M,)
-    out: (M, D)
-    """
-    assert embed_dim % 2 == 0
-    omega = np.arange(embed_dim // 2, dtype=np.float64)
-    omega /= embed_dim / 2.0
-    omega = 1.0 / 10000**omega  # (D/2,)
-    pos = pos.reshape(-1)  # (M,)
-    out = np.einsum("m,d->md", pos, omega)  # (M, D/2), outer product
-    emb_sin = np.sin(out)  # (M, D/2)
-    emb_cos = np.cos(out)  # (M, D/2)
-    return np.concatenate([emb_sin, emb_cos], axis=1)  # (M, D)
-
-
-#################################################################################
-#               Embedding Layers for Timesteps and Class Labels                 #
-#################################################################################
-
-
-class TimestepEmbedder(torch.nn.Module):
-    """Embeds scalar timesteps into vector representations."""
-
-    def __init__(self, hidden_size, frequency_embedding_size=256, dtype=None, device=None):
-        super().__init__()
-        self.mlp = torch.nn.Sequential(
-            torch.nn.Linear(
-                frequency_embedding_size,
-                hidden_size,
-                bias=True,
-                dtype=dtype,
-                device=device,
-            ),
-            torch.nn.SiLU(),
-            torch.nn.Linear(hidden_size, hidden_size, bias=True, dtype=dtype, device=device),
-        )
-        self.frequency_embedding_size = frequency_embedding_size
-
-    @staticmethod
-    def timestep_embedding(t, dim, max_period=10000):
-        """
-        Create sinusoidal timestep embeddings.
-        :param t: a 1-D Tensor of N indices, one per batch element.
-                          These may be fractional.
-        :param dim: the dimension of the output.
-        :param max_period: controls the minimum frequency of the embeddings.
-        :return: an (N, D) Tensor of positional embeddings.
-        """
-        half = dim // 2
-        freqs = torch.exp(-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half).to(
-            device=t.device
-        )
-        args = t[:, None].float() * freqs[None]
-        embedding = torch.cat([torch.cos(args), torch.sin(args)], dim=-1)
-        if dim % 2:
-            embedding = torch.cat([embedding, torch.zeros_like(embedding[:, :1])], dim=-1)
-        if torch.is_floating_point(t):
-            embedding = embedding.to(dtype=t.dtype)
-        return embedding
-
-    def forward(self, t, dtype, **kwargs):
-        t_freq = self.timestep_embedding(t, self.frequency_embedding_size).to(dtype)
-        t_emb = self.mlp(t_freq)
-        return t_emb
-
-
-class VectorEmbedder(torch.nn.Module):
-    """Embeds a flat vector of dimension input_dim"""
-
-    def __init__(self, input_dim: int, hidden_size: int, dtype=None, device=None):
-        super().__init__()
-        self.mlp = torch.nn.Sequential(
-            torch.nn.Linear(input_dim, hidden_size, bias=True, dtype=dtype, device=device),
-            torch.nn.SiLU(),
-            torch.nn.Linear(hidden_size, hidden_size, bias=True, dtype=dtype, device=device),
-        )
-
-    def forward(self, x: torch.Tensor) -> torch.Tensor:
-        return self.mlp(x)
-
-
-#################################################################################
-#                                 Core DiT Model                                #
-#################################################################################
-
-
-def split_qkv(qkv, head_dim):
-    qkv = qkv.reshape(qkv.shape[0], qkv.shape[1], 3, -1, head_dim).movedim(2, 0)
-    return qkv[0], qkv[1], qkv[2]
-
-
-def optimized_attention(qkv, num_heads):
-    return attention(qkv[0], qkv[1], qkv[2], num_heads)
-
-
-class SelfAttention(torch.nn.Module):
-    ATTENTION_MODES = ("xformers", "torch", "torch-hb", "math", "debug")
-
-    def __init__(
-        self,
-        dim: int,
-        num_heads: int = 8,
-        qkv_bias: bool = False,
-        qk_scale: Optional[float] = None,
-        attn_mode: str = "xformers",
-        pre_only: bool = False,
-        qk_norm: Optional[str] = None,
-        rmsnorm: bool = False,
-        dtype=None,
-        device=None,
-    ):
-        super().__init__()
-        self.num_heads = num_heads
-        self.head_dim = dim // num_heads
-
-        self.qkv = torch.nn.Linear(dim, dim * 3, bias=qkv_bias, dtype=dtype, device=device)
-        if not pre_only:
-            self.proj = torch.nn.Linear(dim, dim, dtype=dtype, device=device)
-        assert attn_mode in self.ATTENTION_MODES
-        self.attn_mode = attn_mode
-        self.pre_only = pre_only
-
-        if qk_norm == "rms":
-            self.ln_q = RMSNorm(
-                self.head_dim,
-                elementwise_affine=True,
-                eps=1.0e-6,
-                dtype=dtype,
-                device=device,
-            )
-            self.ln_k = RMSNorm(
-                self.head_dim,
-                elementwise_affine=True,
-                eps=1.0e-6,
-                dtype=dtype,
-                device=device,
-            )
-        elif qk_norm == "ln":
-            self.ln_q = torch.nn.LayerNorm(
-                self.head_dim,
-                elementwise_affine=True,
-                eps=1.0e-6,
-                dtype=dtype,
-                device=device,
-            )
-            self.ln_k = torch.nn.LayerNorm(
-                self.head_dim,
-                elementwise_affine=True,
-                eps=1.0e-6,
-                dtype=dtype,
-                device=device,
-            )
-        elif qk_norm is None:
-            self.ln_q = torch.nn.Identity()
-            self.ln_k = torch.nn.Identity()
-        else:
-            raise ValueError(qk_norm)
-
-    def pre_attention(self, x: torch.Tensor):
-        B, L, C = x.shape
-        qkv = self.qkv(x)
-        q, k, v = split_qkv(qkv, self.head_dim)
-        q = self.ln_q(q).reshape(q.shape[0], q.shape[1], -1)
-        k = self.ln_k(k).reshape(q.shape[0], q.shape[1], -1)
-        return (q, k, v)
-
-    def post_attention(self, x: torch.Tensor) -> torch.Tensor:
-        assert not self.pre_only
-        x = self.proj(x)
-        return x
-
-    def forward(self, x: torch.Tensor) -> torch.Tensor:
-        (q, k, v) = self.pre_attention(x)
-        x = attention(q, k, v, self.num_heads)
-        x = self.post_attention(x)
-        return x
-
-
-class RMSNorm(torch.nn.Module):
-    def __init__(
-        self,
-        dim: int,
-        elementwise_affine: bool = False,
-        eps: float = 1e-6,
-        device=None,
-        dtype=None,
-    ):
-        """
-        Initialize the RMSNorm normalization layer.
-        Args:
-            dim (int): The dimension of the input tensor.
-            eps (float, optional): A small value added to the denominator for numerical stability. Default is 1e-6.
-        Attributes:
-            eps (float): A small value added to the denominator for numerical stability.
-            weight (torch.nn.Parameter): Learnable scaling parameter.
-        """
-        super().__init__()
-        self.eps = eps
-        self.learnable_scale = elementwise_affine
-        if self.learnable_scale:
-            self.weight = torch.nn.Parameter(torch.empty(dim, device=device, dtype=dtype))
-        else:
-            self.register_parameter("weight", None)
-
-    def _norm(self, x):
-        """
-        Apply the RMSNorm normalization to the input tensor.
-        Args:
-            x (torch.Tensor): The input tensor.
-        Returns:
-            torch.Tensor: The normalized tensor.
-        """
-        return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
-
-    def forward(self, x):
-        """
-        Forward pass through the RMSNorm layer.
-        Args:
-            x (torch.Tensor): The input tensor.
-        Returns:
-            torch.Tensor: The output tensor after applying RMSNorm.
-        """
-        x = self._norm(x)
-        if self.learnable_scale:
-            return x * self.weight.to(device=x.device, dtype=x.dtype)
-        else:
-            return x
-
-
-class SwiGLUFeedForward(torch.nn.Module):
-    def __init__(
-        self,
-        dim: int,
-        hidden_dim: int,
-        multiple_of: int,
-        ffn_dim_multiplier: Optional[float] = None,
-    ):
-        """
-        Initialize the FeedForward module.
-
-        Args:
-            dim (int): Input dimension.
-            hidden_dim (int): Hidden dimension of the feedforward layer.
-            multiple_of (int): Value to ensure hidden dimension is a multiple of this value.
-            ffn_dim_multiplier (float, optional): Custom multiplier for hidden dimension. Defaults to None.
-
-        Attributes:
-            w1 (ColumnParallelLinear): Linear transformation for the first layer.
-            w2 (RowParallelLinear): Linear transformation for the second layer.
-            w3 (ColumnParallelLinear): Linear transformation for the third layer.
-
-        """
-        super().__init__()
-        hidden_dim = int(2 * hidden_dim / 3)
-        # custom dim factor multiplier
-        if ffn_dim_multiplier is not None:
-            hidden_dim = int(ffn_dim_multiplier * hidden_dim)
-        hidden_dim = multiple_of * ((hidden_dim + multiple_of - 1) // multiple_of)
-
-        self.w1 = torch.nn.Linear(dim, hidden_dim, bias=False)
-        self.w2 = torch.nn.Linear(hidden_dim, dim, bias=False)
-        self.w3 = torch.nn.Linear(dim, hidden_dim, bias=False)
-
-    def forward(self, x):
-        return self.w2(torch.nn.functional.silu(self.w1(x)) * self.w3(x))
-
-
-class DismantledBlock(torch.nn.Module):
-    """A DiT block with gated adaptive layer norm (adaLN) conditioning."""
-
-    ATTENTION_MODES = ("xformers", "torch", "torch-hb", "math", "debug")
-
-    def __init__(
-        self,
-        hidden_size: int,
-        num_heads: int,
-        mlp_ratio: float = 4.0,
-        attn_mode: str = "xformers",
-        qkv_bias: bool = False,
-        pre_only: bool = False,
-        rmsnorm: bool = False,
-        scale_mod_only: bool = False,
-        swiglu: bool = False,
-        qk_norm: Optional[str] = None,
-        x_block_self_attn: bool = False,
-        dtype=None,
-        device=None,
-        **block_kwargs,
-    ):
-        super().__init__()
-        assert attn_mode in self.ATTENTION_MODES
-        if not rmsnorm:
-            self.norm1 = torch.nn.LayerNorm(
-                hidden_size,
-                elementwise_affine=False,
-                eps=1e-6,
-                dtype=dtype,
-                device=device,
-            )
-        else:
-            self.norm1 = RMSNorm(hidden_size, elementwise_affine=False, eps=1e-6)
-        self.attn = SelfAttention(
-            dim=hidden_size,
-            num_heads=num_heads,
-            qkv_bias=qkv_bias,
-            attn_mode=attn_mode,
-            pre_only=pre_only,
-            qk_norm=qk_norm,
-            rmsnorm=rmsnorm,
-            dtype=dtype,
-            device=device,
-        )
-        if x_block_self_attn:
-            assert not pre_only
-            assert not scale_mod_only
-            self.x_block_self_attn = True
-            self.attn2 = SelfAttention(
-                dim=hidden_size,
-                num_heads=num_heads,
-                qkv_bias=qkv_bias,
-                attn_mode=attn_mode,
-                pre_only=False,
-                qk_norm=qk_norm,
-                rmsnorm=rmsnorm,
-                dtype=dtype,
-                device=device,
-            )
-        else:
-            self.x_block_self_attn = False
-        if not pre_only:
-            if not rmsnorm:
-                self.norm2 = torch.nn.LayerNorm(
-                    hidden_size,
-                    elementwise_affine=False,
-                    eps=1e-6,
-                    dtype=dtype,
-                    device=device,
-                )
-            else:
-                self.norm2 = RMSNorm(hidden_size, elementwise_affine=False, eps=1e-6)
-        mlp_hidden_dim = int(hidden_size * mlp_ratio)
-        if not pre_only:
-            if not swiglu:
-                self.mlp = Mlp(
-                    in_features=hidden_size,
-                    hidden_features=mlp_hidden_dim,
-                    act_layer=torch.nn.GELU(approximate="tanh"),
-                    dtype=dtype,
-                    device=device,
-                )
-            else:
-                self.mlp = SwiGLUFeedForward(dim=hidden_size, hidden_dim=mlp_hidden_dim, multiple_of=256)
-        self.scale_mod_only = scale_mod_only
-        if x_block_self_attn:
-            assert not pre_only
-            assert not scale_mod_only
-            n_mods = 9
-        elif not scale_mod_only:
-            n_mods = 6 if not pre_only else 2
-        else:
-            n_mods = 4 if not pre_only else 1
-        self.adaLN_modulation = torch.nn.Sequential(
-            torch.nn.SiLU(),
-            torch.nn.Linear(hidden_size, n_mods * hidden_size, bias=True, dtype=dtype, device=device),
-        )
-        self.pre_only = pre_only
-
-    def pre_attention(self, x: torch.Tensor, c: torch.Tensor):
-        assert x is not None, "pre_attention called with None input"
-        if not self.pre_only:
-            if not self.scale_mod_only:
-                shift_msa, scale_msa, gate_msa, shift_mlp, scale_mlp, gate_mlp = self.adaLN_modulation(c).chunk(
-                    6, dim=1
-                )
-            else:
-                shift_msa = None
-                shift_mlp = None
-                scale_msa, gate_msa, scale_mlp, gate_mlp = self.adaLN_modulation(c).chunk(4, dim=1)
-            qkv = self.attn.pre_attention(modulate(self.norm1(x), shift_msa, scale_msa))
-            return qkv, (x, gate_msa, shift_mlp, scale_mlp, gate_mlp)
-        else:
-            if not self.scale_mod_only:
-                shift_msa, scale_msa = self.adaLN_modulation(c).chunk(2, dim=1)
-            else:
-                shift_msa = None
-                scale_msa = self.adaLN_modulation(c)
-            qkv = self.attn.pre_attention(modulate(self.norm1(x), shift_msa, scale_msa))
-            return qkv, None
-
-    def post_attention(
-        self,
-        attn: torch.Tensor,
-        x: torch.Tensor,
-        gate_msa: torch.Tensor,
-        shift_mlp: torch.Tensor,
-        scale_mlp: torch.Tensor,
-        gate_mlp: torch.Tensor,
-    ) -> torch.Tensor:
-        assert not self.pre_only
-        x = x + gate_msa.unsqueeze(1) * self.attn.post_attention(attn)
-        x = x + gate_mlp.unsqueeze(1) * self.mlp(modulate(self.norm2(x), shift_mlp, scale_mlp))
-        return x
-
-    def pre_attention_x(
-        self, x: torch.Tensor, c: torch.Tensor
-    ) -> tuple[
-        tuple[torch.Tensor, torch.Tensor, torch.Tensor],
-        tuple[torch.Tensor, torch.Tensor, torch.Tensor],
-        tuple[torch.Tensor, torch.Tensor, torch.Tensor, torch.Tensor, torch.Tensor, torch.Tensor],
-    ]:
-        assert self.x_block_self_attn
-        (
-            shift_msa,
-            scale_msa,
-            gate_msa,
-            shift_mlp,
-            scale_mlp,
-            gate_mlp,
-            shift_msa2,
-            scale_msa2,
-            gate_msa2,
-        ) = self.adaLN_modulation(c).chunk(9, dim=1)
-        x_norm = self.norm1(x)
-        qkv = self.attn.pre_attention(modulate(x_norm, shift_msa, scale_msa))
-        qkv2 = self.attn2.pre_attention(modulate(x_norm, shift_msa2, scale_msa2))
-        return (
-            qkv,
-            qkv2,
-            (
-                x,
-                gate_msa,
-                shift_mlp,
-                scale_mlp,
-                gate_mlp,
-                gate_msa2,
-            ),
-        )
-
-    def post_attention_x(
-        self,
-        attn: torch.Tensor,
-        attn2: torch.Tensor,
-        x: torch.Tensor,
-        gate_msa: torch.Tensor,
-        shift_mlp: torch.Tensor,
-        scale_mlp: torch.Tensor,
-        gate_mlp: torch.Tensor,
-        gate_msa2: torch.Tensor,
-        attn1_dropout: float = 0.0,
-    ):
-        assert not self.pre_only
-        if attn1_dropout > 0.0:
-            # Use torch.bernoulli to implement dropout, only dropout the batch dimension
-            attn1_dropout = torch.bernoulli(torch.full((attn.size(0), 1, 1), 1 - attn1_dropout, device=attn.device))
-            attn_ = gate_msa.unsqueeze(1) * self.attn.post_attention(attn) * attn1_dropout
-        else:
-            attn_ = gate_msa.unsqueeze(1) * self.attn.post_attention(attn)
-        x = x + attn_
-        attn2_ = gate_msa2.unsqueeze(1) * self.attn2.post_attention(attn2)
-        x = x + attn2_
-        mlp_ = gate_mlp.unsqueeze(1) * self.mlp(modulate(self.norm2(x), shift_mlp, scale_mlp))
-        x = x + mlp_
-        return x, (gate_msa, gate_msa2, gate_mlp, attn_, attn2_)
-
-    def forward(self, x: torch.Tensor, c: torch.Tensor):
-        assert not self.pre_only
-        if self.x_block_self_attn:
-            (q, k, v), (q2, k2, v2), intermediates = self.pre_attention_x(x, c)
-            attn = attention(q, k, v, self.attn.num_heads)
-            attn2 = attention(q2, k2, v2, self.attn2.num_heads)
-            return self.post_attention_x(attn, attn2, *intermediates)
-        else:
-            (q, k, v), intermediates = self.pre_attention(x, c)
-            attn = attention(q, k, v, self.attn.num_heads)
-            return self.post_attention(attn, *intermediates)
-
-
-def block_mixing(
-    context: torch.Tensor, x: torch.Tensor, context_block: DismantledBlock, x_block: DismantledBlock, c: torch.Tensor
-):
-    assert context is not None, "block_mixing called with None context"
-    context_qkv, context_intermediates = context_block.pre_attention(context, c)
-
-    if x_block.x_block_self_attn:
-        x_qkv, x_qkv2, x_intermediates = x_block.pre_attention_x(x, c)
-    else:
-        x_qkv, x_intermediates = x_block.pre_attention(x, c)
-
-    o: list[torch.Tensor] = []
-    for t in range(3):
-        o.append(torch.cat((context_qkv[t], x_qkv[t]), dim=1))
-    q, k, v = tuple(o)
-
-    attn = attention(q, k, v, x_block.attn.num_heads)
-    context_attn, x_attn = (
-        attn[:, : context_qkv[0].shape[1]],
-        attn[:, context_qkv[0].shape[1] :],
-    )
-
-    if not context_block.pre_only:
-        context = context_block.post_attention(context_attn, *context_intermediates)
-    else:
-        context = None
-
-    if x_block.x_block_self_attn:
-        x_q2, x_k2, x_v2 = x_qkv2
-        attn2 = attention(x_q2, x_k2, x_v2, x_block.attn2.num_heads)
-    else:
-        x = x_block.post_attention(x_attn, *x_intermediates)
-
-    return context, x
-
-
-class JointBlock(torch.nn.Module):
-    """just a small wrapper to serve as a fsdp unit"""
-
-    def __init__(self, *args, **kwargs):
-        super().__init__()
-        pre_only = kwargs.pop("pre_only")
-        qk_norm = kwargs.pop("qk_norm", None)
-        x_block_self_attn = kwargs.pop("x_block_self_attn", False)
-        self.context_block = DismantledBlock(*args, pre_only=pre_only, qk_norm=qk_norm, **kwargs)
-        self.x_block = DismantledBlock(
-            *args,
-            pre_only=False,
-            qk_norm=qk_norm,
-            x_block_self_attn=x_block_self_attn,
-            **kwargs,
-        )
-
-    def forward(self, *args, **kwargs):
-        return block_mixing(*args, context_block=self.context_block, x_block=self.x_block, **kwargs)
-
-
-class FinalLayer(torch.nn.Module):
-    """
-    The final layer of DiT.
-    """
-
-    def __init__(
-        self,
-        hidden_size: int,
-        patch_size: int,
-        out_channels: int,
-        total_out_channels: Optional[int] = None,
-        dtype: Optional[torch.dtype] = None,
-        device: Optional[torch.device] = None,
-    ):
-        super().__init__()
-        self.norm_final = torch.nn.LayerNorm(
-            hidden_size, elementwise_affine=False, eps=1e-6, dtype=dtype, device=device
-        )
-        self.linear = (
-            torch.nn.Linear(
-                hidden_size,
-                patch_size * patch_size * out_channels,
-                bias=True,
-                dtype=dtype,
-                device=device,
-            )
-            if (total_out_channels is None)
-            else torch.nn.Linear(hidden_size, total_out_channels, bias=True, dtype=dtype, device=device)
-        )
-        self.adaLN_modulation = torch.nn.Sequential(
-            torch.nn.SiLU(),
-            torch.nn.Linear(hidden_size, 2 * hidden_size, bias=True, dtype=dtype, device=device),
-        )
-
-    def forward(self, x: torch.Tensor, c: torch.Tensor) -> torch.Tensor:
-        shift, scale = self.adaLN_modulation(c).chunk(2, dim=1)
-        x = modulate(self.norm_final(x), shift, scale)
-        x = self.linear(x)
-        return x
-
-
-class MMDiTX(torch.nn.Module):
-    """Diffusion model with a Transformer backbone."""
-
-    def __init__(
-        self,
-        input_size: int | None = 32,
-        patch_size: int = 2,
-        in_channels: int = 4,
-        depth: int = 28,
-        mlp_ratio: float = 4.0,
-        learn_sigma: bool = False,
-        adm_in_channels: Optional[int] = None,
-        context_embedder_config: Optional[Dict] = None,
-        register_length: int = 0,
-        attn_mode: str = "torch",
-        rmsnorm: bool = False,
-        scale_mod_only: bool = False,
-        swiglu: bool = False,
-        out_channels: Optional[int] = None,
-        pos_embed_scaling_factor: Optional[float] = None,
-        pos_embed_offset: Optional[float] = None,
-        pos_embed_max_size: Optional[int] = None,
-        num_patches: Optional[int] = None,
-        qk_norm: Optional[str] = None,
-        x_block_self_attn_layers: Optional[List[int]] = None,
-        qkv_bias: bool = True,
-        dtype: Optional[torch.dtype] = None,
-        device: Optional[torch.device] = None,
-        verbose: bool = False,
-    ):
-        super().__init__()
-        if verbose:
-            print(
-                f"mmdit initializing with: {input_size=}, {patch_size=}, {in_channels=}, {depth=}, {mlp_ratio=}, {learn_sigma=}, {adm_in_channels=}, {context_embedder_config=}, {register_length=}, {attn_mode=}, {rmsnorm=}, {scale_mod_only=}, {swiglu=}, {out_channels=}, {pos_embed_scaling_factor=}, {pos_embed_offset=}, {pos_embed_max_size=}, {num_patches=}, {qk_norm=}, {qkv_bias=}, {dtype=}, {device=}"
-            )
-        self.dtype = dtype
-        self.learn_sigma = learn_sigma
-        self.in_channels = in_channels
-        default_out_channels = in_channels * 2 if learn_sigma else in_channels
-        self.out_channels = out_channels if out_channels is not None else default_out_channels
-        self.patch_size = patch_size
-        self.pos_embed_scaling_factor = pos_embed_scaling_factor
-        self.pos_embed_offset = pos_embed_offset
-        self.pos_embed_max_size = pos_embed_max_size
-        self.x_block_self_attn_layers = x_block_self_attn_layers or []
-
-        # apply magic --> this defines a head_size of 64
-        hidden_size = 64 * depth
-        num_heads = depth
-
-        self.num_heads = num_heads
-
-        self.x_embedder = PatchEmbed(
-            input_size,
-            patch_size,
-            in_channels,
-            hidden_size,
-            bias=True,
-            strict_img_size=self.pos_embed_max_size is None,
-            dtype=dtype,
-            device=device,
-        )
-        self.t_embedder = TimestepEmbedder(hidden_size, dtype=dtype, device=device)
-
-        if adm_in_channels is not None:
-            assert isinstance(adm_in_channels, int)
-            self.y_embedder = VectorEmbedder(adm_in_channels, hidden_size, dtype=dtype, device=device)
-
-        self.context_embedder = torch.nn.Identity()
-        if context_embedder_config is not None:
-            if context_embedder_config["target"] == "torch.nn.Linear":
-                self.context_embedder = torch.nn.Linear(**context_embedder_config["params"], dtype=dtype, device=device)
-
-        self.register_length = register_length
-        if self.register_length > 0:
-            self.register = torch.nn.Parameter(torch.randn(1, register_length, hidden_size, dtype=dtype, device=device))
-
-        # num_patches = self.x_embedder.num_patches
-        # Will use fixed sin-cos embedding:
-        # just use a buffer already
-        if num_patches is not None:
-            self.register_buffer(
-                "pos_embed",
-                torch.zeros(1, num_patches, hidden_size, dtype=dtype, device=device),
-            )
-        else:
-            self.pos_embed = None
-
-        self.joint_blocks = torch.nn.ModuleList(
-            [
-                JointBlock(
-                    hidden_size,
-                    num_heads,
-                    mlp_ratio=mlp_ratio,
-                    qkv_bias=qkv_bias,
-                    attn_mode=attn_mode,
-                    pre_only=i == depth - 1,
-                    rmsnorm=rmsnorm,
-                    scale_mod_only=scale_mod_only,
-                    swiglu=swiglu,
-                    qk_norm=qk_norm,
-                    x_block_self_attn=(i in self.x_block_self_attn_layers),
-                    dtype=dtype,
-                    device=device,
-                )
-                for i in range(depth)
-            ]
-        )
-
-        self.final_layer = FinalLayer(hidden_size, patch_size, self.out_channels, dtype=dtype, device=device)
-
-    def cropped_pos_embed(self, hw: torch.Size) -> torch.Tensor:
-        assert self.pos_embed_max_size is not None
-        p = self.x_embedder.patch_size[0]
-        h, w = hw
-        # patched size
-        h = h // p
-        w = w // p
-        assert h <= self.pos_embed_max_size, (h, self.pos_embed_max_size)
-        assert w <= self.pos_embed_max_size, (w, self.pos_embed_max_size)
-        top = (self.pos_embed_max_size - h) // 2
-        left = (self.pos_embed_max_size - w) // 2
-        spatial_pos_embed: torch.Tensor = rearrange(
-            self.pos_embed,
-            "1 (h w) c -> 1 h w c",
-            h=self.pos_embed_max_size,
-            w=self.pos_embed_max_size,
-        )  # type: ignore Type checking does not correctly infer the type of the self.pos_embed buffer.
-        spatial_pos_embed = spatial_pos_embed[:, top : top + h, left : left + w, :]
-        spatial_pos_embed = rearrange(spatial_pos_embed, "1 h w c -> 1 (h w) c")
-        return spatial_pos_embed
-
-    def unpatchify(self, x: torch.Tensor, hw: Optional[torch.Size] = None) -> torch.Tensor:
-        """
-        x: (N, T, patch_size**2 * C)
-        imgs: (N, H, W, C)
-        """
-        c = self.out_channels
-        p = self.x_embedder.patch_size[0]
-        if hw is None:
-            h = w = int(x.shape[1] ** 0.5)
-        else:
-            h, w = hw
-            h = h // p
-            w = w // p
-        assert h * w == x.shape[1]
-
-        x = x.reshape(shape=(x.shape[0], h, w, p, p, c))
-        x = torch.einsum("nhwpqc->nchpwq", x)
-        imgs = x.reshape(shape=(x.shape[0], c, h * p, w * p))
-        return imgs
-
-    def forward_core_with_concat(
-        self,
-        x: torch.Tensor,
-        c_mod: torch.Tensor,
-        context: Optional[torch.Tensor] = None,
-    ) -> torch.Tensor:
-        if self.register_length > 0:
-            context = torch.cat(
-                (
-                    repeat(self.register, "1 ... -> b ...", b=x.shape[0]),
-                    context if context is not None else torch.Tensor([]).type_as(x),
-                ),
-                1,
-            )
-
-        # context is B, L', D
-        # x is B, L, D
-        for block in self.joint_blocks:
-            context, x = block(context, x, c=c_mod)
-
-        x = self.final_layer(x, c_mod)  # (N, T, patch_size ** 2 * out_channels)
-        return x
-
-    def forward(
-        self,
-        x: torch.Tensor,
-        t: torch.Tensor,
-        y: Optional[torch.Tensor] = None,
-        context: Optional[torch.Tensor] = None,
-    ) -> torch.Tensor:
-        """
-        Forward pass of DiT.
-        x: (N, C, H, W) tensor of spatial inputs (images or latent representations of images)
-        t: (N,) tensor of diffusion timesteps
-        y: (N,) tensor of class labels
-        """
-        hw = x.shape[-2:]
-        x = self.x_embedder(x) + self.cropped_pos_embed(hw)
-        c = self.t_embedder(t, dtype=x.dtype)  # (N, D)
-        if y is not None:
-            y = self.y_embedder(y)  # (N, D)
-            c = c + y  # (N, D)
-
-        context = self.context_embedder(context)
-
-        x = self.forward_core_with_concat(x, c, context)
-
-        x = self.unpatchify(x, hw=hw)  # (N, out_channels, H, W)
-        return x

invokeai/backend/sd3/other_impls.py

@@ -1,795 +0,0 @@
-# This file was originally copied from:
-# https://github.com/Stability-AI/sd3.5/blob/19bf11c4e1e37324c5aa5a61f010d4127848a09c/other_impls.py
-
-### This file contains impls for underlying related models (CLIP, T5, etc)
-
-import math
-from typing import Callable, Optional
-
-import torch
-from transformers import CLIPTokenizer, T5TokenizerFast
-
-#################################################################################################
-### Core/Utility
-#################################################################################################
-
-
-def attention(
-    q: torch.Tensor, k: torch.Tensor, v: torch.Tensor, heads: int, mask: Optional[torch.Tensor] = None
-) -> torch.Tensor:
-    """Convenience wrapper around a basic attention operation"""
-    b, _, dim_head = q.shape
-    dim_head //= heads
-    q, k, v = map(lambda t: t.view(b, -1, heads, dim_head).transpose(1, 2), (q, k, v))
-    out = torch.nn.functional.scaled_dot_product_attention(q, k, v, attn_mask=mask, dropout_p=0.0, is_causal=False)
-    return out.transpose(1, 2).reshape(b, -1, heads * dim_head)
-
-
-class Mlp(torch.nn.Module):
-    """MLP as used in Vision Transformer, MLP-Mixer and related networks"""
-
-    def __init__(
-        self,
-        in_features: int,
-        hidden_features: Optional[int] = None,
-        out_features: Optional[int] = None,
-        act_layer: Callable[[torch.Tensor], torch.Tensor] | None = None,
-        bias: bool = True,
-        dtype: Optional[torch.dtype] = None,
-        device: Optional[torch.device] = None,
-    ):
-        super().__init__()
-        out_features = out_features or in_features
-        hidden_features = hidden_features or in_features
-        if act_layer is None:
-            act_layer = torch.nn.functional.gelu
-
-        self.fc1 = torch.nn.Linear(in_features, hidden_features, bias=bias, dtype=dtype, device=device)
-        self.act = act_layer
-        self.fc2 = torch.nn.Linear(hidden_features, out_features, bias=bias, dtype=dtype, device=device)
-
-    def forward(self, x: torch.Tensor) -> torch.Tensor:
-        x = self.fc1(x)
-        x = self.act(x)
-        x = self.fc2(x)
-        return x
-
-
-#################################################################################################
-### CLIP
-#################################################################################################
-
-
-class CLIPAttention(torch.nn.Module):
-    def __init__(self, embed_dim, heads, dtype, device):
-        super().__init__()
-        self.heads = heads
-        self.q_proj = torch.nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
-        self.k_proj = torch.nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
-        self.v_proj = torch.nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
-        self.out_proj = torch.nn.Linear(embed_dim, embed_dim, bias=True, dtype=dtype, device=device)
-
-    def forward(self, x, mask=None):
-        q = self.q_proj(x)
-        k = self.k_proj(x)
-        v = self.v_proj(x)
-        out = attention(q, k, v, self.heads, mask)
-        return self.out_proj(out)
-
-
-ACTIVATIONS = {
-    "quick_gelu": lambda a: a * torch.sigmoid(1.702 * a),
-    "gelu": torch.nn.functional.gelu,
-}
-
-
-class CLIPLayer(torch.nn.Module):
-    def __init__(
-        self,
-        embed_dim,
-        heads,
-        intermediate_size,
-        intermediate_activation,
-        dtype,
-        device,
-    ):
-        super().__init__()
-        self.layer_norm1 = torch.nn.LayerNorm(embed_dim, dtype=dtype, device=device)
-        self.self_attn = CLIPAttention(embed_dim, heads, dtype, device)
-        self.layer_norm2 = torch.nn.LayerNorm(embed_dim, dtype=dtype, device=device)
-        # self.mlp = CLIPMLP(embed_dim, intermediate_size, intermediate_activation, dtype, device)
-        self.mlp = Mlp(
-            embed_dim,
-            intermediate_size,
-            embed_dim,
-            act_layer=ACTIVATIONS[intermediate_activation],
-            dtype=dtype,
-            device=device,
-        )
-
-    def forward(self, x, mask=None):
-        x += self.self_attn(self.layer_norm1(x), mask)
-        x += self.mlp(self.layer_norm2(x))
-        return x
-
-
-class CLIPEncoder(torch.nn.Module):
-    def __init__(
-        self,
-        num_layers,
-        embed_dim,
-        heads,
-        intermediate_size,
-        intermediate_activation,
-        dtype,
-        device,
-    ):
-        super().__init__()
-        self.layers = torch.nn.ModuleList(
-            [
-                CLIPLayer(
-                    embed_dim,
-                    heads,
-                    intermediate_size,
-                    intermediate_activation,
-                    dtype,
-                    device,
-                )
-                for i in range(num_layers)
-            ]
-        )
-
-    def forward(self, x, mask=None, intermediate_output=None):
-        if intermediate_output is not None:
-            if intermediate_output < 0:
-                intermediate_output = len(self.layers) + intermediate_output
-        intermediate = None
-        for i, l in enumerate(self.layers):
-            x = l(x, mask)
-            if i == intermediate_output:
-                intermediate = x.clone()
-        return x, intermediate
-
-
-class CLIPEmbeddings(torch.nn.Module):
-    def __init__(self, embed_dim, vocab_size=49408, num_positions=77, dtype=None, device=None):
-        super().__init__()
-        self.token_embedding = torch.nn.Embedding(vocab_size, embed_dim, dtype=dtype, device=device)
-        self.position_embedding = torch.nn.Embedding(num_positions, embed_dim, dtype=dtype, device=device)
-
-    def forward(self, input_tokens):
-        return self.token_embedding(input_tokens) + self.position_embedding.weight
-
-
-class CLIPTextModel_(torch.nn.Module):
-    def __init__(self, config_dict, dtype, device):
-        num_layers = config_dict["num_hidden_layers"]
-        embed_dim = config_dict["hidden_size"]
-        heads = config_dict["num_attention_heads"]
-        intermediate_size = config_dict["intermediate_size"]
-        intermediate_activation = config_dict["hidden_act"]
-        super().__init__()
-        self.embeddings = CLIPEmbeddings(embed_dim, dtype=torch.float32, device=device)
-        self.encoder = CLIPEncoder(
-            num_layers,
-            embed_dim,
-            heads,
-            intermediate_size,
-            intermediate_activation,
-            dtype,
-            device,
-        )
-        self.final_layer_norm = torch.nn.LayerNorm(embed_dim, dtype=dtype, device=device)
-
-    def forward(self, input_tokens, intermediate_output=None, final_layer_norm_intermediate=True):
-        x = self.embeddings(input_tokens)
-        causal_mask = torch.empty(x.shape[1], x.shape[1], dtype=x.dtype, device=x.device).fill_(float("-inf")).triu_(1)
-        x, i = self.encoder(x, mask=causal_mask, intermediate_output=intermediate_output)
-        x = self.final_layer_norm(x)
-        if i is not None and final_layer_norm_intermediate:
-            i = self.final_layer_norm(i)
-        pooled_output = x[
-            torch.arange(x.shape[0], device=x.device),
-            input_tokens.to(dtype=torch.int, device=x.device).argmax(dim=-1),
-        ]
-        return x, i, pooled_output
-
-
-class CLIPTextModel(torch.nn.Module):
-    def __init__(self, config_dict, dtype, device):
-        super().__init__()
-        self.num_layers = config_dict["num_hidden_layers"]
-        self.text_model = CLIPTextModel_(config_dict, dtype, device)
-        embed_dim = config_dict["hidden_size"]
-        self.text_projection = torch.nn.Linear(embed_dim, embed_dim, bias=False, dtype=dtype, device=device)
-        self.text_projection.weight.copy_(torch.eye(embed_dim))
-        self.dtype = dtype
-
-    def get_input_embeddings(self):
-        return self.text_model.embeddings.token_embedding
-
-    def set_input_embeddings(self, embeddings):
-        self.text_model.embeddings.token_embedding = embeddings
-
-    def forward(self, *args, **kwargs):
-        x = self.text_model(*args, **kwargs)
-        out = self.text_projection(x[2])
-        return (x[0], x[1], out, x[2])
-
-
-def parse_parentheses(string):
-    result = []
-    current_item = ""
-    nesting_level = 0
-    for char in string:
-        if char == "(":
-            if nesting_level == 0:
-                if current_item:
-                    result.append(current_item)
-                    current_item = "("
-                else:
-                    current_item = "("
-            else:
-                current_item += char
-            nesting_level += 1
-        elif char == ")":
-            nesting_level -= 1
-            if nesting_level == 0:
-                result.append(current_item + ")")
-                current_item = ""
-            else:
-                current_item += char
-        else:
-            current_item += char
-    if current_item:
-        result.append(current_item)
-    return result
-
-
-def token_weights(string, current_weight):
-    a = parse_parentheses(string)
-    out = []
-    for x in a:
-        weight = current_weight
-        if len(x) >= 2 and x[-1] == ")" and x[0] == "(":
-            x = x[1:-1]
-            xx = x.rfind(":")
-            weight *= 1.1
-            if xx > 0:
-                try:
-                    weight = float(x[xx + 1 :])
-                    x = x[:xx]
-                except:
-                    pass
-            out += token_weights(x, weight)
-        else:
-            out += [(x, current_weight)]
-    return out
-
-
-def escape_important(text):
-    text = text.replace("\\)", "\0\1")
-    text = text.replace("\\(", "\0\2")
-    return text
-
-
-def unescape_important(text):
-    text = text.replace("\0\1", ")")
-    text = text.replace("\0\2", "(")
-    return text
-
-
-class SDTokenizer:
-    def __init__(
-        self,
-        max_length=77,
-        pad_with_end=True,
-        tokenizer=None,
-        has_start_token=True,
-        pad_to_max_length=True,
-        min_length=None,
-        extra_padding_token=None,
-    ):
-        self.tokenizer = tokenizer
-        self.max_length = max_length
-        self.min_length = min_length
-
-        empty = self.tokenizer("")["input_ids"]
-        if has_start_token:
-            self.tokens_start = 1
-            self.start_token = empty[0]
-            self.end_token = empty[1]
-        else:
-            self.tokens_start = 0
-            self.start_token = None
-            self.end_token = empty[0]
-        self.pad_with_end = pad_with_end
-        self.pad_to_max_length = pad_to_max_length
-        self.extra_padding_token = extra_padding_token
-
-        vocab = self.tokenizer.get_vocab()
-        self.inv_vocab = {v: k for k, v in vocab.items()}
-        self.max_word_length = 8
-
-    def tokenize_with_weights(self, text: str, return_word_ids=False):
-        """
-        Tokenize the text, with weight values - presume 1.0 for all and ignore other features here.
-        The details aren't relevant for a reference impl, and weights themselves has weak effect on SD3.
-        """
-        if self.pad_with_end:
-            pad_token = self.end_token
-        else:
-            pad_token = 0
-
-        text = escape_important(text)
-        parsed_weights = token_weights(text, 1.0)
-
-        # tokenize words
-        tokens = []
-        for weighted_segment, weight in parsed_weights:
-            to_tokenize = unescape_important(weighted_segment).replace("\n", " ").split(" ")
-            to_tokenize = [x for x in to_tokenize if x != ""]
-            for word in to_tokenize:
-                # parse word
-                tokens.append([(t, weight) for t in self.tokenizer(word)["input_ids"][self.tokens_start : -1]])
-
-        # reshape token array to CLIP input size
-        batched_tokens = []
-        batch = []
-        if self.start_token is not None:
-            batch.append((self.start_token, 1.0, 0))
-        batched_tokens.append(batch)
-        for i, t_group in enumerate(tokens):
-            # determine if we're going to try and keep the tokens in a single batch
-            is_large = len(t_group) >= self.max_word_length
-
-            while len(t_group) > 0:
-                if len(t_group) + len(batch) > self.max_length - 1:
-                    remaining_length = self.max_length - len(batch) - 1
-                    # break word in two and add end token
-                    if is_large:
-                        batch.extend([(t, w, i + 1) for t, w in t_group[:remaining_length]])
-                        batch.append((self.end_token, 1.0, 0))
-                        t_group = t_group[remaining_length:]
-                    # add end token and pad
-                    else:
-                        batch.append((self.end_token, 1.0, 0))
-                        if self.pad_to_max_length:
-                            batch.extend([(pad_token, 1.0, 0)] * (remaining_length))
-                    # start new batch
-                    batch = []
-                    if self.start_token is not None:
-                        batch.append((self.start_token, 1.0, 0))
-                    batched_tokens.append(batch)
-                else:
-                    batch.extend([(t, w, i + 1) for t, w in t_group])
-                    t_group = []
-
-        # pad extra padding token first befor getting to the end token
-        if self.extra_padding_token is not None:
-            batch.extend([(self.extra_padding_token, 1.0, 0)] * (self.min_length - len(batch) - 1))
-        # fill last batch
-        batch.append((self.end_token, 1.0, 0))
-        if self.pad_to_max_length:
-            batch.extend([(pad_token, 1.0, 0)] * (self.max_length - len(batch)))
-        if self.min_length is not None and len(batch) < self.min_length:
-            batch.extend([(pad_token, 1.0, 0)] * (self.min_length - len(batch)))
-
-        if not return_word_ids:
-            batched_tokens = [[(t, w) for t, w, _ in x] for x in batched_tokens]
-
-        return batched_tokens
-
-    def untokenize(self, token_weight_pair):
-        return list(map(lambda a: (a, self.inv_vocab[a[0]]), token_weight_pair))
-
-
-class SDXLClipGTokenizer(SDTokenizer):
-    def __init__(self, tokenizer):
-        super().__init__(pad_with_end=False, tokenizer=tokenizer)
-
-
-class SD3Tokenizer:
-    def __init__(self):
-        clip_tokenizer = CLIPTokenizer.from_pretrained("openai/clip-vit-large-patch14")
-        self.clip_l = SDTokenizer(tokenizer=clip_tokenizer)
-        self.clip_g = SDXLClipGTokenizer(clip_tokenizer)
-        self.t5xxl = T5XXLTokenizer()
-
-    def tokenize_with_weights(self, text: str):
-        out = {}
-        out["l"] = self.clip_l.tokenize_with_weights(text)
-        out["g"] = self.clip_g.tokenize_with_weights(text)
-        out["t5xxl"] = self.t5xxl.tokenize_with_weights(text[:226])
-        return out
-
-
-class ClipTokenWeightEncoder:
-    def encode_token_weights(self, token_weight_pairs):
-        tokens = list(map(lambda a: a[0], token_weight_pairs[0]))
-        out, pooled = self([tokens])
-        if pooled is not None:
-            first_pooled = pooled[0:1].cpu()
-        else:
-            first_pooled = pooled
-        output = [out[0:1]]
-        return torch.cat(output, dim=-2).cpu(), first_pooled
-
-
-class SDClipModel(torch.nn.Module, ClipTokenWeightEncoder):
-    """Uses the CLIP transformer encoder for text (from huggingface)"""
-
-    LAYERS = ["last", "pooled", "hidden"]
-
-    def __init__(
-        self,
-        device="cpu",
-        max_length=77,
-        layer="last",
-        layer_idx=None,
-        textmodel_json_config=None,
-        dtype=None,
-        model_class=CLIPTextModel,
-        special_tokens={"start": 49406, "end": 49407, "pad": 49407},
-        layer_norm_hidden_state=True,
-        return_projected_pooled=True,
-    ):
-        super().__init__()
-        assert layer in self.LAYERS
-        self.transformer = model_class(textmodel_json_config, dtype, device)
-        self.num_layers = self.transformer.num_layers
-        self.max_length = max_length
-        self.transformer = self.transformer.eval()
-        for param in self.parameters():
-            param.requires_grad = False
-        self.layer = layer
-        self.layer_idx = None
-        self.special_tokens = special_tokens
-        self.logit_scale = torch.nn.Parameter(torch.tensor(4.6055))
-        self.layer_norm_hidden_state = layer_norm_hidden_state
-        self.return_projected_pooled = return_projected_pooled
-        if layer == "hidden":
-            assert layer_idx is not None
-            assert abs(layer_idx) < self.num_layers
-            self.set_clip_options({"layer": layer_idx})
-        self.options_default = (
-            self.layer,
-            self.layer_idx,
-            self.return_projected_pooled,
-        )
-
-    def set_clip_options(self, options):
-        layer_idx = options.get("layer", self.layer_idx)
-        self.return_projected_pooled = options.get("projected_pooled", self.return_projected_pooled)
-        if layer_idx is None or abs(layer_idx) > self.num_layers:
-            self.layer = "last"
-        else:
-            self.layer = "hidden"
-            self.layer_idx = layer_idx
-
-    def forward(self, tokens):
-        backup_embeds = self.transformer.get_input_embeddings()
-        device = backup_embeds.weight.device
-        tokens = torch.LongTensor(tokens).to(device)
-        outputs = self.transformer(
-            tokens,
-            intermediate_output=self.layer_idx,
-            final_layer_norm_intermediate=self.layer_norm_hidden_state,
-        )
-        self.transformer.set_input_embeddings(backup_embeds)
-        if self.layer == "last":
-            z = outputs[0]
-        else:
-            z = outputs[1]
-        pooled_output = None
-        if len(outputs) >= 3:
-            if not self.return_projected_pooled and len(outputs) >= 4 and outputs[3] is not None:
-                pooled_output = outputs[3].float()
-            elif outputs[2] is not None:
-                pooled_output = outputs[2].float()
-        return z.float(), pooled_output
-
-
-class SDXLClipG(SDClipModel):
-    """Wraps the CLIP-G model into the SD-CLIP-Model interface"""
-
-    def __init__(self, config, device="cpu", layer="penultimate", layer_idx=None, dtype=None):
-        if layer == "penultimate":
-            layer = "hidden"
-            layer_idx = -2
-        super().__init__(
-            device=device,
-            layer=layer,
-            layer_idx=layer_idx,
-            textmodel_json_config=config,
-            dtype=dtype,
-            special_tokens={"start": 49406, "end": 49407, "pad": 0},
-            layer_norm_hidden_state=False,
-        )
-
-
-class T5XXLModel(SDClipModel):
-    """Wraps the T5-XXL model into the SD-CLIP-Model interface for convenience"""
-
-    def __init__(self, config, device="cpu", layer="last", layer_idx=None, dtype=None):
-        super().__init__(
-            device=device,
-            layer=layer,
-            layer_idx=layer_idx,
-            textmodel_json_config=config,
-            dtype=dtype,
-            special_tokens={"end": 1, "pad": 0},
-            model_class=T5,
-        )
-
-
-#################################################################################################
-### T5 implementation, for the T5-XXL text encoder portion, largely pulled from upstream impl
-#################################################################################################
-
-
-class T5XXLTokenizer(SDTokenizer):
-    """Wraps the T5 Tokenizer from HF into the SDTokenizer interface"""
-
-    def __init__(self):
-        super().__init__(
-            pad_with_end=False,
-            tokenizer=T5TokenizerFast.from_pretrained("google/t5-v1_1-xxl"),
-            has_start_token=False,
-            pad_to_max_length=False,
-            max_length=99999999,
-            min_length=77,
-        )
-
-
-class T5LayerNorm(torch.nn.Module):
-    def __init__(self, hidden_size, eps=1e-6, dtype=None, device=None):
-        super().__init__()
-        self.weight = torch.nn.Parameter(torch.ones(hidden_size, dtype=dtype, device=device))
-        self.variance_epsilon = eps
-
-    def forward(self, x):
-        variance = x.pow(2).mean(-1, keepdim=True)
-        x = x * torch.rsqrt(variance + self.variance_epsilon)
-        return self.weight.to(device=x.device, dtype=x.dtype) * x
-
-
-class T5DenseGatedActDense(torch.nn.Module):
-    def __init__(self, model_dim, ff_dim, dtype, device):
-        super().__init__()
-        self.wi_0 = torch.nn.Linear(model_dim, ff_dim, bias=False, dtype=dtype, device=device)
-        self.wi_1 = torch.nn.Linear(model_dim, ff_dim, bias=False, dtype=dtype, device=device)
-        self.wo = torch.nn.Linear(ff_dim, model_dim, bias=False, dtype=dtype, device=device)
-
-    def forward(self, x):
-        hidden_gelu = torch.nn.functional.gelu(self.wi_0(x), approximate="tanh")
-        hidden_linear = self.wi_1(x)
-        x = hidden_gelu * hidden_linear
-        x = self.wo(x)
-        return x
-
-
-class T5LayerFF(torch.nn.Module):
-    def __init__(self, model_dim, ff_dim, dtype, device):
-        super().__init__()
-        self.DenseReluDense = T5DenseGatedActDense(model_dim, ff_dim, dtype, device)
-        self.layer_norm = T5LayerNorm(model_dim, dtype=dtype, device=device)
-
-    def forward(self, x):
-        forwarded_states = self.layer_norm(x)
-        forwarded_states = self.DenseReluDense(forwarded_states)
-        x += forwarded_states
-        return x
-
-
-class T5Attention(torch.nn.Module):
-    def __init__(self, model_dim, inner_dim, num_heads, relative_attention_bias, dtype, device):
-        super().__init__()
-        # Mesh TensorFlow initialization to avoid scaling before softmax
-        self.q = torch.nn.Linear(model_dim, inner_dim, bias=False, dtype=dtype, device=device)
-        self.k = torch.nn.Linear(model_dim, inner_dim, bias=False, dtype=dtype, device=device)
-        self.v = torch.nn.Linear(model_dim, inner_dim, bias=False, dtype=dtype, device=device)
-        self.o = torch.nn.Linear(inner_dim, model_dim, bias=False, dtype=dtype, device=device)
-        self.num_heads = num_heads
-        self.relative_attention_bias = None
-        if relative_attention_bias:
-            self.relative_attention_num_buckets = 32
-            self.relative_attention_max_distance = 128
-            self.relative_attention_bias = torch.nn.Embedding(
-                self.relative_attention_num_buckets, self.num_heads, device=device
-            )
-
-    @staticmethod
-    def _relative_position_bucket(relative_position, bidirectional=True, num_buckets=32, max_distance=128):
-        """
-        Adapted from Mesh Tensorflow:
-        https://github.com/tensorflow/mesh/blob/0cb87fe07da627bf0b7e60475d59f95ed6b5be3d/mesh_tensorflow/transformer/transformer_layers.py#L593
-
-        Translate relative position to a bucket number for relative attention. The relative position is defined as
-        memory_position - query_position, i.e. the distance in tokens from the attending position to the attended-to
-        position. If bidirectional=False, then positive relative positions are invalid. We use smaller buckets for
-        small absolute relative_position and larger buckets for larger absolute relative_positions. All relative
-        positions >=max_distance map to the same bucket. All relative positions <=-max_distance map to the same bucket.
-        This should allow for more graceful generalization to longer sequences than the model has been trained on
-
-        Args:
-            relative_position: an int32 Tensor
-            bidirectional: a boolean - whether the attention is bidirectional
-            num_buckets: an integer
-            max_distance: an integer
-
-        Returns:
-            a Tensor with the same shape as relative_position, containing int32 values in the range [0, num_buckets)
-        """
-        relative_buckets = 0
-        if bidirectional:
-            num_buckets //= 2
-            relative_buckets += (relative_position > 0).to(torch.long) * num_buckets
-            relative_position = torch.abs(relative_position)
-        else:
-            relative_position = -torch.min(relative_position, torch.zeros_like(relative_position))
-        # now relative_position is in the range [0, inf)
-        # half of the buckets are for exact increments in positions
-        max_exact = num_buckets // 2
-        is_small = relative_position < max_exact
-        # The other half of the buckets are for logarithmically bigger bins in positions up to max_distance
-        relative_position_if_large = max_exact + (
-            torch.log(relative_position.float() / max_exact)
-            / math.log(max_distance / max_exact)
-            * (num_buckets - max_exact)
-        ).to(torch.long)
-        relative_position_if_large = torch.min(
-            relative_position_if_large,
-            torch.full_like(relative_position_if_large, num_buckets - 1),
-        )
-        relative_buckets += torch.where(is_small, relative_position, relative_position_if_large)
-        return relative_buckets
-
-    def compute_bias(self, query_length, key_length, device):
-        """Compute binned relative position bias"""
-        context_position = torch.arange(query_length, dtype=torch.long, device=device)[:, None]
-        memory_position = torch.arange(key_length, dtype=torch.long, device=device)[None, :]
-        relative_position = memory_position - context_position  # shape (query_length, key_length)
-        relative_position_bucket = self._relative_position_bucket(
-            relative_position,  # shape (query_length, key_length)
-            bidirectional=True,
-            num_buckets=self.relative_attention_num_buckets,
-            max_distance=self.relative_attention_max_distance,
-        )
-        values = self.relative_attention_bias(relative_position_bucket)  # shape (query_length, key_length, num_heads)
-        values = values.permute([2, 0, 1]).unsqueeze(0)  # shape (1, num_heads, query_length, key_length)
-        return values
-
-    def forward(self, x, past_bias=None):
-        q = self.q(x)
-        k = self.k(x)
-        v = self.v(x)
-        if self.relative_attention_bias is not None:
-            past_bias = self.compute_bias(x.shape[1], x.shape[1], x.device)
-        if past_bias is not None:
-            mask = past_bias
-        out = attention(q, k * ((k.shape[-1] / self.num_heads) ** 0.5), v, self.num_heads, mask)
-        return self.o(out), past_bias
-
-
-class T5LayerSelfAttention(torch.nn.Module):
-    def __init__(
-        self,
-        model_dim,
-        inner_dim,
-        ff_dim,
-        num_heads,
-        relative_attention_bias,
-        dtype,
-        device,
-    ):
-        super().__init__()
-        self.SelfAttention = T5Attention(model_dim, inner_dim, num_heads, relative_attention_bias, dtype, device)
-        self.layer_norm = T5LayerNorm(model_dim, dtype=dtype, device=device)
-
-    def forward(self, x, past_bias=None):
-        output, past_bias = self.SelfAttention(self.layer_norm(x), past_bias=past_bias)
-        x += output
-        return x, past_bias
-
-
-class T5Block(torch.nn.Module):
-    def __init__(
-        self,
-        model_dim,
-        inner_dim,
-        ff_dim,
-        num_heads,
-        relative_attention_bias,
-        dtype,
-        device,
-    ):
-        super().__init__()
-        self.layer = torch.nn.ModuleList()
-        self.layer.append(
-            T5LayerSelfAttention(
-                model_dim,
-                inner_dim,
-                ff_dim,
-                num_heads,
-                relative_attention_bias,
-                dtype,
-                device,
-            )
-        )
-        self.layer.append(T5LayerFF(model_dim, ff_dim, dtype, device))
-
-    def forward(self, x, past_bias=None):
-        x, past_bias = self.layer[0](x, past_bias)
-        x = self.layer[-1](x)
-        return x, past_bias
-
-
-class T5Stack(torch.nn.Module):
-    def __init__(
-        self,
-        num_layers,
-        model_dim,
-        inner_dim,
-        ff_dim,
-        num_heads,
-        vocab_size,
-        dtype,
-        device,
-    ):
-        super().__init__()
-        self.embed_tokens = torch.nn.Embedding(vocab_size, model_dim, device=device)
-        self.block = torch.nn.ModuleList(
-            [
-                T5Block(
-                    model_dim,
-                    inner_dim,
-                    ff_dim,
-                    num_heads,
-                    relative_attention_bias=(i == 0),
-                    dtype=dtype,
-                    device=device,
-                )
-                for i in range(num_layers)
-            ]
-        )
-        self.final_layer_norm = T5LayerNorm(model_dim, dtype=dtype, device=device)
-
-    def forward(self, input_ids, intermediate_output=None, final_layer_norm_intermediate=True):
-        intermediate = None
-        x = self.embed_tokens(input_ids)
-        past_bias = None
-        for i, l in enumerate(self.block):
-            x, past_bias = l(x, past_bias)
-            if i == intermediate_output:
-                intermediate = x.clone()
-        x = self.final_layer_norm(x)
-        if intermediate is not None and final_layer_norm_intermediate:
-            intermediate = self.final_layer_norm(intermediate)
-        return x, intermediate
-
-
-class T5(torch.nn.Module):
-    def __init__(self, config_dict, dtype, device):
-        super().__init__()
-        self.num_layers = config_dict["num_layers"]
-        self.encoder = T5Stack(
-            self.num_layers,
-            config_dict["d_model"],
-            config_dict["d_model"],
-            config_dict["d_ff"],
-            config_dict["num_heads"],
-            config_dict["vocab_size"],
-            dtype,
-            device,
-        )
-        self.dtype = dtype
-
-    def get_input_embeddings(self):
-        return self.encoder.embed_tokens
-
-    def set_input_embeddings(self, embeddings):
-        self.encoder.embed_tokens = embeddings
-
-    def forward(self, *args, **kwargs):
-        return self.encoder(*args, **kwargs)

invokeai/backend/sd3/sd3_impls.py

@@ -1,609 +0,0 @@
-# This file was originally copied from:
-# https://github.com/Stability-AI/sd3.5/blob/19bf11c4e1e37324c5aa5a61f010d4127848a09c/sd3_impls.py
-
-
-### Impls of the SD3 core diffusion model and VAE
-
-import math
-import re
-
-import einops
-import torch
-from PIL import Image
-from tqdm import tqdm
-
-from invokeai.backend.sd3.mmditx import MMDiTX
-
-#################################################################################################
-### MMDiT Model Wrapping
-#################################################################################################
-
-
-class ModelSamplingDiscreteFlow(torch.nn.Module):
-    """Helper for sampler scheduling (ie timestep/sigma calculations) for Discrete Flow models"""
-
-    def __init__(self, shift: float = 1.0):
-        super().__init__()
-        self.shift = shift
-        timesteps = 1000
-        ts = self.sigma(torch.arange(1, timesteps + 1, 1))
-        self.register_buffer("sigmas", ts)
-
-    @property
-    def sigma_min(self):
-        return self.sigmas[0]
-
-    @property
-    def sigma_max(self):
-        return self.sigmas[-1]
-
-    def timestep(self, sigma: torch.Tensor) -> torch.Tensor:
-        return sigma * 1000
-
-    def sigma(self, timestep: torch.Tensor):
-        timestep = timestep / 1000.0
-        if self.shift == 1.0:
-            return timestep
-        return self.shift * timestep / (1 + (self.shift - 1) * timestep)
-
-    def calculate_denoised(
-        self, sigma: torch.Tensor, model_output: torch.Tensor, model_input: torch.Tensor
-    ) -> torch.Tensor:
-        sigma = sigma.view(sigma.shape[:1] + (1,) * (model_output.ndim - 1))
-        return model_input - model_output * sigma
-
-    def noise_scaling(self, sigma, noise, latent_image, max_denoise=False):
-        return sigma * noise + (1.0 - sigma) * latent_image
-
-
-class BaseModel(torch.nn.Module):
-    """Wrapper around the core MM-DiT model"""
-
-    def __init__(
-        self,
-        shift=1.0,
-        device=None,
-        dtype=torch.float32,
-        file=None,
-        prefix="",
-        verbose=False,
-    ):
-        super().__init__()
-        # Important configuration values can be quickly determined by checking shapes in the source file
-        # Some of these will vary between models (eg 2B vs 8B primarily differ in their depth, but also other details change)
-        patch_size = file.get_tensor(f"{prefix}x_embedder.proj.weight").shape[2]
-        depth = file.get_tensor(f"{prefix}x_embedder.proj.weight").shape[0] // 64
-        num_patches = file.get_tensor(f"{prefix}pos_embed").shape[1]
-        pos_embed_max_size = round(math.sqrt(num_patches))
-        adm_in_channels = file.get_tensor(f"{prefix}y_embedder.mlp.0.weight").shape[1]
-        context_shape = file.get_tensor(f"{prefix}context_embedder.weight").shape
-        qk_norm = "rms" if f"{prefix}joint_blocks.0.context_block.attn.ln_k.weight" in file.keys() else None
-        x_block_self_attn_layers = sorted(
-            [
-                int(key.split(".x_block.attn2.ln_k.weight")[0].split(".")[-1])
-                for key in list(filter(re.compile(".*.x_block.attn2.ln_k.weight").match, file.keys()))
-            ]
-        )
-
-        context_embedder_config = {
-            "target": "torch.nn.Linear",
-            "params": {
-                "in_features": context_shape[1],
-                "out_features": context_shape[0],
-            },
-        }
-        self.diffusion_model = MMDiTX(
-            input_size=None,
-            pos_embed_scaling_factor=None,
-            pos_embed_offset=None,
-            pos_embed_max_size=pos_embed_max_size,
-            patch_size=patch_size,
-            in_channels=16,
-            depth=depth,
-            num_patches=num_patches,
-            adm_in_channels=adm_in_channels,
-            context_embedder_config=context_embedder_config,
-            qk_norm=qk_norm,
-            x_block_self_attn_layers=x_block_self_attn_layers,
-            device=device,
-            dtype=dtype,
-            verbose=verbose,
-        )
-        self.model_sampling = ModelSamplingDiscreteFlow(shift=shift)
-
-    def apply_model(
-        self, x: torch.Tensor, sigma: float, c_crossattn: torch.Tensor | None = None, y: torch.Tensor | None = None
-    ):
-        dtype = self.get_dtype()
-        timestep = self.model_sampling.timestep(sigma).float()
-        model_output = self.diffusion_model(x.to(dtype), timestep, context=c_crossattn.to(dtype), y=y.to(dtype)).float()
-        return self.model_sampling.calculate_denoised(sigma, model_output, x)
-
-    def forward(self, *args, **kwargs):
-        return self.apply_model(*args, **kwargs)
-
-    def get_dtype(self):
-        return self.diffusion_model.dtype
-
-
-class CFGDenoiser(torch.nn.Module):
-    """Helper for applying CFG Scaling to diffusion outputs"""
-
-    def __init__(self, model):
-        super().__init__()
-        self.model = model
-
-    def forward(self, x, timestep, cond, uncond, cond_scale):
-        # Run cond and uncond in a batch together
-        batched = self.model.apply_model(
-            torch.cat([x, x]),
-            torch.cat([timestep, timestep]),
-            c_crossattn=torch.cat([cond["c_crossattn"], uncond["c_crossattn"]]),
-            y=torch.cat([cond["y"], uncond["y"]]),
-        )
-        # Then split and apply CFG Scaling
-        pos_out, neg_out = batched.chunk(2)
-        scaled = neg_out + (pos_out - neg_out) * cond_scale
-        return scaled
-
-
-class SD3LatentFormat:
-    """Latents are slightly shifted from center - this class must be called after VAE Decode to correct for the shift"""
-
-    def __init__(self):
-        self.scale_factor = 1.5305
-        self.shift_factor = 0.0609
-
-    def process_in(self, latent):
-        return (latent - self.shift_factor) * self.scale_factor
-
-    def process_out(self, latent):
-        return (latent / self.scale_factor) + self.shift_factor
-
-    def decode_latent_to_preview(self, x0):
-        """Quick RGB approximate preview of sd3 latents"""
-        factors = torch.tensor(
-            [
-                [-0.0645, 0.0177, 0.1052],
-                [0.0028, 0.0312, 0.0650],
-                [0.1848, 0.0762, 0.0360],
-                [0.0944, 0.0360, 0.0889],
-                [0.0897, 0.0506, -0.0364],
-                [-0.0020, 0.1203, 0.0284],
-                [0.0855, 0.0118, 0.0283],
-                [-0.0539, 0.0658, 0.1047],
-                [-0.0057, 0.0116, 0.0700],
-                [-0.0412, 0.0281, -0.0039],
-                [0.1106, 0.1171, 0.1220],
-                [-0.0248, 0.0682, -0.0481],
-                [0.0815, 0.0846, 0.1207],
-                [-0.0120, -0.0055, -0.0867],
-                [-0.0749, -0.0634, -0.0456],
-                [-0.1418, -0.1457, -0.1259],
-            ],
-            device="cpu",
-        )
-        latent_image = x0[0].permute(1, 2, 0).cpu() @ factors
-
-        latents_ubyte = (
-            ((latent_image + 1) / 2)
-            .clamp(0, 1)  # change scale from -1..1 to 0..1
-            .mul(0xFF)  # to 0..255
-            .byte()
-        ).cpu()
-
-        return Image.fromarray(latents_ubyte.numpy())
-
-
-#################################################################################################
-### Samplers
-#################################################################################################
-
-
-def append_dims(x, target_dims):
-    """Appends dimensions to the end of a tensor until it has target_dims dimensions."""
-    dims_to_append = target_dims - x.ndim
-    return x[(...,) + (None,) * dims_to_append]
-
-
-def to_d(x, sigma, denoised):
-    """Converts a denoiser output to a Karras ODE derivative."""
-    return (x - denoised) / append_dims(sigma, x.ndim)
-
-
-@torch.no_grad()
-@torch.autocast("cuda", dtype=torch.float16)
-def sample_euler(model, x, sigmas, extra_args=None):
-    """Implements Algorithm 2 (Euler steps) from Karras et al. (2022)."""
-    extra_args = {} if extra_args is None else extra_args
-    s_in = x.new_ones([x.shape[0]])
-    for i in tqdm(range(len(sigmas) - 1)):
-        sigma_hat = sigmas[i]
-        denoised = model(x, sigma_hat * s_in, **extra_args)
-        d = to_d(x, sigma_hat, denoised)
-        dt = sigmas[i + 1] - sigma_hat
-        # Euler method
-        x = x + d * dt
-    return x
-
-
-@torch.no_grad()
-@torch.autocast("cuda", dtype=torch.float16)
-def sample_dpmpp_2m(model, x, sigmas, extra_args=None):
-    """DPM-Solver++(2M)."""
-    extra_args = {} if extra_args is None else extra_args
-    s_in = x.new_ones([x.shape[0]])
-    sigma_fn = lambda t: t.neg().exp()
-    t_fn = lambda sigma: sigma.log().neg()
-    old_denoised = None
-    for i in tqdm(range(len(sigmas) - 1)):
-        denoised = model(x, sigmas[i] * s_in, **extra_args)
-        t, t_next = t_fn(sigmas[i]), t_fn(sigmas[i + 1])
-        h = t_next - t
-        if old_denoised is None or sigmas[i + 1] == 0:
-            x = (sigma_fn(t_next) / sigma_fn(t)) * x - (-h).expm1() * denoised
-        else:
-            h_last = t - t_fn(sigmas[i - 1])
-            r = h_last / h
-            denoised_d = (1 + 1 / (2 * r)) * denoised - (1 / (2 * r)) * old_denoised
-            x = (sigma_fn(t_next) / sigma_fn(t)) * x - (-h).expm1() * denoised_d
-        old_denoised = denoised
-    return x
-
-
-#################################################################################################
-### VAE
-#################################################################################################
-
-
-def Normalize(in_channels, num_groups=32, dtype=torch.float32, device=None):
-    return torch.nn.GroupNorm(
-        num_groups=num_groups,
-        num_channels=in_channels,
-        eps=1e-6,
-        affine=True,
-        dtype=dtype,
-        device=device,
-    )
-
-
-class ResnetBlock(torch.nn.Module):
-    def __init__(self, *, in_channels, out_channels=None, dtype=torch.float32, device=None):
-        super().__init__()
-        self.in_channels = in_channels
-        out_channels = in_channels if out_channels is None else out_channels
-        self.out_channels = out_channels
-
-        self.norm1 = Normalize(in_channels, dtype=dtype, device=device)
-        self.conv1 = torch.nn.Conv2d(
-            in_channels,
-            out_channels,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-        self.norm2 = Normalize(out_channels, dtype=dtype, device=device)
-        self.conv2 = torch.nn.Conv2d(
-            out_channels,
-            out_channels,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-        if self.in_channels != self.out_channels:
-            self.nin_shortcut = torch.nn.Conv2d(
-                in_channels,
-                out_channels,
-                kernel_size=1,
-                stride=1,
-                padding=0,
-                dtype=dtype,
-                device=device,
-            )
-        else:
-            self.nin_shortcut = None
-        self.swish = torch.nn.SiLU(inplace=True)
-
-    def forward(self, x):
-        hidden = x
-        hidden = self.norm1(hidden)
-        hidden = self.swish(hidden)
-        hidden = self.conv1(hidden)
-        hidden = self.norm2(hidden)
-        hidden = self.swish(hidden)
-        hidden = self.conv2(hidden)
-        if self.in_channels != self.out_channels:
-            x = self.nin_shortcut(x)
-        return x + hidden
-
-
-class AttnBlock(torch.nn.Module):
-    def __init__(self, in_channels, dtype=torch.float32, device=None):
-        super().__init__()
-        self.norm = Normalize(in_channels, dtype=dtype, device=device)
-        self.q = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0,
-            dtype=dtype,
-            device=device,
-        )
-        self.k = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0,
-            dtype=dtype,
-            device=device,
-        )
-        self.v = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0,
-            dtype=dtype,
-            device=device,
-        )
-        self.proj_out = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=1,
-            stride=1,
-            padding=0,
-            dtype=dtype,
-            device=device,
-        )
-
-    def forward(self, x):
-        hidden = self.norm(x)
-        q = self.q(hidden)
-        k = self.k(hidden)
-        v = self.v(hidden)
-        b, c, h, w = q.shape
-        q, k, v = map(
-            lambda x: einops.rearrange(x, "b c h w -> b 1 (h w) c").contiguous(),
-            (q, k, v),
-        )
-        hidden = torch.nn.functional.scaled_dot_product_attention(q, k, v)  # scale is dim ** -0.5 per default
-        hidden = einops.rearrange(hidden, "b 1 (h w) c -> b c h w", h=h, w=w, c=c, b=b)
-        hidden = self.proj_out(hidden)
-        return x + hidden
-
-
-class Downsample(torch.nn.Module):
-    def __init__(self, in_channels, dtype=torch.float32, device=None):
-        super().__init__()
-        self.conv = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=3,
-            stride=2,
-            padding=0,
-            dtype=dtype,
-            device=device,
-        )
-
-    def forward(self, x):
-        pad = (0, 1, 0, 1)
-        x = torch.nn.functional.pad(x, pad, mode="constant", value=0)
-        x = self.conv(x)
-        return x
-
-
-class Upsample(torch.nn.Module):
-    def __init__(self, in_channels, dtype=torch.float32, device=None):
-        super().__init__()
-        self.conv = torch.nn.Conv2d(
-            in_channels,
-            in_channels,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-
-    def forward(self, x):
-        x = torch.nn.functional.interpolate(x, scale_factor=2.0, mode="nearest")
-        x = self.conv(x)
-        return x
-
-
-class VAEEncoder(torch.nn.Module):
-    def __init__(
-        self,
-        ch=128,
-        ch_mult=(1, 2, 4, 4),
-        num_res_blocks=2,
-        in_channels=3,
-        z_channels=16,
-        dtype=torch.float32,
-        device=None,
-    ):
-        super().__init__()
-        self.num_resolutions = len(ch_mult)
-        self.num_res_blocks = num_res_blocks
-        # downsampling
-        self.conv_in = torch.nn.Conv2d(
-            in_channels,
-            ch,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-        in_ch_mult = (1,) + tuple(ch_mult)
-        self.in_ch_mult = in_ch_mult
-        self.down = torch.nn.ModuleList()
-        for i_level in range(self.num_resolutions):
-            block = torch.nn.ModuleList()
-            attn = torch.nn.ModuleList()
-            block_in = ch * in_ch_mult[i_level]
-            block_out = ch * ch_mult[i_level]
-            for i_block in range(num_res_blocks):
-                block.append(
-                    ResnetBlock(
-                        in_channels=block_in,
-                        out_channels=block_out,
-                        dtype=dtype,
-                        device=device,
-                    )
-                )
-                block_in = block_out
-            down = torch.nn.Module()
-            down.block = block
-            down.attn = attn
-            if i_level != self.num_resolutions - 1:
-                down.downsample = Downsample(block_in, dtype=dtype, device=device)
-            self.down.append(down)
-        # middle
-        self.mid = torch.nn.Module()
-        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
-        self.mid.attn_1 = AttnBlock(block_in, dtype=dtype, device=device)
-        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
-        # end
-        self.norm_out = Normalize(block_in, dtype=dtype, device=device)
-        self.conv_out = torch.nn.Conv2d(
-            block_in,
-            2 * z_channels,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-        self.swish = torch.nn.SiLU(inplace=True)
-
-    def forward(self, x):
-        # downsampling
-        hs = [self.conv_in(x)]
-        for i_level in range(self.num_resolutions):
-            for i_block in range(self.num_res_blocks):
-                h = self.down[i_level].block[i_block](hs[-1])
-                hs.append(h)
-            if i_level != self.num_resolutions - 1:
-                hs.append(self.down[i_level].downsample(hs[-1]))
-        # middle
-        h = hs[-1]
-        h = self.mid.block_1(h)
-        h = self.mid.attn_1(h)
-        h = self.mid.block_2(h)
-        # end
-        h = self.norm_out(h)
-        h = self.swish(h)
-        h = self.conv_out(h)
-        return h
-
-
-class VAEDecoder(torch.nn.Module):
-    def __init__(
-        self,
-        ch=128,
-        out_ch=3,
-        ch_mult=(1, 2, 4, 4),
-        num_res_blocks=2,
-        resolution=256,
-        z_channels=16,
-        dtype=torch.float32,
-        device=None,
-    ):
-        super().__init__()
-        self.num_resolutions = len(ch_mult)
-        self.num_res_blocks = num_res_blocks
-        block_in = ch * ch_mult[self.num_resolutions - 1]
-        curr_res = resolution // 2 ** (self.num_resolutions - 1)
-        # z to block_in
-        self.conv_in = torch.nn.Conv2d(
-            z_channels,
-            block_in,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-        # middle
-        self.mid = torch.nn.Module()
-        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
-        self.mid.attn_1 = AttnBlock(block_in, dtype=dtype, device=device)
-        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in, dtype=dtype, device=device)
-        # upsampling
-        self.up = torch.nn.ModuleList()
-        for i_level in reversed(range(self.num_resolutions)):
-            block = torch.nn.ModuleList()
-            block_out = ch * ch_mult[i_level]
-            for i_block in range(self.num_res_blocks + 1):
-                block.append(
-                    ResnetBlock(
-                        in_channels=block_in,
-                        out_channels=block_out,
-                        dtype=dtype,
-                        device=device,
-                    )
-                )
-                block_in = block_out
-            up = torch.nn.Module()
-            up.block = block
-            if i_level != 0:
-                up.upsample = Upsample(block_in, dtype=dtype, device=device)
-                curr_res = curr_res * 2
-            self.up.insert(0, up)  # prepend to get consistent order
-        # end
-        self.norm_out = Normalize(block_in, dtype=dtype, device=device)
-        self.conv_out = torch.nn.Conv2d(
-            block_in,
-            out_ch,
-            kernel_size=3,
-            stride=1,
-            padding=1,
-            dtype=dtype,
-            device=device,
-        )
-        self.swish = torch.nn.SiLU(inplace=True)
-
-    def forward(self, z):
-        # z to block_in
-        hidden = self.conv_in(z)
-        # middle
-        hidden = self.mid.block_1(hidden)
-        hidden = self.mid.attn_1(hidden)
-        hidden = self.mid.block_2(hidden)
-        # upsampling
-        for i_level in reversed(range(self.num_resolutions)):
-            for i_block in range(self.num_res_blocks + 1):
-                hidden = self.up[i_level].block[i_block](hidden)
-            if i_level != 0:
-                hidden = self.up[i_level].upsample(hidden)
-        # end
-        hidden = self.norm_out(hidden)
-        hidden = self.swish(hidden)
-        hidden = self.conv_out(hidden)
-        return hidden
-
-
-class SDVAE(torch.nn.Module):
-    def __init__(self, dtype=torch.float32, device=None):
-        super().__init__()
-        self.encoder = VAEEncoder(dtype=dtype, device=device)
-        self.decoder = VAEDecoder(dtype=dtype, device=device)
-
-    @torch.autocast("cuda", dtype=torch.float16)
-    def decode(self, latent):
-        return self.decoder(latent)
-
-    @torch.autocast("cuda", dtype=torch.float16)
-    def encode(self, image):
-        hidden = self.encoder(image)
-        mean, logvar = torch.chunk(hidden, 2, dim=1)
-        logvar = torch.clamp(logvar, -30.0, 20.0)
-        std = torch.exp(0.5 * logvar)
-        return mean + std * torch.randn_like(mean)

invokeai/backend/sd3/sd3_infer.py

@@ -1,426 +0,0 @@
-# This file was originally copied from:
-# https://github.com/Stability-AI/sd3.5/blob/19bf11c4e1e37324c5aa5a61f010d4127848a09c/sd3_infer.py
-
-# NOTE: Must have folder `models` with the following files:
-# - `clip_g.safetensors` (openclip bigG, same as SDXL)
-# - `clip_l.safetensors` (OpenAI CLIP-L, same as SDXL)
-# - `t5xxl.safetensors` (google T5-v1.1-XXL)
-# - `sd3_medium.safetensors` (or whichever main MMDiT model file)
-# Also can have
-# - `sd3_vae.safetensors` (holds the VAE separately if needed)
-
-import datetime
-import math
-import os
-
-import fire
-import numpy as np
-import sd3_impls
-import torch
-from other_impls import SD3Tokenizer, SDClipModel, SDXLClipG, T5XXLModel
-from PIL import Image
-from safetensors import safe_open
-from sd3_impls import SDVAE, BaseModel, CFGDenoiser, SD3LatentFormat
-from tqdm import tqdm
-
-#################################################################################################
-### Wrappers for model parts
-#################################################################################################
-
-
-def load_into(f, model, prefix, device, dtype=None):
-    """Just a debugging-friendly hack to apply the weights in a safetensors file to the pytorch module."""
-    for key in f.keys():
-        if key.startswith(prefix) and not key.startswith("loss."):
-            path = key[len(prefix) :].split(".")
-            obj = model
-            for p in path:
-                if obj is list:
-                    obj = obj[int(p)]
-                else:
-                    obj = getattr(obj, p, None)
-                    if obj is None:
-                        print(f"Skipping key '{key}' in safetensors file as '{p}' does not exist in python model")
-                        break
-            if obj is None:
-                continue
-            try:
-                tensor = f.get_tensor(key).to(device=device)
-                if dtype is not None:
-                    tensor = tensor.to(dtype=dtype)
-                obj.requires_grad_(False)
-                obj.set_(tensor)
-            except Exception as e:
-                print(f"Failed to load key '{key}' in safetensors file: {e}")
-                raise e
-
-
-CLIPG_CONFIG = {
-    "hidden_act": "gelu",
-    "hidden_size": 1280,
-    "intermediate_size": 5120,
-    "num_attention_heads": 20,
-    "num_hidden_layers": 32,
-}
-
-
-class ClipG:
-    def __init__(self):
-        with safe_open("models/clip_g.safetensors", framework="pt", device="cpu") as f:
-            self.model = SDXLClipG(CLIPG_CONFIG, device="cpu", dtype=torch.float32)
-            load_into(f, self.model.transformer, "", "cpu", torch.float32)
-
-
-CLIPL_CONFIG = {
-    "hidden_act": "quick_gelu",
-    "hidden_size": 768,
-    "intermediate_size": 3072,
-    "num_attention_heads": 12,
-    "num_hidden_layers": 12,
-}
-
-
-class ClipL:
-    def __init__(self):
-        with safe_open("models/clip_l.safetensors", framework="pt", device="cpu") as f:
-            self.model = SDClipModel(
-                layer="hidden",
-                layer_idx=-2,
-                device="cpu",
-                dtype=torch.float32,
-                layer_norm_hidden_state=False,
-                return_projected_pooled=False,
-                textmodel_json_config=CLIPL_CONFIG,
-            )
-            load_into(f, self.model.transformer, "", "cpu", torch.float32)
-
-
-T5_CONFIG = {
-    "d_ff": 10240,
-    "d_model": 4096,
-    "num_heads": 64,
-    "num_layers": 24,
-    "vocab_size": 32128,
-}
-
-
-class T5XXL:
-    def __init__(self):
-        with safe_open("models/t5xxl.safetensors", framework="pt", device="cpu") as f:
-            self.model = T5XXLModel(T5_CONFIG, device="cpu", dtype=torch.float32)
-            load_into(f, self.model.transformer, "", "cpu", torch.float32)
-
-
-class SD3:
-    def __init__(self, model, shift, verbose=False):
-        with safe_open(model, framework="pt", device="cpu") as f:
-            self.model = BaseModel(
-                shift=shift,
-                file=f,
-                prefix="model.diffusion_model.",
-                device="cpu",
-                dtype=torch.float16,
-                verbose=verbose,
-            ).eval()
-            load_into(f, self.model, "model.", "cpu", torch.float16)
-
-
-class VAE:
-    def __init__(self, model):
-        with safe_open(model, framework="pt", device="cpu") as f:
-            self.model = SDVAE(device="cpu", dtype=torch.float16).eval().cpu()
-            prefix = ""
-            if any(k.startswith("first_stage_model.") for k in f.keys()):
-                prefix = "first_stage_model."
-            load_into(f, self.model, prefix, "cpu", torch.float16)
-
-
-#################################################################################################
-### Main inference logic
-#################################################################################################
-
-
-# Note: Sigma shift value, publicly released models use 3.0
-SHIFT = 3.0
-# Naturally, adjust to the width/height of the model you have
-WIDTH = 1024
-HEIGHT = 1024
-# Pick your prompt
-PROMPT = "a photo of a cat"
-# Most models prefer the range of 4-5, but still work well around 7
-CFG_SCALE = 4.5
-# Different models want different step counts but most will be good at 50, albeit that's slow to run
-# sd3_medium is quite decent at 28 steps
-STEPS = 40
-# Seed
-SEED = 23
-# SEEDTYPE = "fixed"
-SEEDTYPE = "rand"
-# SEEDTYPE = "roll"
-# Actual model file path
-# MODEL = "models/sd3_medium.safetensors"
-# MODEL = "models/sd3.5_large_turbo.safetensors"
-MODEL = "models/sd3.5_large.safetensors"
-# VAE model file path, or set None to use the same model file
-VAEFile = None  # "models/sd3_vae.safetensors"
-# Optional init image file path
-INIT_IMAGE = None
-# If init_image is given, this is the percentage of denoising steps to run (1.0 = full denoise, 0.0 = no denoise at all)
-DENOISE = 0.6
-# Output file path
-OUTDIR = "outputs"
-# SAMPLER
-# SAMPLER = "euler"
-SAMPLER = "dpmpp_2m"
-
-
-class SD3Inferencer:
-    def print(self, txt):
-        if self.verbose:
-            print(txt)
-
-    def load(self, model=MODEL, vae=VAEFile, shift=SHIFT, verbose=False):
-        self.verbose = verbose
-        print("Loading tokenizers...")
-        # NOTE: if you need a reference impl for a high performance CLIP tokenizer instead of just using the HF transformers one,
-        # check https://github.com/Stability-AI/StableSwarmUI/blob/master/src/Utils/CliplikeTokenizer.cs
-        # (T5 tokenizer is different though)
-        self.tokenizer = SD3Tokenizer()
-        print("Loading OpenAI CLIP L...")
-        self.clip_l = ClipL()
-        print("Loading OpenCLIP bigG...")
-        self.clip_g = ClipG()
-        print("Loading Google T5-v1-XXL...")
-        self.t5xxl = T5XXL()
-        print(f"Loading SD3 model {os.path.basename(model)}...")
-        self.sd3 = SD3(model, shift, verbose)
-        print("Loading VAE model...")
-        self.vae = VAE(vae or model)
-        print("Models loaded.")
-
-    def get_empty_latent(self, width, height):
-        self.print("Prep an empty latent...")
-        return torch.ones(1, 16, height // 8, width // 8, device="cpu") * 0.0609
-
-    def get_sigmas(self, sampling, steps):
-        start = sampling.timestep(sampling.sigma_max)
-        end = sampling.timestep(sampling.sigma_min)
-        timesteps = torch.linspace(start, end, steps)
-        sigs = []
-        for x in range(len(timesteps)):
-            ts = timesteps[x]
-            sigs.append(sampling.sigma(ts))
-        sigs += [0.0]
-        return torch.FloatTensor(sigs)
-
-    def get_noise(self, seed, latent):
-        generator = torch.manual_seed(seed)
-        self.print(f"dtype = {latent.dtype}, layout = {latent.layout}, device = {latent.device}")
-        return torch.randn(
-            latent.size(),
-            dtype=torch.float32,
-            layout=latent.layout,
-            generator=generator,
-            device="cpu",
-        ).to(latent.dtype)
-
-    def get_cond(self, prompt):
-        self.print("Encode prompt...")
-        tokens = self.tokenizer.tokenize_with_weights(prompt)
-        l_out, l_pooled = self.clip_l.model.encode_token_weights(tokens["l"])
-        g_out, g_pooled = self.clip_g.model.encode_token_weights(tokens["g"])
-        t5_out, t5_pooled = self.t5xxl.model.encode_token_weights(tokens["t5xxl"])
-        lg_out = torch.cat([l_out, g_out], dim=-1)
-        lg_out = torch.nn.functional.pad(lg_out, (0, 4096 - lg_out.shape[-1]))
-        return torch.cat([lg_out, t5_out], dim=-2), torch.cat((l_pooled, g_pooled), dim=-1)
-
-    def max_denoise(self, sigmas):
-        max_sigma = float(self.sd3.model.model_sampling.sigma_max)
-        sigma = float(sigmas[0])
-        return math.isclose(max_sigma, sigma, rel_tol=1e-05) or sigma > max_sigma
-
-    def fix_cond(self, cond):
-        cond, pooled = (cond[0].half().cuda(), cond[1].half().cuda())
-        return {"c_crossattn": cond, "y": pooled}
-
-    def do_sampling(
-        self,
-        latent,
-        seed,
-        conditioning,
-        neg_cond,
-        steps,
-        cfg_scale,
-        sampler="dpmpp_2m",
-        denoise=1.0,
-    ) -> torch.Tensor:
-        self.print("Sampling...")
-        latent = latent.half().cuda()
-        self.sd3.model = self.sd3.model.cuda()
-        noise = self.get_noise(seed, latent).cuda()
-        sigmas = self.get_sigmas(self.sd3.model.model_sampling, steps).cuda()
-        sigmas = sigmas[int(steps * (1 - denoise)) :]
-        conditioning = self.fix_cond(conditioning)
-        neg_cond = self.fix_cond(neg_cond)
-        extra_args = {"cond": conditioning, "uncond": neg_cond, "cond_scale": cfg_scale}
-        noise_scaled = self.sd3.model.model_sampling.noise_scaling(sigmas[0], noise, latent, self.max_denoise(sigmas))
-        sample_fn = getattr(sd3_impls, f"sample_{sampler}")
-        latent = sample_fn(CFGDenoiser(self.sd3.model), noise_scaled, sigmas, extra_args=extra_args)
-        latent = SD3LatentFormat().process_out(latent)
-        self.sd3.model = self.sd3.model.cpu()
-        self.print("Sampling done")
-        return latent
-
-    def vae_encode(self, image) -> torch.Tensor:
-        self.print("Encoding image to latent...")
-        image = image.convert("RGB")
-        image_np = np.array(image).astype(np.float32) / 255.0
-        image_np = np.moveaxis(image_np, 2, 0)
-        batch_images = np.expand_dims(image_np, axis=0).repeat(1, axis=0)
-        image_torch = torch.from_numpy(batch_images)
-        image_torch = 2.0 * image_torch - 1.0
-        image_torch = image_torch.cuda()
-        self.vae.model = self.vae.model.cuda()
-        latent = self.vae.model.encode(image_torch).cpu()
-        self.vae.model = self.vae.model.cpu()
-        self.print("Encoded")
-        return latent
-
-    def vae_decode(self, latent) -> Image.Image:
-        self.print("Decoding latent to image...")
-        latent = latent.cuda()
-        self.vae.model = self.vae.model.cuda()
-        image = self.vae.model.decode(latent)
-        image = image.float()
-        self.vae.model = self.vae.model.cpu()
-        image = torch.clamp((image + 1.0) / 2.0, min=0.0, max=1.0)[0]
-        decoded_np = 255.0 * np.moveaxis(image.cpu().numpy(), 0, 2)
-        decoded_np = decoded_np.astype(np.uint8)
-        out_image = Image.fromarray(decoded_np)
-        self.print("Decoded")
-        return out_image
-
-    def gen_image(
-        self,
-        prompts=[PROMPT],
-        width=WIDTH,
-        height=HEIGHT,
-        steps=STEPS,
-        cfg_scale=CFG_SCALE,
-        sampler=SAMPLER,
-        seed=SEED,
-        seed_type=SEEDTYPE,
-        out_dir=OUTDIR,
-        init_image=INIT_IMAGE,
-        denoise=DENOISE,
-    ):
-        latent = self.get_empty_latent(width, height)
-        if init_image:
-            image_data = Image.open(init_image)
-            image_data = image_data.resize((width, height), Image.LANCZOS)
-            latent = self.vae_encode(image_data)
-            latent = SD3LatentFormat().process_in(latent)
-        neg_cond = self.get_cond("")
-        seed_num = None
-        pbar = tqdm(enumerate(prompts), total=len(prompts), position=0, leave=True)
-        for i, prompt in pbar:
-            if seed_type == "roll":
-                seed_num = seed if seed_num is None else seed_num + 1
-            elif seed_type == "rand":
-                seed_num = torch.randint(0, 100000, (1,)).item()
-            else:  # fixed
-                seed_num = seed
-            conditioning = self.get_cond(prompt)
-            sampled_latent = self.do_sampling(
-                latent,
-                seed_num,
-                conditioning,
-                neg_cond,
-                steps,
-                cfg_scale,
-                sampler,
-                denoise if init_image else 1.0,
-            )
-            image = self.vae_decode(sampled_latent)
-            save_path = os.path.join(out_dir, f"{i:06d}.png")
-            self.print(f"Will save to {save_path}")
-            image.save(save_path)
-            self.print("Done")
-
-
-CONFIGS = {
-    "sd3_medium": {
-        "shift": 1.0,
-        "cfg": 5.0,
-        "steps": 50,
-        "sampler": "dpmpp_2m",
-    },
-    "sd3.5_large": {
-        "shift": 3.0,
-        "cfg": 4.5,
-        "steps": 40,
-        "sampler": "dpmpp_2m",
-    },
-    "sd3.5_large_turbo": {"shift": 3.0, "cfg": 1.0, "steps": 4, "sampler": "euler"},
-}
-
-
-@torch.no_grad()
-def main(
-    prompt=PROMPT,
-    model=MODEL,
-    out_dir=OUTDIR,
-    postfix=None,
-    seed=SEED,
-    seed_type=SEEDTYPE,
-    sampler=None,
-    steps=None,
-    cfg=None,
-    shift=None,
-    width=WIDTH,
-    height=HEIGHT,
-    vae=VAEFile,
-    init_image=INIT_IMAGE,
-    denoise=DENOISE,
-    verbose=False,
-):
-    steps = steps or CONFIGS[os.path.splitext(os.path.basename(model))[0]]["steps"]
-    cfg = cfg or CONFIGS[os.path.splitext(os.path.basename(model))[0]]["cfg"]
-    shift = shift or CONFIGS[os.path.splitext(os.path.basename(model))[0]]["shift"]
-    sampler = sampler or CONFIGS[os.path.splitext(os.path.basename(model))[0]]["sampler"]
-
-    inferencer = SD3Inferencer()
-    inferencer.load(model, vae, shift, verbose)
-
-    if isinstance(prompt, str):
-        if os.path.splitext(prompt)[-1] == ".txt":
-            with open(prompt, "r") as f:
-                prompts = [l.strip() for l in f.readlines()]
-        else:
-            prompts = [prompt]
-
-    out_dir = os.path.join(
-        out_dir,
-        os.path.splitext(os.path.basename(model))[0],
-        os.path.splitext(os.path.basename(prompt))[0][:50]
-        + (postfix or datetime.datetime.now().strftime("_%Y-%m-%dT%H-%M-%S")),
-    )
-    print(f"Saving images to {out_dir}")
-    os.makedirs(out_dir, exist_ok=False)
-
-    inferencer.gen_image(
-        prompts,
-        width,
-        height,
-        steps,
-        cfg,
-        sampler,
-        seed,
-        seed_type,
-        out_dir,
-        init_image,
-        denoise,
-    )
-
-
-fire.Fire(main)

invokeai/backend/sd3/sd3_mmditx.py

@@ -1,72 +0,0 @@
-from dataclasses import dataclass
-from typing import Literal, TypedDict
-
-import torch
-
-from invokeai.backend.sd3.mmditx import MMDiTX
-from invokeai.backend.sd3.sd3_impls import ModelSamplingDiscreteFlow
-
-
-class ContextEmbedderConfig(TypedDict):
-    target: Literal["torch.nn.Linear"]
-    params: dict[str, int]
-
-
-@dataclass
-class Sd3MMDiTXParams:
-    patch_size: int
-    depth: int
-    num_patches: int
-    pos_embed_max_size: int
-    adm_in_channels: int
-    context_shape: tuple[int, int]
-    qk_norm: Literal["rms", None]
-    x_block_self_attn_layers: list[int]
-    context_embedder_config: ContextEmbedderConfig
-
-
-class Sd3MMDiTX(torch.nn.Module):
-    """This class is based closely on
-    https://github.com/Stability-AI/sd3.5/blob/19bf11c4e1e37324c5aa5a61f010d4127848a09c/sd3_impls.py#L53
-    but has more standard model loading semantics.
-    """
-
-    def __init__(
-        self,
-        params: Sd3MMDiTXParams,
-        shift: float = 1.0,
-        device: torch.device | None = None,
-        dtype: torch.dtype | None = None,
-        verbose: bool = False,
-    ):
-        super().__init__()
-        self.diffusion_model = MMDiTX(
-            input_size=None,
-            pos_embed_scaling_factor=None,
-            pos_embed_offset=None,
-            pos_embed_max_size=params.pos_embed_max_size,
-            patch_size=params.patch_size,
-            in_channels=16,
-            depth=params.depth,
-            num_patches=params.num_patches,
-            adm_in_channels=params.adm_in_channels,
-            context_embedder_config=params.context_embedder_config,
-            qk_norm=params.qk_norm,
-            x_block_self_attn_layers=params.x_block_self_attn_layers,
-            device=device,
-            dtype=dtype,
-            verbose=verbose,
-        )
-        self.model_sampling = ModelSamplingDiscreteFlow(shift=shift)
-
-    def apply_model(self, x: torch.Tensor, sigma: torch.Tensor, c_crossattn: torch.Tensor, y: torch.Tensor):
-        dtype = self.get_dtype()
-        timestep = self.model_sampling.timestep(sigma).float()
-        model_output = self.diffusion_model(x.to(dtype), timestep, context=c_crossattn.to(dtype), y=y.to(dtype)).float()
-        return self.model_sampling.calculate_denoised(sigma, model_output, x)
-
-    def forward(self, x: torch.Tensor, sigma: float, c_crossattn: torch.Tensor, y: torch.Tensor):
-        return self.apply_model(x=x, sigma=sigma, c_crossattn=c_crossattn, y=y)
-
-    def get_dtype(self):
-        return self.diffusion_model.dtype

invokeai/backend/sd3/sd3_state_dict_utils.py

@@ -1,70 +0,0 @@
-import math
-import re
-from typing import Any, Dict
-
-from invokeai.backend.sd3.sd3_mmditx import ContextEmbedderConfig, Sd3MMDiTXParams
-
-
-def is_sd3_checkpoint(sd: Dict[str, Any]) -> bool:
-    """Is the state dict for an SD3 checkpoint like this one?:
-    https://huggingface.co/stabilityai/stable-diffusion-3.5-large/blob/main/sd3.5_large.safetensors
-
-    Note that this checkpoint format contains both the VAE and the MMDiTX model.
-
-    This is intended to be a reasonably high-precision detector, but it is not guaranteed to have perfect precision.
-    """
-    # If all of the expected keys are present, then this is very likely a SD3 checkpoint.
-    expected_keys = {
-        # VAE decoder and encoder keys.
-        "first_stage_model.decoder.conv_in.bias",
-        "first_stage_model.decoder.conv_in.weight",
-        "first_stage_model.encoder.conv_in.bias",
-        "first_stage_model.encoder.conv_in.weight",
-        # MMDiTX keys.
-        "model.diffusion_model.final_layer.linear.bias",
-        "model.diffusion_model.final_layer.linear.weight",
-        "model.diffusion_model.joint_blocks.0.context_block.attn.ln_k.weight",
-        "model.diffusion_model.joint_blocks.0.context_block.attn.ln_q.weight",
-    }
-
-    return expected_keys.issubset(sd.keys())
-
-
-def infer_sd3_mmditx_params(sd: Dict[str, Any], prefix: str = "model.diffusion_model.") -> Sd3MMDiTXParams:
-    """Infer the MMDiTX model parameters from the state dict.
-
-    This logic is based on:
-    https://github.com/Stability-AI/sd3.5/blob/19bf11c4e1e37324c5aa5a61f010d4127848a09c/sd3_impls.py#L68-L88
-    """
-    patch_size = sd[f"{prefix}x_embedder.proj.weight"].shape[2]
-    depth = sd[f"{prefix}x_embedder.proj.weight"].shape[0] // 64
-    num_patches = sd[f"{prefix}pos_embed"].shape[1]
-    pos_embed_max_size = round(math.sqrt(num_patches))
-    adm_in_channels = sd[f"{prefix}y_embedder.mlp.0.weight"].shape[1]
-    context_shape = sd[f"{prefix}context_embedder.weight"].shape
-    qk_norm = "rms" if f"{prefix}joint_blocks.0.context_block.attn.ln_k.weight" in sd else None
-    x_block_self_attn_layers = sorted(
-        [
-            int(key.split(".x_block.attn2.ln_k.weight")[0].split(".")[-1])
-            for key in list(filter(re.compile(".*.x_block.attn2.ln_k.weight").match, sd.keys()))
-        ]
-    )
-
-    context_embedder_config: ContextEmbedderConfig = {
-        "target": "torch.nn.Linear",
-        "params": {
-            "in_features": context_shape[1],
-            "out_features": context_shape[0],
-        },
-    }
-    return Sd3MMDiTXParams(
-        patch_size=patch_size,
-        depth=depth,
-        num_patches=num_patches,
-        pos_embed_max_size=pos_embed_max_size,
-        adm_in_channels=adm_in_channels,
-        context_shape=context_shape,
-        qk_norm=qk_norm,
-        x_block_self_attn_layers=x_block_self_attn_layers,
-        context_embedder_config=context_embedder_config,
-    )

invokeai/frontend/web/package.json

@@ -58,7 +58,7 @@
     "@dnd-kit/sortable": "^8.0.0",
     "@dnd-kit/utilities": "^3.2.2",
     "@fontsource-variable/inter": "^5.1.0",
-    "@invoke-ai/ui-library": "^0.0.42",
+    "@invoke-ai/ui-library": "^0.0.43",
     "@nanostores/react": "^0.7.3",
     "@reduxjs/toolkit": "2.2.3",
     "@roarr/browser-log-writer": "^1.3.0",

invokeai/frontend/web/pnpm-lock.yaml

@@ -24,8 +24,8 @@ dependencies:
     specifier: ^5.1.0
     version: 5.1.0
   '@invoke-ai/ui-library':
-    specifier: ^0.0.42
-    version: 0.0.42(@chakra-ui/form-control@2.2.0)(@chakra-ui/icon@3.2.0)(@chakra-ui/media-query@3.3.0)(@chakra-ui/menu@2.2.1)(@chakra-ui/spinner@2.1.0)(@chakra-ui/system@2.6.2)(@fontsource-variable/inter@5.1.0)(@types/react@18.3.11)(i18next@23.15.1)(react-dom@18.3.1)(react@18.3.1)
+    specifier: ^0.0.43
+    version: 0.0.43(@chakra-ui/form-control@2.2.0)(@chakra-ui/icon@3.2.0)(@chakra-ui/media-query@3.3.0)(@chakra-ui/menu@2.2.1)(@chakra-ui/spinner@2.1.0)(@chakra-ui/system@2.6.2)(@fontsource-variable/inter@5.1.0)(@types/react@18.3.11)(i18next@23.15.1)(react-dom@18.3.1)(react@18.3.1)
   '@nanostores/react':
     specifier: ^0.7.3
     version: 0.7.3(nanostores@0.11.3)(react@18.3.1)
@@ -1696,20 +1696,20 @@ packages:
       prettier: 3.3.3
     dev: true
 
-  /@invoke-ai/ui-library@0.0.42(@chakra-ui/form-control@2.2.0)(@chakra-ui/icon@3.2.0)(@chakra-ui/media-query@3.3.0)(@chakra-ui/menu@2.2.1)(@chakra-ui/spinner@2.1.0)(@chakra-ui/system@2.6.2)(@fontsource-variable/inter@5.1.0)(@types/react@18.3.11)(i18next@23.15.1)(react-dom@18.3.1)(react@18.3.1):
-    resolution: {integrity: sha512-OuDXRipBO5mu+Nv4qN8cd8MiwiGBdq6h4PirVgPI9/ltbdcIzePgUJ0dJns26lflHSTRWW38I16wl4YTw3mNWA==}
+  /@invoke-ai/ui-library@0.0.43(@chakra-ui/form-control@2.2.0)(@chakra-ui/icon@3.2.0)(@chakra-ui/media-query@3.3.0)(@chakra-ui/menu@2.2.1)(@chakra-ui/spinner@2.1.0)(@chakra-ui/system@2.6.2)(@fontsource-variable/inter@5.1.0)(@types/react@18.3.11)(i18next@23.15.1)(react-dom@18.3.1)(react@18.3.1):
+    resolution: {integrity: sha512-t3fPYyks07ue3dEBPJuTHbeDLnDckDCOrtvc07mMDbLOnlPEZ0StaeiNGH+oO8qLzAuMAlSTdswgHfzTc2MmPw==}
     peerDependencies:
       '@fontsource-variable/inter': ^5.0.16
       react: ^18.2.0
       react-dom: ^18.2.0
     dependencies:
-      '@chakra-ui/anatomy': 2.2.2
+      '@chakra-ui/anatomy': 2.3.4
       '@chakra-ui/icons': 2.2.4(@chakra-ui/react@2.10.2)(react@18.3.1)
       '@chakra-ui/layout': 2.3.1(@chakra-ui/system@2.6.2)(react@18.3.1)
       '@chakra-ui/portal': 2.1.0(react-dom@18.3.1)(react@18.3.1)
       '@chakra-ui/react': 2.10.2(@emotion/react@11.13.3)(@emotion/styled@11.13.0)(@types/react@18.3.11)(framer-motion@11.10.0)(react-dom@18.3.1)(react@18.3.1)
-      '@chakra-ui/styled-system': 2.9.2
-      '@chakra-ui/theme-tools': 2.1.2(@chakra-ui/styled-system@2.9.2)
+      '@chakra-ui/styled-system': 2.11.2(react@18.3.1)
+      '@chakra-ui/theme-tools': 2.2.6(@chakra-ui/styled-system@2.11.2)(react@18.3.1)
       '@emotion/react': 11.13.3(@types/react@18.3.11)(react@18.3.1)
       '@emotion/styled': 11.13.0(@emotion/react@11.13.3)(@types/react@18.3.11)(react@18.3.1)
       '@fontsource-variable/inter': 5.1.0

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

@@ -94,6 +94,7 @@
         "close": "Close",
         "copy": "Copy",
         "copyError": "$t(gallery.copy) Error",
+        "clipboard": "Clipboard",
         "on": "On",
         "off": "Off",
         "or": "or",
@@ -1251,6 +1252,33 @@
             "heading": "Mask Adjustments",
             "paragraphs": ["Adjust the mask."]
         },
+        "inpainting": {
+            "heading": "Inpainting",
+            "paragraphs": ["Controls which area is modified, guided by Denoising Strength."]
+        },
+        "rasterLayer": {
+            "heading": "Raster Layer",
+            "paragraphs": ["Pixel-based content of your canvas, used during image generation."]
+        },
+        "regionalGuidance": {
+            "heading": "Regional Guidance",
+            "paragraphs": ["Brush to guide where elements from global prompts should appear."]
+        },
+        "regionalGuidanceAndReferenceImage": {
+            "heading": "Regional Guidance and Regional Reference Image",
+            "paragraphs": [
+                "For Regional Guidance, brush to guide where elements from global prompts should appear.",
+                "For Regional Reference Image, brush to apply a reference image to specific areas."
+            ]
+        },
+        "globalReferenceImage": {
+            "heading": "Global Reference Image",
+            "paragraphs": ["Applies a reference image to influence the entire generation."]
+        },
+        "regionalReferenceImage": {
+            "heading": "Regional Reference Image",
+            "paragraphs": ["Brush to apply a reference image to specific areas."]
+        },
         "controlNet": {
             "heading": "ControlNet",
             "paragraphs": [
@@ -1688,8 +1716,18 @@
         "layer_other": "Layers",
         "layer_withCount_one": "Layer ({{count}})",
         "layer_withCount_other": "Layers ({{count}})",
-        "convertToControlLayer": "Convert to Control Layer",
-        "convertToRasterLayer": "Convert to Raster Layer",
+        "convertRasterLayerTo": "Convert $t(controlLayers.rasterLayer) To",
+        "convertControlLayerTo": "Convert $t(controlLayers.controlLayer) To",
+        "convertInpaintMaskTo": "Convert $t(controlLayers.inpaintMask) To",
+        "convertRegionalGuidanceTo": "Convert $t(controlLayers.regionalGuidance) To",
+        "copyRasterLayerTo": "Copy $t(controlLayers.rasterLayer) To",
+        "copyControlLayerTo": "Copy $t(controlLayers.controlLayer) To",
+        "copyInpaintMaskTo": "Copy $t(controlLayers.inpaintMask) To",
+        "copyRegionalGuidanceTo": "Copy $t(controlLayers.regionalGuidance) To",
+        "newRasterLayer": "New $t(controlLayers.rasterLayer)",
+        "newControlLayer": "New $t(controlLayers.controlLayer)",
+        "newInpaintMask": "New $t(controlLayers.inpaintMask)",
+        "newRegionalGuidance": "New $t(controlLayers.regionalGuidance)",
         "transparency": "Transparency",
         "enableTransparencyEffect": "Enable Transparency Effect",
         "disableTransparencyEffect": "Disable Transparency Effect",
@@ -1845,11 +1883,11 @@
         "segment": {
             "autoMask": "Auto Mask",
             "pointType": "Point Type",
-            "foreground": "Foreground",
-            "background": "Background",
+            "include": "Include",
+            "exclude": "Exclude",
             "neutral": "Neutral",
             "reset": "Reset",
-            "apply": "Apply",
+            "saveAs": "Save As",
             "cancel": "Cancel",
             "process": "Process"
         },

invokeai/frontend/web/src/common/components/IconMenuItem.tsx

@@ -26,5 +26,9 @@ export const IconMenuItem = ({ tooltip, icon, ...props }: Props) => {
 };
 
 export const IconMenuItemGroup = ({ children }: { children: ReactNode }) => {
-  return <Flex gap={2}>{children}</Flex>;
+  return (
+    <Flex gap={2} justifyContent="space-between">
+      {children}
+    </Flex>
+  );
 };

invokeai/frontend/web/src/common/components/InformationalPopover/constants.ts

@@ -23,8 +23,10 @@ export type Feature =
   | 'dynamicPrompts'
   | 'dynamicPromptsMaxPrompts'
   | 'dynamicPromptsSeedBehaviour'
+  | 'globalReferenceImage'
   | 'imageFit'
   | 'infillMethod'
+  | 'inpainting'
   | 'ipAdapterMethod'
   | 'lora'
   | 'loraWeight'
@@ -46,6 +48,7 @@ export type Feature =
   | 'paramVAEPrecision'
   | 'paramWidth'
   | 'patchmatchDownScaleSize'
+  | 'rasterLayer'
   | 'refinerModel'
   | 'refinerNegativeAestheticScore'
   | 'refinerPositiveAestheticScore'
@@ -53,6 +56,9 @@ export type Feature =
   | 'refinerStart'
   | 'refinerSteps'
   | 'refinerCfgScale'
+  | 'regionalGuidance'
+  | 'regionalGuidanceAndReferenceImage'
+  | 'regionalReferenceImage'
   | 'scaleBeforeProcessing'
   | 'seamlessTilingXAxis'
   | 'seamlessTilingYAxis'
@@ -76,6 +82,24 @@ export const POPOVER_DATA: { [key in Feature]?: PopoverData } = {
   clipSkip: {
     href: 'https://support.invoke.ai/support/solutions/articles/151000178161-advanced-settings',
   },
+  inpainting: {
+    href: 'https://support.invoke.ai/support/solutions/articles/151000096702-inpainting-outpainting-and-bounding-box',
+  },
+  rasterLayer: {
+    href: 'https://support.invoke.ai/support/solutions/articles/151000094998-raster-layers-and-initial-images',
+  },
+  regionalGuidance: {
+    href: 'https://support.invoke.ai/support/solutions/articles/151000165024-regional-guidance-layers',
+  },
+  regionalGuidanceAndReferenceImage: {
+    href: 'https://support.invoke.ai/support/solutions/articles/151000165024-regional-guidance-layers',
+  },
+  globalReferenceImage: {
+    href: 'https://support.invoke.ai/support/solutions/articles/151000159340-global-and-regional-reference-images-ip-adapters-',
+  },
+  regionalReferenceImage: {
+    href: 'https://support.invoke.ai/support/solutions/articles/151000159340-global-and-regional-reference-images-ip-adapters-',
+  },
   controlNet: {
     href: 'https://support.invoke.ai/support/solutions/articles/151000105880',
   },

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

@@ -127,8 +127,6 @@ export const buildUseDisclosure = (defaultIsOpen: boolean): [() => UseDisclosure
  *
  * Hook to manage a boolean state. Use this for a local boolean state.
  * @param defaultIsOpen Initial state of the disclosure
- *
- * @knipignore
  */
 export const useDisclosure = (defaultIsOpen: boolean): UseDisclosure => {
   const [isOpen, set] = useState(defaultIsOpen);

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

@@ -16,6 +16,7 @@ type UseGroupedModelComboboxArg<T extends AnyModelConfig> = {
   getIsDisabled?: (model: T) => boolean;
   isLoading?: boolean;
   groupByType?: boolean;
+  showDescriptions?: boolean;
 };
 
 type UseGroupedModelComboboxReturn = {
@@ -37,7 +38,15 @@ export const useGroupedModelCombobox = <T extends AnyModelConfig>(
 ): UseGroupedModelComboboxReturn => {
   const { t } = useTranslation();
   const base = useAppSelector(selectBaseWithSDXLFallback);
-  const { modelConfigs, selectedModel, getIsDisabled, onChange, isLoading, groupByType = false } = arg;
+  const {
+    modelConfigs,
+    selectedModel,
+    getIsDisabled,
+    onChange,
+    isLoading,
+    groupByType = false,
+    showDescriptions = false,
+  } = arg;
   const options = useMemo<GroupBase<ComboboxOption>[]>(() => {
     if (!modelConfigs) {
       return [];
@@ -51,6 +60,7 @@ export const useGroupedModelCombobox = <T extends AnyModelConfig>(
           options: val.map((model) => ({
             label: model.name,
             value: model.key,
+            description: (showDescriptions && model.description) || undefined,
             isDisabled: getIsDisabled ? getIsDisabled(model) : false,
           })),
         });
@@ -60,7 +70,7 @@ export const useGroupedModelCombobox = <T extends AnyModelConfig>(
     );
     _options.sort((a) => (a.label?.split('/')[0]?.toLowerCase().includes(base) ? -1 : 1));
     return _options;
-  }, [modelConfigs, groupByType, getIsDisabled, base]);
+  }, [modelConfigs, groupByType, getIsDisabled, base, showDescriptions]);
 
   const value = useMemo(
     () =>

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

@@ -0,0 +1,161 @@
+import type { MenuButtonProps, MenuItemProps, MenuListProps, MenuProps } from '@invoke-ai/ui-library';
+import { Box, Flex, Icon, Text } from '@invoke-ai/ui-library';
+import { useDisclosure } from 'common/hooks/useBoolean';
+import type { FocusEventHandler, PointerEvent, RefObject } from 'react';
+import { useCallback, useEffect, useRef } from 'react';
+import { PiCaretRightBold } from 'react-icons/pi';
+import { useDebouncedCallback } from 'use-debounce';
+
+const offset: [number, number] = [0, 8];
+
+type UseSubMenuReturn = {
+  parentMenuItemProps: Partial<MenuItemProps>;
+  menuProps: Partial<MenuProps>;
+  menuButtonProps: Partial<MenuButtonProps>;
+  menuListProps: Partial<MenuListProps> & { ref: RefObject<HTMLDivElement> };
+};
+
+/**
+ * A hook that provides the necessary props to create a sub-menu within a menu.
+ *
+ * The sub-menu should be wrapped inside a parent `MenuItem` component.
+ *
+ * Use SubMenuButtonContent to render a button with a label and a right caret icon.
+ *
+ * TODO(psyche): Add keyboard handling for sub-menu.
+ *
+ * @example
+ * ```tsx
+ * const SubMenuExample = () => {
+ *   const subMenu = useSubMenu();
+ *   return (
+ *     <Menu>
+ *       <MenuButton>Open Parent Menu</MenuButton>
+ *       <MenuList>
+ *         <MenuItem>Parent Item 1</MenuItem>
+ *         <MenuItem>Parent Item 2</MenuItem>
+ *         <MenuItem>Parent Item 3</MenuItem>
+ *         <MenuItem {...subMenu.parentMenuItemProps} icon={<PiImageBold />}>
+ *           <Menu {...subMenu.menuProps}>
+ *             <MenuButton {...subMenu.menuButtonProps}>
+ *               <SubMenuButtonContent label="Open Sub Menu" />
+ *             </MenuButton>
+ *             <MenuList {...subMenu.menuListProps}>
+ *               <MenuItem>Sub Item 1</MenuItem>
+ *               <MenuItem>Sub Item 2</MenuItem>
+ *               <MenuItem>Sub Item 3</MenuItem>
+ *             </MenuList>
+ *           </Menu>
+ *         </MenuItem>
+ *       </MenuList>
+ *     </Menu>
+ *   );
+ * };
+ * ```
+ */
+export const useSubMenu = (): UseSubMenuReturn => {
+  const subMenu = useDisclosure(false);
+  const menuListRef = useRef<HTMLDivElement>(null);
+  const closeDebounced = useDebouncedCallback(subMenu.close, 300);
+  const openAndCancelPendingClose = useCallback(() => {
+    closeDebounced.cancel();
+    subMenu.open();
+  }, [closeDebounced, subMenu]);
+  const toggleAndCancelPendingClose = useCallback(() => {
+    if (subMenu.isOpen) {
+      subMenu.close();
+      return;
+    } else {
+      closeDebounced.cancel();
+      subMenu.toggle();
+    }
+  }, [closeDebounced, subMenu]);
+  const onBlurMenuList = useCallback<FocusEventHandler<HTMLDivElement>>(
+    (e) => {
+      // Don't trigger blur if focus is moving to a child element - e.g. from a sub-menu item to another sub-menu item
+      if (e.currentTarget.contains(e.relatedTarget)) {
+        closeDebounced.cancel();
+        return;
+      }
+      subMenu.close();
+    },
+    [closeDebounced, subMenu]
+  );
+
+  const onParentMenuItemPointerLeave = useCallback(
+    (e: PointerEvent<HTMLButtonElement>) => {
+      /**
+       * The pointerleave event is triggered when the pen or touch device is lifted, which would close the sub-menu.
+       * However, we want to keep the sub-menu open until the pen or touch device pressed some other element. This
+       * will be handled in the useEffect below - just ignore the pointerleave event for pen and touch devices.
+       */
+      if (e.pointerType === 'pen' || e.pointerType === 'touch') {
+        return;
+      }
+      subMenu.close();
+    },
+    [subMenu]
+  );
+
+  /**
+   * When using a mouse, the pointerleave events close the menu. But when using a pen or touch device, we need to close
+   * the sub-menu when the user taps outside of the menu list. So we need to listen for clicks outside of the menu list
+   * and close the menu accordingly.
+   */
+  useEffect(() => {
+    const el = menuListRef.current;
+    if (!el) {
+      return;
+    }
+    const controller = new AbortController();
+    window.addEventListener(
+      'click',
+      (e) => {
+        if (menuListRef.current?.contains(e.target as Node)) {
+          return;
+        }
+        subMenu.close();
+      },
+      { signal: controller.signal }
+    );
+    return () => {
+      controller.abort();
+    };
+  }, [subMenu]);
+
+  return {
+    parentMenuItemProps: {
+      onClick: toggleAndCancelPendingClose,
+      onPointerEnter: openAndCancelPendingClose,
+      onPointerLeave: onParentMenuItemPointerLeave,
+      closeOnSelect: false,
+    },
+    menuProps: {
+      isOpen: subMenu.isOpen,
+      onClose: subMenu.close,
+      placement: 'right',
+      offset: offset,
+      closeOnBlur: false,
+    },
+    menuButtonProps: {
+      as: Box,
+      width: 'full',
+      height: 'full',
+    },
+    menuListProps: {
+      ref: menuListRef,
+      onPointerEnter: openAndCancelPendingClose,
+      onPointerLeave: closeDebounced,
+      onBlur: onBlurMenuList,
+    },
+  };
+};
+
+export const SubMenuButtonContent = ({ label }: { label: string }) => {
+  return (
+    <Flex w="full" h="full" flexDir="row" justifyContent="space-between" alignItems="center">
+      <Text>{label}</Text>
+      <Icon as={PiCaretRightBold} />
+    </Flex>
+  );
+};

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

@@ -1,5 +1,6 @@
 import { Button, Flex, Heading } from '@invoke-ai/ui-library';
 import { useAppSelector } from 'app/store/storeHooks';
+import { InformationalPopover } from 'common/components/InformationalPopover/InformationalPopover';
 import {
   useAddControlLayer,
   useAddGlobalReferenceImage,
@@ -28,69 +29,80 @@ export const CanvasAddEntityButtons = memo(() => {
       <Flex position="relative" flexDir="column" gap={4} top="20%">
         <Flex flexDir="column" justifyContent="flex-start" gap={2}>
           <Heading size="xs">{t('controlLayers.global')}</Heading>
-          <Button
-            size="sm"
-            variant="ghost"
-            justifyContent="flex-start"
-            leftIcon={<PiPlusBold />}
-            onClick={addGlobalReferenceImage}
-          >
-            {t('controlLayers.globalReferenceImage')}
-          </Button>
+          <InformationalPopover feature="globalReferenceImage">
+            <Button
+              size="sm"
+              variant="ghost"
+              justifyContent="flex-start"
+              leftIcon={<PiPlusBold />}
+              onClick={addGlobalReferenceImage}
+            >
+              {t('controlLayers.globalReferenceImage')}
+            </Button>
+          </InformationalPopover>
         </Flex>
         <Flex flexDir="column" gap={2}>
           <Heading size="xs">{t('controlLayers.regional')}</Heading>
-          <Button
-            size="sm"
-            variant="ghost"
-            justifyContent="flex-start"
-            leftIcon={<PiPlusBold />}
-            onClick={addInpaintMask}
-          >
-            {t('controlLayers.inpaintMask')}
-          </Button>
-          <Button
-            size="sm"
-            variant="ghost"
-            justifyContent="flex-start"
-            leftIcon={<PiPlusBold />}
-            onClick={addRegionalGuidance}
-            isDisabled={isFLUX}
-          >
-            {t('controlLayers.regionalGuidance')}
-          </Button>
-          <Button
-            size="sm"
-            variant="ghost"
-            justifyContent="flex-start"
-            leftIcon={<PiPlusBold />}
-            onClick={addRegionalReferenceImage}
-            isDisabled={isFLUX}
-          >
-            {t('controlLayers.regionalReferenceImage')}
-          </Button>
+          <InformationalPopover feature="inpainting">
+            <Button
+              size="sm"
+              variant="ghost"
+              justifyContent="flex-start"
+              leftIcon={<PiPlusBold />}
+              onClick={addInpaintMask}
+            >
+              {t('controlLayers.inpaintMask')}
+            </Button>
+          </InformationalPopover>
+          <InformationalPopover feature="regionalGuidance">
+            <Button
+              size="sm"
+              variant="ghost"
+              justifyContent="flex-start"
+              leftIcon={<PiPlusBold />}
+              onClick={addRegionalGuidance}
+              isDisabled={isFLUX}
+            >
+              {t('controlLayers.regionalGuidance')}
+            </Button>
+          </InformationalPopover>
+          <InformationalPopover feature="regionalReferenceImage">
+            <Button
+              size="sm"
+              variant="ghost"
+              justifyContent="flex-start"
+              leftIcon={<PiPlusBold />}
+              onClick={addRegionalReferenceImage}
+              isDisabled={isFLUX}
+            >
+              {t('controlLayers.regionalReferenceImage')}
+            </Button>
+          </InformationalPopover>
         </Flex>
         <Flex flexDir="column" justifyContent="flex-start" gap={2}>
           <Heading size="xs">{t('controlLayers.layer_other')}</Heading>
-
-          <Button
-            size="sm"
-            variant="ghost"
-            justifyContent="flex-start"
-            leftIcon={<PiPlusBold />}
-            onClick={addControlLayer}
-          >
-            {t('controlLayers.controlLayer')}
-          </Button>
-          <Button
-            size="sm"
-            variant="ghost"
-            justifyContent="flex-start"
-            leftIcon={<PiPlusBold />}
-            onClick={addRasterLayer}
-          >
-            {t('controlLayers.rasterLayer')}
-          </Button>
+          <InformationalPopover feature="controlNet">
+            <Button
+              size="sm"
+              variant="ghost"
+              justifyContent="flex-start"
+              leftIcon={<PiPlusBold />}
+              onClick={addControlLayer}
+            >
+              {t('controlLayers.controlLayer')}
+            </Button>
+          </InformationalPopover>
+          <InformationalPopover feature="rasterLayer">
+            <Button
+              size="sm"
+              variant="ghost"
+              justifyContent="flex-start"
+              leftIcon={<PiPlusBold />}
+              onClick={addRasterLayer}
+            >
+              {t('controlLayers.rasterLayer')}
+            </Button>
+          </InformationalPopover>
         </Flex>
       </Flex>
     </Flex>

invokeai/frontend/web/src/features/controlLayers/components/CanvasContextMenu/CanvasContextMenuGlobalMenuItems.tsx

@@ -1,4 +1,5 @@
-import { MenuGroup, MenuItem } from '@invoke-ai/ui-library';
+import { Menu, MenuButton, MenuGroup, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
 import { CanvasContextMenuItemsCropCanvasToBbox } from 'features/controlLayers/components/CanvasContextMenu/CanvasContextMenuItemsCropCanvasToBbox';
 import { NewLayerIcon } from 'features/controlLayers/components/common/icons';
 import {
@@ -16,6 +17,8 @@ import { PiFloppyDiskBold } from 'react-icons/pi';
 
 export const CanvasContextMenuGlobalMenuItems = memo(() => {
   const { t } = useTranslation();
+  const saveSubMenu = useSubMenu();
+  const newSubMenu = useSubMenu();
   const isBusy = useCanvasIsBusy();
   const saveCanvasToGallery = useSaveCanvasToGallery();
   const saveBboxToGallery = useSaveBboxToGallery();
@@ -28,27 +31,41 @@ export const CanvasContextMenuGlobalMenuItems = memo(() => {
     <>
       <MenuGroup title={t('controlLayers.canvasContextMenu.canvasGroup')}>
         <CanvasContextMenuItemsCropCanvasToBbox />
-      </MenuGroup>
-      <MenuGroup title={t('controlLayers.canvasContextMenu.saveToGalleryGroup')}>
-        <MenuItem icon={<PiFloppyDiskBold />} isDisabled={isBusy} onClick={saveCanvasToGallery}>
-          {t('controlLayers.canvasContextMenu.saveCanvasToGallery')}
+        <MenuItem {...saveSubMenu.parentMenuItemProps} icon={<PiFloppyDiskBold />}>
+          <Menu {...saveSubMenu.menuProps}>
+            <MenuButton {...saveSubMenu.menuButtonProps}>
+              <SubMenuButtonContent label={t('controlLayers.canvasContextMenu.saveToGalleryGroup')} />
+            </MenuButton>
+            <MenuList {...saveSubMenu.menuListProps}>
+              <MenuItem icon={<PiFloppyDiskBold />} isDisabled={isBusy} onClick={saveCanvasToGallery}>
+                {t('controlLayers.canvasContextMenu.saveCanvasToGallery')}
+              </MenuItem>
+              <MenuItem icon={<PiFloppyDiskBold />} isDisabled={isBusy} onClick={saveBboxToGallery}>
+                {t('controlLayers.canvasContextMenu.saveBboxToGallery')}
+              </MenuItem>
+            </MenuList>
+          </Menu>
         </MenuItem>
-        <MenuItem icon={<PiFloppyDiskBold />} isDisabled={isBusy} onClick={saveBboxToGallery}>
-          {t('controlLayers.canvasContextMenu.saveBboxToGallery')}
-        </MenuItem>
-      </MenuGroup>
-      <MenuGroup title={t('controlLayers.canvasContextMenu.bboxGroup')}>
-        <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newGlobalReferenceImageFromBbox}>
-          {t('controlLayers.canvasContextMenu.newGlobalReferenceImage')}
-        </MenuItem>
-        <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newRegionalReferenceImageFromBbox}>
-          {t('controlLayers.canvasContextMenu.newRegionalReferenceImage')}
-        </MenuItem>
-        <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newControlLayerFromBbox}>
-          {t('controlLayers.canvasContextMenu.newControlLayer')}
-        </MenuItem>
-        <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newRasterLayerFromBbox}>
-          {t('controlLayers.canvasContextMenu.newRasterLayer')}
+        <MenuItem {...newSubMenu.parentMenuItemProps} icon={<NewLayerIcon />}>
+          <Menu {...newSubMenu.menuProps}>
+            <MenuButton {...newSubMenu.menuButtonProps}>
+              <SubMenuButtonContent label={t('controlLayers.canvasContextMenu.bboxGroup')} />
+            </MenuButton>
+            <MenuList {...newSubMenu.menuListProps}>
+              <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newGlobalReferenceImageFromBbox}>
+                {t('controlLayers.canvasContextMenu.newGlobalReferenceImage')}
+              </MenuItem>
+              <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newRegionalReferenceImageFromBbox}>
+                {t('controlLayers.canvasContextMenu.newRegionalReferenceImage')}
+              </MenuItem>
+              <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newControlLayerFromBbox}>
+                {t('controlLayers.canvasContextMenu.newControlLayer')}
+              </MenuItem>
+              <MenuItem icon={<NewLayerIcon />} isDisabled={isBusy} onClick={newRasterLayerFromBbox}>
+                {t('controlLayers.canvasContextMenu.newRasterLayer')}
+              </MenuItem>
+            </MenuList>
+          </Menu>
         </MenuItem>
       </MenuGroup>
     </>

invokeai/frontend/web/src/features/controlLayers/components/CanvasContextMenu/CanvasContextMenuSelectedEntityMenuItems.tsx

@@ -1,42 +1,40 @@
-import { MenuGroup } from '@invoke-ai/ui-library';
 import { useAppSelector } from 'app/store/storeHooks';
-import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
-import { CanvasEntityMenuItemsCropToBbox } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCropToBbox';
-import { CanvasEntityMenuItemsDelete } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDelete';
-import { CanvasEntityMenuItemsFilter } from 'features/controlLayers/components/common/CanvasEntityMenuItemsFilter';
-import { CanvasEntityMenuItemsSave } from 'features/controlLayers/components/common/CanvasEntityMenuItemsSave';
-import { CanvasEntityMenuItemsSegment } from 'features/controlLayers/components/common/CanvasEntityMenuItemsSegment';
-import { CanvasEntityMenuItemsTransform } from 'features/controlLayers/components/common/CanvasEntityMenuItemsTransform';
+import { ControlLayerMenuItems } from 'features/controlLayers/components/ControlLayer/ControlLayerMenuItems';
+import { InpaintMaskMenuItems } from 'features/controlLayers/components/InpaintMask/InpaintMaskMenuItems';
+import { IPAdapterMenuItems } from 'features/controlLayers/components/IPAdapter/IPAdapterMenuItems';
+import { RasterLayerMenuItems } from 'features/controlLayers/components/RasterLayer/RasterLayerMenuItems';
+import { RegionalGuidanceMenuItems } from 'features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItems';
 import {
   EntityIdentifierContext,
   useEntityIdentifierContext,
 } from 'features/controlLayers/contexts/EntityIdentifierContext';
-import { useEntityTitle } from 'features/controlLayers/hooks/useEntityTitle';
 import { selectSelectedEntityIdentifier } from 'features/controlLayers/store/selectors';
-import {
-  isFilterableEntityIdentifier,
-  isSaveableEntityIdentifier,
-  isSegmentableEntityIdentifier,
-  isTransformableEntityIdentifier,
-} from 'features/controlLayers/store/types';
 import { memo } from 'react';
+import type { Equals } from 'tsafe';
+import { assert } from 'tsafe';
 
 const CanvasContextMenuSelectedEntityMenuItemsContent = memo(() => {
   const entityIdentifier = useEntityIdentifierContext();
-  const title = useEntityTitle(entityIdentifier);
 
-  return (
-    <MenuGroup title={title}>
-      {isFilterableEntityIdentifier(entityIdentifier) && <CanvasEntityMenuItemsFilter />}
-      {isTransformableEntityIdentifier(entityIdentifier) && <CanvasEntityMenuItemsTransform />}
-      {isSegmentableEntityIdentifier(entityIdentifier) && <CanvasEntityMenuItemsSegment />}
-      {isSaveableEntityIdentifier(entityIdentifier) && <CanvasEntityMenuItemsCopyToClipboard />}
-      {isSaveableEntityIdentifier(entityIdentifier) && <CanvasEntityMenuItemsSave />}
-      {isTransformableEntityIdentifier(entityIdentifier) && <CanvasEntityMenuItemsCropToBbox />}
-      <CanvasEntityMenuItemsDelete />
-    </MenuGroup>
-  );
+  if (entityIdentifier.type === 'raster_layer') {
+    return <RasterLayerMenuItems />;
+  }
+  if (entityIdentifier.type === 'control_layer') {
+    return <ControlLayerMenuItems />;
+  }
+  if (entityIdentifier.type === 'inpaint_mask') {
+    return <InpaintMaskMenuItems />;
+  }
+  if (entityIdentifier.type === 'regional_guidance') {
+    return <RegionalGuidanceMenuItems />;
+  }
+  if (entityIdentifier.type === 'reference_image') {
+    return <IPAdapterMenuItems />;
+  }
+
+  assert<Equals<typeof entityIdentifier.type, never>>(false);
 });
+
 CanvasContextMenuSelectedEntityMenuItemsContent.displayName = 'CanvasContextMenuSelectedEntityMenuItemsContent';
 
 export const CanvasContextMenuSelectedEntityMenuItems = memo(() => {

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

@@ -1,5 +1,6 @@
 import { Flex, Spacer } from '@invoke-ai/ui-library';
 import { EntityListGlobalActionBarAddLayerMenu } from 'features/controlLayers/components/CanvasEntityList/EntityListGlobalActionBarAddLayerMenu';
+import { EntityListSelectedEntityActionBarAutoMaskButton } from 'features/controlLayers/components/CanvasEntityList/EntityListSelectedEntityActionBarAutoMaskButton';
 import { EntityListSelectedEntityActionBarDuplicateButton } from 'features/controlLayers/components/CanvasEntityList/EntityListSelectedEntityActionBarDuplicateButton';
 import { EntityListSelectedEntityActionBarFill } from 'features/controlLayers/components/CanvasEntityList/EntityListSelectedEntityActionBarFill';
 import { EntityListSelectedEntityActionBarFilterButton } from 'features/controlLayers/components/CanvasEntityList/EntityListSelectedEntityActionBarFilterButton';
@@ -16,6 +17,7 @@ export const EntityListSelectedEntityActionBar = memo(() => {
       <Spacer />
       <EntityListSelectedEntityActionBarFill />
       <Flex h="full">
+        <EntityListSelectedEntityActionBarAutoMaskButton />
         <EntityListSelectedEntityActionBarFilterButton />
         <EntityListSelectedEntityActionBarTransformButton />
         <EntityListSelectedEntityActionBarSaveToAssetsButton />

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

@@ -0,0 +1,37 @@
+import { IconButton } from '@invoke-ai/ui-library';
+import { useAppSelector } from 'app/store/storeHooks';
+import { useEntitySegmentAnything } from 'features/controlLayers/hooks/useEntitySegmentAnything';
+import { selectSelectedEntityIdentifier } from 'features/controlLayers/store/selectors';
+import { isSegmentableEntityIdentifier } from 'features/controlLayers/store/types';
+import { memo } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiMaskHappyBold } from 'react-icons/pi';
+
+export const EntityListSelectedEntityActionBarAutoMaskButton = memo(() => {
+  const { t } = useTranslation();
+  const selectedEntityIdentifier = useAppSelector(selectSelectedEntityIdentifier);
+  const segment = useEntitySegmentAnything(selectedEntityIdentifier);
+
+  if (!selectedEntityIdentifier) {
+    return null;
+  }
+
+  if (!isSegmentableEntityIdentifier(selectedEntityIdentifier)) {
+    return null;
+  }
+
+  return (
+    <IconButton
+      onClick={segment.start}
+      isDisabled={segment.isDisabled}
+      size="sm"
+      variant="link"
+      alignSelf="stretch"
+      aria-label={t('controlLayers.segment.autoMask')}
+      tooltip={t('controlLayers.segment.autoMask')}
+      icon={<PiMaskHappyBold />}
+    />
+  );
+});
+
+EntityListSelectedEntityActionBarAutoMaskButton.displayName = 'EntityListSelectedEntityActionBarAutoMaskButton';

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

@@ -25,8 +25,8 @@ const MenuContent = () => {
   return (
     <CanvasManagerProviderGate>
       <MenuList>
-        <CanvasContextMenuGlobalMenuItems />
         <CanvasContextMenuSelectedEntityMenuItems />
+        <CanvasContextMenuGlobalMenuItems />
       </MenuList>
     </CanvasManagerProviderGate>
   );

invokeai/frontend/web/src/features/controlLayers/components/ControlLayer/ControlLayerMenuItems.tsx

@@ -1,7 +1,6 @@
 import { MenuDivider } from '@invoke-ai/ui-library';
 import { IconMenuItemGroup } from 'common/components/IconMenuItem';
 import { CanvasEntityMenuItemsArrange } from 'features/controlLayers/components/common/CanvasEntityMenuItemsArrange';
-import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
 import { CanvasEntityMenuItemsCropToBbox } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCropToBbox';
 import { CanvasEntityMenuItemsDelete } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDelete';
 import { CanvasEntityMenuItemsDuplicate } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDuplicate';
@@ -9,7 +8,8 @@ import { CanvasEntityMenuItemsFilter } from 'features/controlLayers/components/c
 import { CanvasEntityMenuItemsSave } from 'features/controlLayers/components/common/CanvasEntityMenuItemsSave';
 import { CanvasEntityMenuItemsSegment } from 'features/controlLayers/components/common/CanvasEntityMenuItemsSegment';
 import { CanvasEntityMenuItemsTransform } from 'features/controlLayers/components/common/CanvasEntityMenuItemsTransform';
-import { ControlLayerMenuItemsConvertControlToRaster } from 'features/controlLayers/components/ControlLayer/ControlLayerMenuItemsConvertControlToRaster';
+import { ControlLayerMenuItemsConvertToSubMenu } from 'features/controlLayers/components/ControlLayer/ControlLayerMenuItemsConvertToSubMenu';
+import { ControlLayerMenuItemsCopyToSubMenu } from 'features/controlLayers/components/ControlLayer/ControlLayerMenuItemsCopyToSubMenu';
 import { ControlLayerMenuItemsTransparencyEffect } from 'features/controlLayers/components/ControlLayer/ControlLayerMenuItemsTransparencyEffect';
 import { memo } from 'react';
 
@@ -25,12 +25,13 @@ export const ControlLayerMenuItems = memo(() => {
       <CanvasEntityMenuItemsTransform />
       <CanvasEntityMenuItemsFilter />
       <CanvasEntityMenuItemsSegment />
-      <ControlLayerMenuItemsConvertControlToRaster />
       <ControlLayerMenuItemsTransparencyEffect />
       <MenuDivider />
       <CanvasEntityMenuItemsCropToBbox />
-      <CanvasEntityMenuItemsCopyToClipboard />
       <CanvasEntityMenuItemsSave />
+      <MenuDivider />
+      <ControlLayerMenuItemsConvertToSubMenu />
+      <ControlLayerMenuItemsCopyToSubMenu />
     </>
   );
 });

invokeai/frontend/web/src/features/controlLayers/components/ControlLayer/ControlLayerMenuItemsConvertControlToRaster.tsx

@@ -1,27 +0,0 @@
-import { MenuItem } from '@invoke-ai/ui-library';
-import { useAppDispatch } from 'app/store/storeHooks';
-import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
-import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
-import { controlLayerConvertedToRasterLayer } from 'features/controlLayers/store/canvasSlice';
-import { memo, useCallback } from 'react';
-import { useTranslation } from 'react-i18next';
-import { PiLightningBold } from 'react-icons/pi';
-
-export const ControlLayerMenuItemsConvertControlToRaster = memo(() => {
-  const { t } = useTranslation();
-  const dispatch = useAppDispatch();
-  const entityIdentifier = useEntityIdentifierContext('control_layer');
-  const isInteractable = useIsEntityInteractable(entityIdentifier);
-
-  const convertControlLayerToRasterLayer = useCallback(() => {
-    dispatch(controlLayerConvertedToRasterLayer({ entityIdentifier }));
-  }, [dispatch, entityIdentifier]);
-
-  return (
-    <MenuItem onClick={convertControlLayerToRasterLayer} icon={<PiLightningBold />} isDisabled={!isInteractable}>
-      {t('controlLayers.convertToRasterLayer')}
-    </MenuItem>
-  );
-});
-
-ControlLayerMenuItemsConvertControlToRaster.displayName = 'ControlLayerMenuItemsConvertControlToRaster';

invokeai/frontend/web/src/features/controlLayers/components/ControlLayer/ControlLayerMenuItemsConvertToSubMenu.tsx

@@ -0,0 +1,56 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import {
+  controlLayerConvertedToInpaintMask,
+  controlLayerConvertedToRasterLayer,
+  controlLayerConvertedToRegionalGuidance,
+} from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiSwapBold } from 'react-icons/pi';
+
+export const ControlLayerMenuItemsConvertToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('control_layer');
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const convertToInpaintMask = useCallback(() => {
+    dispatch(controlLayerConvertedToInpaintMask({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  const convertToRegionalGuidance = useCallback(() => {
+    dispatch(controlLayerConvertedToRegionalGuidance({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  const convertToRasterLayer = useCallback(() => {
+    dispatch(controlLayerConvertedToRasterLayer({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiSwapBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.convertControlLayerTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <MenuItem onClick={convertToInpaintMask} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.inpaintMask')}
+          </MenuItem>
+          <MenuItem onClick={convertToRegionalGuidance} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.regionalGuidance')}
+          </MenuItem>
+          <MenuItem onClick={convertToRasterLayer} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.rasterLayer')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+ControlLayerMenuItemsConvertToSubMenu.displayName = 'ControlLayerMenuItemsConvertToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/ControlLayer/ControlLayerMenuItemsCopyToSubMenu.tsx

@@ -0,0 +1,58 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import {
+  controlLayerConvertedToInpaintMask,
+  controlLayerConvertedToRasterLayer,
+  controlLayerConvertedToRegionalGuidance,
+} from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiCopyBold } from 'react-icons/pi';
+
+export const ControlLayerMenuItemsCopyToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('control_layer');
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const copyToInpaintMask = useCallback(() => {
+    dispatch(controlLayerConvertedToInpaintMask({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  const copyToRegionalGuidance = useCallback(() => {
+    dispatch(controlLayerConvertedToRegionalGuidance({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  const copyToRasterLayer = useCallback(() => {
+    dispatch(controlLayerConvertedToRasterLayer({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiCopyBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.copyControlLayerTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <CanvasEntityMenuItemsCopyToClipboard />
+          <MenuItem onClick={copyToInpaintMask} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newInpaintMask')}
+          </MenuItem>
+          <MenuItem onClick={copyToRegionalGuidance} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newRegionalGuidance')}
+          </MenuItem>
+          <MenuItem onClick={copyToRasterLayer} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newRasterLayer')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+ControlLayerMenuItemsCopyToSubMenu.displayName = 'ControlLayerMenuItemsCopyToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/IPAdapter/IPAdapterMenuItemPullBbox.tsx

@@ -0,0 +1,22 @@
+import { MenuItem } from '@invoke-ai/ui-library';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { usePullBboxIntoGlobalReferenceImage } from 'features/controlLayers/hooks/saveCanvasHooks';
+import { useCanvasIsBusy } from 'features/controlLayers/hooks/useCanvasIsBusy';
+import { memo } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiBoundingBoxBold } from 'react-icons/pi';
+
+export const IPAdapterMenuItemPullBbox = memo(() => {
+  const { t } = useTranslation();
+  const entityIdentifier = useEntityIdentifierContext('reference_image');
+  const pullBboxIntoIPAdapter = usePullBboxIntoGlobalReferenceImage(entityIdentifier);
+  const isBusy = useCanvasIsBusy();
+
+  return (
+    <MenuItem onClick={pullBboxIntoIPAdapter} icon={<PiBoundingBoxBold />} isDisabled={isBusy}>
+      {t('controlLayers.pullBboxIntoReferenceImage')}
+    </MenuItem>
+  );
+});
+
+IPAdapterMenuItemPullBbox.displayName = 'IPAdapterMenuItemPullBbox';

invokeai/frontend/web/src/features/controlLayers/components/IPAdapter/IPAdapterMenuItems.tsx

@@ -1,16 +1,22 @@
+import { MenuDivider } from '@invoke-ai/ui-library';
 import { IconMenuItemGroup } from 'common/components/IconMenuItem';
 import { CanvasEntityMenuItemsArrange } from 'features/controlLayers/components/common/CanvasEntityMenuItemsArrange';
 import { CanvasEntityMenuItemsDelete } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDelete';
 import { CanvasEntityMenuItemsDuplicate } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDuplicate';
+import { IPAdapterMenuItemPullBbox } from 'features/controlLayers/components/IPAdapter/IPAdapterMenuItemPullBbox';
 import { memo } from 'react';
 
 export const IPAdapterMenuItems = memo(() => {
   return (
-    <IconMenuItemGroup>
-      <CanvasEntityMenuItemsArrange />
-      <CanvasEntityMenuItemsDuplicate />
-      <CanvasEntityMenuItemsDelete asIcon />
-    </IconMenuItemGroup>
+    <>
+      <IconMenuItemGroup>
+        <CanvasEntityMenuItemsArrange />
+        <CanvasEntityMenuItemsDuplicate />
+        <CanvasEntityMenuItemsDelete asIcon />
+      </IconMenuItemGroup>
+      <MenuDivider />
+      <IPAdapterMenuItemPullBbox />
+    </>
   );
 });
 

invokeai/frontend/web/src/features/controlLayers/components/InpaintMask/InpaintMaskMenuItems.tsx

@@ -5,6 +5,8 @@ import { CanvasEntityMenuItemsCropToBbox } from 'features/controlLayers/componen
 import { CanvasEntityMenuItemsDelete } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDelete';
 import { CanvasEntityMenuItemsDuplicate } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDuplicate';
 import { CanvasEntityMenuItemsTransform } from 'features/controlLayers/components/common/CanvasEntityMenuItemsTransform';
+import { InpaintMaskMenuItemsConvertToSubMenu } from 'features/controlLayers/components/InpaintMask/InpaintMaskMenuItemsConvertToSubMenu';
+import { InpaintMaskMenuItemsCopyToSubMenu } from 'features/controlLayers/components/InpaintMask/InpaintMaskMenuItemsCopyToSubMenu';
 import { memo } from 'react';
 
 export const InpaintMaskMenuItems = memo(() => {
@@ -19,6 +21,9 @@ export const InpaintMaskMenuItems = memo(() => {
       <CanvasEntityMenuItemsTransform />
       <MenuDivider />
       <CanvasEntityMenuItemsCropToBbox />
+      <MenuDivider />
+      <InpaintMaskMenuItemsConvertToSubMenu />
+      <InpaintMaskMenuItemsCopyToSubMenu />
     </>
   );
 });

invokeai/frontend/web/src/features/controlLayers/components/InpaintMask/InpaintMaskMenuItemsConvertToSubMenu.tsx

@@ -0,0 +1,38 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import { inpaintMaskConvertedToRegionalGuidance } from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiSwapBold } from 'react-icons/pi';
+
+export const InpaintMaskMenuItemsConvertToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('inpaint_mask');
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const convertToRegionalGuidance = useCallback(() => {
+    dispatch(inpaintMaskConvertedToRegionalGuidance({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiSwapBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.convertInpaintMaskTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <MenuItem onClick={convertToRegionalGuidance} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.regionalGuidance')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+InpaintMaskMenuItemsConvertToSubMenu.displayName = 'InpaintMaskMenuItemsConvertToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/InpaintMask/InpaintMaskMenuItemsCopyToSubMenu.tsx

@@ -0,0 +1,40 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import { inpaintMaskConvertedToRegionalGuidance } from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiCopyBold } from 'react-icons/pi';
+
+export const InpaintMaskMenuItemsCopyToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('inpaint_mask');
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const copyToRegionalGuidance = useCallback(() => {
+    dispatch(inpaintMaskConvertedToRegionalGuidance({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiCopyBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.copyInpaintMaskTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <CanvasEntityMenuItemsCopyToClipboard />
+          <MenuItem onClick={copyToRegionalGuidance} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newRegionalGuidance')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+InpaintMaskMenuItemsCopyToSubMenu.displayName = 'InpaintMaskMenuItemsCopyToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/RasterLayer/RasterLayerMenuItems.tsx

@@ -1,7 +1,6 @@
 import { MenuDivider } from '@invoke-ai/ui-library';
 import { IconMenuItemGroup } from 'common/components/IconMenuItem';
 import { CanvasEntityMenuItemsArrange } from 'features/controlLayers/components/common/CanvasEntityMenuItemsArrange';
-import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
 import { CanvasEntityMenuItemsCropToBbox } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCropToBbox';
 import { CanvasEntityMenuItemsDelete } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDelete';
 import { CanvasEntityMenuItemsDuplicate } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDuplicate';
@@ -9,7 +8,8 @@ import { CanvasEntityMenuItemsFilter } from 'features/controlLayers/components/c
 import { CanvasEntityMenuItemsSave } from 'features/controlLayers/components/common/CanvasEntityMenuItemsSave';
 import { CanvasEntityMenuItemsSegment } from 'features/controlLayers/components/common/CanvasEntityMenuItemsSegment';
 import { CanvasEntityMenuItemsTransform } from 'features/controlLayers/components/common/CanvasEntityMenuItemsTransform';
-import { RasterLayerMenuItemsConvertRasterToControl } from 'features/controlLayers/components/RasterLayer/RasterLayerMenuItemsConvertRasterToControl';
+import { RasterLayerMenuItemsConvertToSubMenu } from 'features/controlLayers/components/RasterLayer/RasterLayerMenuItemsConvertToSubMenu';
+import { RasterLayerMenuItemsCopyToSubMenu } from 'features/controlLayers/components/RasterLayer/RasterLayerMenuItemsCopyToSubMenu';
 import { memo } from 'react';
 
 export const RasterLayerMenuItems = memo(() => {
@@ -24,11 +24,12 @@ export const RasterLayerMenuItems = memo(() => {
       <CanvasEntityMenuItemsTransform />
       <CanvasEntityMenuItemsFilter />
       <CanvasEntityMenuItemsSegment />
-      <RasterLayerMenuItemsConvertRasterToControl />
       <MenuDivider />
       <CanvasEntityMenuItemsCropToBbox />
-      <CanvasEntityMenuItemsCopyToClipboard />
       <CanvasEntityMenuItemsSave />
+      <MenuDivider />
+      <RasterLayerMenuItemsConvertToSubMenu />
+      <RasterLayerMenuItemsCopyToSubMenu />
     </>
   );
 });

invokeai/frontend/web/src/features/controlLayers/components/RasterLayer/RasterLayerMenuItemsConvertRasterToControl.tsx

@@ -1,36 +0,0 @@
-import { MenuItem } from '@invoke-ai/ui-library';
-import { useAppDispatch, useAppSelector } from 'app/store/storeHooks';
-import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
-import { selectDefaultControlAdapter } from 'features/controlLayers/hooks/addLayerHooks';
-import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
-import { rasterLayerConvertedToControlLayer } from 'features/controlLayers/store/canvasSlice';
-import { memo, useCallback } from 'react';
-import { useTranslation } from 'react-i18next';
-import { PiLightningBold } from 'react-icons/pi';
-
-export const RasterLayerMenuItemsConvertRasterToControl = memo(() => {
-  const { t } = useTranslation();
-  const dispatch = useAppDispatch();
-  const entityIdentifier = useEntityIdentifierContext('raster_layer');
-  const defaultControlAdapter = useAppSelector(selectDefaultControlAdapter);
-  const isInteractable = useIsEntityInteractable(entityIdentifier);
-
-  const onClick = useCallback(() => {
-    dispatch(
-      rasterLayerConvertedToControlLayer({
-        entityIdentifier,
-        overrides: {
-          controlAdapter: defaultControlAdapter,
-        },
-      })
-    );
-  }, [defaultControlAdapter, dispatch, entityIdentifier]);
-
-  return (
-    <MenuItem onClick={onClick} icon={<PiLightningBold />} isDisabled={!isInteractable}>
-      {t('controlLayers.convertToControlLayer')}
-    </MenuItem>
-  );
-});
-
-RasterLayerMenuItemsConvertRasterToControl.displayName = 'RasterLayerMenuItemsConvertRasterToControl';

invokeai/frontend/web/src/features/controlLayers/components/RasterLayer/RasterLayerMenuItemsConvertToSubMenu.tsx

@@ -0,0 +1,65 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch, useAppSelector } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { selectDefaultControlAdapter } from 'features/controlLayers/hooks/addLayerHooks';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import {
+  rasterLayerConvertedToControlLayer,
+  rasterLayerConvertedToInpaintMask,
+  rasterLayerConvertedToRegionalGuidance,
+} from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiSwapBold } from 'react-icons/pi';
+
+export const RasterLayerMenuItemsConvertToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('raster_layer');
+  const defaultControlAdapter = useAppSelector(selectDefaultControlAdapter);
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const convertToInpaintMask = useCallback(() => {
+    dispatch(rasterLayerConvertedToInpaintMask({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  const convertToRegionalGuidance = useCallback(() => {
+    dispatch(rasterLayerConvertedToRegionalGuidance({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  const convertToControlLayer = useCallback(() => {
+    dispatch(
+      rasterLayerConvertedToControlLayer({
+        entityIdentifier,
+        replace: true,
+        overrides: { controlAdapter: defaultControlAdapter },
+      })
+    );
+  }, [defaultControlAdapter, dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiSwapBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.convertRasterLayerTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <MenuItem onClick={convertToInpaintMask} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.inpaintMask')}
+          </MenuItem>
+          <MenuItem onClick={convertToRegionalGuidance} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.regionalGuidance')}
+          </MenuItem>
+          <MenuItem onClick={convertToControlLayer} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.controlLayer')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+RasterLayerMenuItemsConvertToSubMenu.displayName = 'RasterLayerMenuItemsConvertToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/RasterLayer/RasterLayerMenuItemsCopyToSubMenu.tsx

@@ -0,0 +1,66 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch, useAppSelector } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { selectDefaultControlAdapter } from 'features/controlLayers/hooks/addLayerHooks';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import {
+  rasterLayerConvertedToControlLayer,
+  rasterLayerConvertedToInpaintMask,
+  rasterLayerConvertedToRegionalGuidance,
+} from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiCopyBold } from 'react-icons/pi';
+
+export const RasterLayerMenuItemsCopyToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('raster_layer');
+  const defaultControlAdapter = useAppSelector(selectDefaultControlAdapter);
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const copyToInpaintMask = useCallback(() => {
+    dispatch(rasterLayerConvertedToInpaintMask({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  const copyToRegionalGuidance = useCallback(() => {
+    dispatch(rasterLayerConvertedToRegionalGuidance({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  const copyToControlLayer = useCallback(() => {
+    dispatch(
+      rasterLayerConvertedToControlLayer({
+        entityIdentifier,
+        overrides: { controlAdapter: defaultControlAdapter },
+      })
+    );
+  }, [defaultControlAdapter, dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiCopyBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.copyRasterLayerTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <CanvasEntityMenuItemsCopyToClipboard />
+          <MenuItem onClick={copyToInpaintMask} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newInpaintMask')}
+          </MenuItem>
+          <MenuItem onClick={copyToRegionalGuidance} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newRegionalGuidance')}
+          </MenuItem>
+          <MenuItem onClick={copyToControlLayer} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newControlLayer')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+RasterLayerMenuItemsCopyToSubMenu.displayName = 'RasterLayerMenuItemsCopyToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItems.tsx

@@ -1,4 +1,5 @@
-import { Flex, MenuDivider } from '@invoke-ai/ui-library';
+import { MenuDivider } from '@invoke-ai/ui-library';
+import { IconMenuItemGroup } from 'common/components/IconMenuItem';
 import { CanvasEntityMenuItemsArrange } from 'features/controlLayers/components/common/CanvasEntityMenuItemsArrange';
 import { CanvasEntityMenuItemsCropToBbox } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCropToBbox';
 import { CanvasEntityMenuItemsDelete } from 'features/controlLayers/components/common/CanvasEntityMenuItemsDelete';
@@ -6,16 +7,18 @@ import { CanvasEntityMenuItemsDuplicate } from 'features/controlLayers/component
 import { CanvasEntityMenuItemsTransform } from 'features/controlLayers/components/common/CanvasEntityMenuItemsTransform';
 import { RegionalGuidanceMenuItemsAddPromptsAndIPAdapter } from 'features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItemsAddPromptsAndIPAdapter';
 import { RegionalGuidanceMenuItemsAutoNegative } from 'features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItemsAutoNegative';
+import { RegionalGuidanceMenuItemsConvertToSubMenu } from 'features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItemsConvertToSubMenu';
+import { RegionalGuidanceMenuItemsCopyToSubMenu } from 'features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItemsCopyToSubMenu';
 import { memo } from 'react';
 
 export const RegionalGuidanceMenuItems = memo(() => {
   return (
     <>
-      <Flex gap={2}>
+      <IconMenuItemGroup>
         <CanvasEntityMenuItemsArrange />
         <CanvasEntityMenuItemsDuplicate />
         <CanvasEntityMenuItemsDelete asIcon />
-      </Flex>
+      </IconMenuItemGroup>
       <MenuDivider />
       <RegionalGuidanceMenuItemsAddPromptsAndIPAdapter />
       <MenuDivider />
@@ -23,6 +26,9 @@ export const RegionalGuidanceMenuItems = memo(() => {
       <RegionalGuidanceMenuItemsAutoNegative />
       <MenuDivider />
       <CanvasEntityMenuItemsCropToBbox />
+      <MenuDivider />
+      <RegionalGuidanceMenuItemsConvertToSubMenu />
+      <RegionalGuidanceMenuItemsCopyToSubMenu />
     </>
   );
 });

invokeai/frontend/web/src/features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItemsConvertToSubMenu.tsx

@@ -0,0 +1,38 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import { rgConvertedToInpaintMask } from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiSwapBold } from 'react-icons/pi';
+
+export const RegionalGuidanceMenuItemsConvertToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('regional_guidance');
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const convertToInpaintMask = useCallback(() => {
+    dispatch(rgConvertedToInpaintMask({ entityIdentifier, replace: true }));
+  }, [dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiSwapBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.convertRegionalGuidanceTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <MenuItem onClick={convertToInpaintMask} icon={<PiSwapBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.inpaintMask')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+RegionalGuidanceMenuItemsConvertToSubMenu.displayName = 'RegionalGuidanceMenuItemsConvertToSubMenu';

invokeai/frontend/web/src/features/controlLayers/components/RegionalGuidance/RegionalGuidanceMenuItemsCopyToSubMenu.tsx

@@ -0,0 +1,40 @@
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { useAppDispatch } from 'app/store/storeHooks';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
+import { CanvasEntityMenuItemsCopyToClipboard } from 'features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard';
+import { useEntityIdentifierContext } from 'features/controlLayers/contexts/EntityIdentifierContext';
+import { useIsEntityInteractable } from 'features/controlLayers/hooks/useEntityIsInteractable';
+import { rgConvertedToInpaintMask } from 'features/controlLayers/store/canvasSlice';
+import { memo, useCallback } from 'react';
+import { useTranslation } from 'react-i18next';
+import { PiCopyBold } from 'react-icons/pi';
+
+export const RegionalGuidanceMenuItemsCopyToSubMenu = memo(() => {
+  const { t } = useTranslation();
+  const subMenu = useSubMenu();
+  const dispatch = useAppDispatch();
+  const entityIdentifier = useEntityIdentifierContext('regional_guidance');
+  const isInteractable = useIsEntityInteractable(entityIdentifier);
+
+  const copyToInpaintMask = useCallback(() => {
+    dispatch(rgConvertedToInpaintMask({ entityIdentifier }));
+  }, [dispatch, entityIdentifier]);
+
+  return (
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiCopyBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label={t('controlLayers.copyRegionalGuidanceTo')} />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <CanvasEntityMenuItemsCopyToClipboard />
+          <MenuItem onClick={copyToInpaintMask} icon={<PiCopyBold />} isDisabled={!isInteractable}>
+            {t('controlLayers.newInpaintMask')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
+  );
+});
+
+RegionalGuidanceMenuItemsCopyToSubMenu.displayName = 'RegionalGuidanceMenuItemsCopyToSubMenu';

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

@@ -1,4 +1,14 @@
-import { Button, ButtonGroup, Flex, Heading, Spacer } from '@invoke-ai/ui-library';
+import {
+  Button,
+  ButtonGroup,
+  Flex,
+  Heading,
+  Menu,
+  MenuButton,
+  MenuItem,
+  MenuList,
+  Spacer,
+} from '@invoke-ai/ui-library';
 import { useStore } from '@nanostores/react';
 import { useAppSelector } from 'app/store/storeHooks';
 import { useFocusRegion, useIsRegionFocused } from 'common/hooks/focus';
@@ -10,9 +20,9 @@ import type { CanvasEntityAdapterControlLayer } from 'features/controlLayers/kon
 import type { CanvasEntityAdapterRasterLayer } from 'features/controlLayers/konva/CanvasEntity/CanvasEntityAdapterRasterLayer';
 import { selectAutoProcess } from 'features/controlLayers/store/canvasSettingsSlice';
 import { useRegisteredHotkeys } from 'features/system/components/HotkeysModal/useHotkeyData';
-import { memo, useRef } from 'react';
+import { memo, useCallback, useRef } from 'react';
 import { useTranslation } from 'react-i18next';
-import { PiArrowsCounterClockwiseBold, PiCheckBold, PiStarBold, PiXBold } from 'react-icons/pi';
+import { PiArrowsCounterClockwiseBold, PiFloppyDiskBold, PiStarBold, PiXBold } from 'react-icons/pi';
 
 const SegmentAnythingContent = memo(
   ({ adapter }: { adapter: CanvasEntityAdapterRasterLayer | CanvasEntityAdapterControlLayer }) => {
@@ -22,8 +32,25 @@ const SegmentAnythingContent = memo(
     const isCanvasFocused = useIsRegionFocused('canvas');
     const isProcessing = useStore(adapter.segmentAnything.$isProcessing);
     const hasPoints = useStore(adapter.segmentAnything.$hasPoints);
+    const hasImageState = useStore(adapter.segmentAnything.$hasImageState);
     const autoProcess = useAppSelector(selectAutoProcess);
 
+    const saveAsInpaintMask = useCallback(() => {
+      adapter.segmentAnything.saveAs('inpaint_mask');
+    }, [adapter.segmentAnything]);
+
+    const saveAsRegionalGuidance = useCallback(() => {
+      adapter.segmentAnything.saveAs('regional_guidance');
+    }, [adapter.segmentAnything]);
+
+    const saveAsRasterLayer = useCallback(() => {
+      adapter.segmentAnything.saveAs('raster_layer');
+    }, [adapter.segmentAnything]);
+
+    const saveAsControlLayer = useCallback(() => {
+      adapter.segmentAnything.saveAs('control_layer');
+    }, [adapter.segmentAnything]);
+
     useRegisteredHotkeys({
       id: 'applySegmentAnything',
       category: 'canvas',
@@ -86,15 +113,32 @@ const SegmentAnythingContent = memo(
           >
             {t('controlLayers.segment.reset')}
           </Button>
-          <Button
-            leftIcon={<PiCheckBold />}
-            onClick={adapter.segmentAnything.apply}
-            isLoading={isProcessing}
-            loadingText={t('controlLayers.segment.apply')}
-            variant="ghost"
-          >
-            {t('controlLayers.segment.apply')}
-          </Button>
+          <Menu>
+            <MenuButton
+              as={Button}
+              leftIcon={<PiFloppyDiskBold />}
+              isLoading={isProcessing}
+              loadingText={t('controlLayers.segment.saveAs')}
+              variant="ghost"
+              isDisabled={!hasImageState}
+            >
+              {t('controlLayers.segment.saveAs')}
+            </MenuButton>
+            <MenuList>
+              <MenuItem isDisabled={!hasImageState} onClick={saveAsInpaintMask}>
+                {t('controlLayers.inpaintMask')}
+              </MenuItem>
+              <MenuItem isDisabled={!hasImageState} onClick={saveAsRegionalGuidance}>
+                {t('controlLayers.regionalGuidance')}
+              </MenuItem>
+              <MenuItem isDisabled={!hasImageState} onClick={saveAsControlLayer}>
+                {t('controlLayers.controlLayer')}
+              </MenuItem>
+              <MenuItem isDisabled={!hasImageState} onClick={saveAsRasterLayer}>
+                {t('controlLayers.rasterLayer')}
+              </MenuItem>
+            </MenuList>
+          </Menu>
           <Button
             leftIcon={<PiXBold />}
             onClick={adapter.segmentAnything.cancel}

invokeai/frontend/web/src/features/controlLayers/components/SegmentAnything/SegmentAnythingPointType.tsx

@@ -26,13 +26,10 @@ export const SegmentAnythingPointType = memo(
         <RadioGroup value={pointType} onChange={onChange} w="full" size="md">
           <Flex alignItems="center" w="full" gap={4} fontWeight="semibold" color="base.300">
             <Radio value="foreground">
-              <Text>{t('controlLayers.segment.foreground')}</Text>
+              <Text>{t('controlLayers.segment.include')}</Text>
             </Radio>
             <Radio value="background">
-              <Text>{t('controlLayers.segment.background')}</Text>
-            </Radio>
-            <Radio value="neutral">
-              <Text>{t('controlLayers.segment.neutral')}</Text>
+              <Text>{t('controlLayers.segment.exclude')}</Text>
             </Radio>
           </Flex>
         </RadioGroup>

invokeai/frontend/web/src/features/controlLayers/components/common/CanvasEntityGroupList.tsx

@@ -1,9 +1,11 @@
 import type { SystemStyleObject } from '@invoke-ai/ui-library';
 import { Button, Collapse, Flex, Icon, Spacer, Text } from '@invoke-ai/ui-library';
+import { InformationalPopover } from 'common/components/InformationalPopover/InformationalPopover';
 import { useBoolean } from 'common/hooks/useBoolean';
 import { CanvasEntityAddOfTypeButton } from 'features/controlLayers/components/common/CanvasEntityAddOfTypeButton';
 import { CanvasEntityMergeVisibleButton } from 'features/controlLayers/components/common/CanvasEntityMergeVisibleButton';
 import { CanvasEntityTypeIsHiddenToggle } from 'features/controlLayers/components/common/CanvasEntityTypeIsHiddenToggle';
+import { useEntityTypeInformationalPopover } from 'features/controlLayers/hooks/useEntityTypeInformationalPopover';
 import { useEntityTypeTitle } from 'features/controlLayers/hooks/useEntityTypeTitle';
 import type { CanvasEntityIdentifier } from 'features/controlLayers/store/types';
 import type { PropsWithChildren } from 'react';
@@ -21,6 +23,7 @@ const _hover: SystemStyleObject = {
 
 export const CanvasEntityGroupList = memo(({ isSelected, type, children }: Props) => {
   const title = useEntityTypeTitle(type);
+  const informationalPopoverFeature = useEntityTypeInformationalPopover(type);
   const collapse = useBoolean(true);
   const canMergeVisible = useMemo(() => type === 'raster_layer' || type === 'inpaint_mask', [type]);
   const canHideAll = useMemo(() => type !== 'reference_image', [type]);
@@ -47,15 +50,30 @@ export const CanvasEntityGroupList = memo(({ isSelected, type, children }: Props
             transitionProperty="common"
             transitionDuration="fast"
           />
-          <Text
-            fontWeight="semibold"
-            color={isSelected ? 'base.200' : 'base.500'}
-            userSelect="none"
-            transitionProperty="common"
-            transitionDuration="fast"
-          >
-            {title}
-          </Text>
+          {informationalPopoverFeature ? (
+            <InformationalPopover feature={informationalPopoverFeature}>
+              <Text
+                fontWeight="semibold"
+                color={isSelected ? 'base.200' : 'base.500'}
+                userSelect="none"
+                transitionProperty="common"
+                transitionDuration="fast"
+              >
+                {title}
+              </Text>
+            </InformationalPopover>
+          ) : (
+            <Text
+              fontWeight="semibold"
+              color={isSelected ? 'base.200' : 'base.500'}
+              userSelect="none"
+              transitionProperty="common"
+              transitionDuration="fast"
+            >
+              {title}
+            </Text>
+          )}
+
           <Spacer />
         </Flex>
         {canMergeVisible && <CanvasEntityMergeVisibleButton type={type} />}

invokeai/frontend/web/src/features/controlLayers/components/common/CanvasEntityMenuItemsCopyToClipboard.tsx

@@ -20,7 +20,7 @@ export const CanvasEntityMenuItemsCopyToClipboard = memo(() => {
 
   return (
     <MenuItem onClick={onClick} icon={<PiCopyBold />} isDisabled={!isInteractable}>
-      {t('controlLayers.copyToClipboard')}
+      {t('common.clipboard')}
     </MenuItem>
   );
 });

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

@@ -5,11 +5,13 @@ import { useEntityAdapterSafe } from 'features/controlLayers/contexts/EntityAdap
 import { useCanvasIsBusy } from 'features/controlLayers/hooks/useCanvasIsBusy';
 import type { CanvasEntityIdentifier } from 'features/controlLayers/store/types';
 import { isFilterableEntityIdentifier } from 'features/controlLayers/store/types';
+import { useImageViewer } from 'features/gallery/components/ImageViewer/useImageViewer';
 import { useCallback, useMemo } from 'react';
 
 export const useEntityFilter = (entityIdentifier: CanvasEntityIdentifier | null) => {
   const canvasManager = useCanvasManager();
   const adapter = useEntityAdapterSafe(entityIdentifier);
+  const imageViewer = useImageViewer();
   const isBusy = useCanvasIsBusy();
   const isInteractable = useStore(adapter?.$isInteractable ?? $false);
   const isEmpty = useStore(adapter?.$isEmpty ?? $false);
@@ -50,8 +52,9 @@ export const useEntityFilter = (entityIdentifier: CanvasEntityIdentifier | null)
     if (!adapter) {
       return;
     }
+    imageViewer.close();
     adapter.filterer.start();
-  }, [isDisabled, entityIdentifier, canvasManager]);
+  }, [isDisabled, entityIdentifier, canvasManager, imageViewer]);
 
   return { isDisabled, start } as const;
 };

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

@@ -5,11 +5,13 @@ import { useEntityAdapterSafe } from 'features/controlLayers/contexts/EntityAdap
 import { useCanvasIsBusy } from 'features/controlLayers/hooks/useCanvasIsBusy';
 import type { CanvasEntityIdentifier } from 'features/controlLayers/store/types';
 import { isSegmentableEntityIdentifier } from 'features/controlLayers/store/types';
+import { useImageViewer } from 'features/gallery/components/ImageViewer/useImageViewer';
 import { useCallback, useMemo } from 'react';
 
 export const useEntitySegmentAnything = (entityIdentifier: CanvasEntityIdentifier | null) => {
   const canvasManager = useCanvasManager();
   const adapter = useEntityAdapterSafe(entityIdentifier);
+  const imageViewer = useImageViewer();
   const isBusy = useCanvasIsBusy();
   const isInteractable = useStore(adapter?.$isInteractable ?? $false);
   const isEmpty = useStore(adapter?.$isEmpty ?? $false);
@@ -50,8 +52,9 @@ export const useEntitySegmentAnything = (entityIdentifier: CanvasEntityIdentifie
     if (!adapter) {
       return;
     }
+    imageViewer.close();
     adapter.segmentAnything.start();
-  }, [isDisabled, entityIdentifier, canvasManager]);
+  }, [isDisabled, entityIdentifier, canvasManager, imageViewer]);
 
   return { isDisabled, start } as const;
 };

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

@@ -5,11 +5,13 @@ import { useEntityAdapterSafe } from 'features/controlLayers/contexts/EntityAdap
 import { useCanvasIsBusy } from 'features/controlLayers/hooks/useCanvasIsBusy';
 import type { CanvasEntityIdentifier } from 'features/controlLayers/store/types';
 import { isTransformableEntityIdentifier } from 'features/controlLayers/store/types';
+import { useImageViewer } from 'features/gallery/components/ImageViewer/useImageViewer';
 import { useCallback, useMemo } from 'react';
 
 export const useEntityTransform = (entityIdentifier: CanvasEntityIdentifier | null) => {
   const canvasManager = useCanvasManager();
   const adapter = useEntityAdapterSafe(entityIdentifier);
+  const imageViewer = useImageViewer();
   const isBusy = useCanvasIsBusy();
   const isInteractable = useStore(adapter?.$isInteractable ?? $false);
   const isEmpty = useStore(adapter?.$isEmpty ?? $false);
@@ -67,10 +69,11 @@ export const useEntityTransform = (entityIdentifier: CanvasEntityIdentifier | nu
     if (!adapter) {
       return;
     }
+    imageViewer.close();
     await adapter.transformer.startTransform({ silent: true });
     adapter.transformer.fitToBboxContain();
     await adapter.transformer.applyTransform();
-  }, [canvasManager, entityIdentifier, isDisabled]);
+  }, [canvasManager, entityIdentifier, imageViewer, isDisabled]);
 
   return { isDisabled, start, fitToBbox } as const;
 };

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

@@ -0,0 +1,25 @@
+import type { Feature } from 'common/components/InformationalPopover/constants';
+import type { CanvasEntityIdentifier } from 'features/controlLayers/store/types';
+import { useMemo } from 'react';
+
+export const useEntityTypeInformationalPopover = (type: CanvasEntityIdentifier['type']): Feature | undefined => {
+  const feature = useMemo(() => {
+    switch (type) {
+      case 'control_layer':
+        return 'controlNet';
+      case 'inpaint_mask':
+        return 'inpainting';
+      case 'raster_layer':
+        return 'rasterLayer';
+      case 'regional_guidance':
+        return 'regionalGuidanceAndReferenceImage';
+      case 'reference_image':
+        return 'globalReferenceImage';
+
+      default:
+        return undefined;
+    }
+  }, [type]);
+
+  return feature;
+};

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

@@ -6,15 +6,22 @@ import type { CanvasEntityAdapterRasterLayer } from 'features/controlLayers/konv
 import type { CanvasManager } from 'features/controlLayers/konva/CanvasManager';
 import { CanvasModuleBase } from 'features/controlLayers/konva/CanvasModuleBase';
 import { CanvasObjectImage } from 'features/controlLayers/konva/CanvasObject/CanvasObjectImage';
-import { addCoords, getKonvaNodeDebugAttrs, getPrefixedId, offsetCoord } from 'features/controlLayers/konva/util';
+import {
+  addCoords,
+  getKonvaNodeDebugAttrs,
+  getPrefixedId,
+  offsetCoord,
+  roundCoord,
+} from 'features/controlLayers/konva/util';
 import { selectAutoProcess } from 'features/controlLayers/store/canvasSettingsSlice';
 import type {
+  CanvasEntityType,
   CanvasImageState,
   Coordinate,
   RgbaColor,
-  SAMPoint,
   SAMPointLabel,
   SAMPointLabelString,
+  SAMPointWithId,
 } from 'features/controlLayers/store/types';
 import { SAM_POINT_LABEL_NUMBER_TO_STRING } from 'features/controlLayers/store/types';
 import { imageDTOToImageObject } from 'features/controlLayers/store/util';
@@ -27,6 +34,9 @@ import { atom, computed } from 'nanostores';
 import type { Logger } from 'roarr';
 import { serializeError } from 'serialize-error';
 import type { ImageDTO } from 'services/api/types';
+import stableHash from 'stable-hash';
+import type { Equals } from 'tsafe';
+import { assert } from 'tsafe';
 
 type CanvasSegmentAnythingModuleConfig = {
   /**
@@ -70,7 +80,7 @@ const DEFAULT_CONFIG: CanvasSegmentAnythingModuleConfig = {
   SAM_POINT_FOREGROUND_COLOR: { r: 50, g: 255, b: 0, a: 1 }, // light green
   SAM_POINT_BACKGROUND_COLOR: { r: 255, g: 0, b: 50, a: 1 }, // red-ish
   SAM_POINT_NEUTRAL_COLOR: { r: 0, g: 225, b: 255, a: 1 }, // cyan
-  MASK_COLOR: { r: 0, g: 200, b: 200, a: 0.5 }, // cyan with 50% opacity
+  MASK_COLOR: { r: 0, g: 225, b: 255, a: 1 }, // cyan
   PROCESS_DEBOUNCE_MS: 1000,
 };
 
@@ -85,6 +95,7 @@ const DEFAULT_CONFIG: CanvasSegmentAnythingModuleConfig = {
 type SAMPointState = {
   id: string;
   label: SAMPointLabel;
+  coord: Coordinate;
   konva: {
     circle: Konva.Circle;
   };
@@ -113,9 +124,9 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
   $isSegmenting = atom<boolean>(false);
 
   /**
-   * Whether the current set of points has been processed.
+   * The hash of the last processed points. This is used to prevent re-processing the same points.
    */
-  $hasProcessed = atom<boolean>(false);
+  $lastProcessedHash = atom<string>('');
 
   /**
    * Whether the module is currently processing the points.
@@ -144,10 +155,15 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
   /**
    * The ephemeral image state of the processed image. Only used while segmenting.
    */
-  imageState: CanvasImageState | null = null;
+  $imageState = atom<CanvasImageState | null>(null);
 
   /**
-   * The current input points.
+   * Whether the module has an image state. This is a computed value based on $imageState.
+   */
+  $hasImageState = computed(this.$imageState, (imageState) => imageState !== null);
+
+  /**
+   * The current input points. A listener is added to this atom to process the points when they change.
    */
   $points = atom<SAMPointState[]>([]);
 
@@ -187,6 +203,10 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
      * It's rendered with a globalCompositeOperation of 'source-atop' to preview the mask as a semi-transparent overlay.
      */
     compositingRect: Konva.Rect;
+    /**
+     * A tween for pulsing the mask group's opacity.
+     */
+    maskTween: Konva.Tween | null;
   };
 
   KONVA_CIRCLE_NAME = `${this.type}:circle`;
@@ -209,7 +229,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     this.konva = {
       group: new Konva.Group({ name: this.KONVA_GROUP_NAME }),
       pointGroup: new Konva.Group({ name: this.KONVA_POINT_GROUP_NAME }),
-      maskGroup: new Konva.Group({ name: this.KONVA_MASK_GROUP_NAME }),
+      maskGroup: new Konva.Group({ name: this.KONVA_MASK_GROUP_NAME, opacity: 0.6 }),
       compositingRect: new Konva.Rect({
         name: this.KONVA_COMPOSITING_RECT_NAME,
         fill: rgbaColorToString(this.config.MASK_COLOR),
@@ -219,6 +239,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
         perfectDrawEnabled: false,
         visible: false,
       }),
+      maskTween: null,
     };
 
     // Points should always be rendered above the mask group
@@ -250,10 +271,12 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
   createPoint(coord: Coordinate, label: SAMPointLabel): SAMPointState {
     const id = getPrefixedId('sam_point');
 
+    const roundedCoord = roundCoord(coord);
+
     const circle = new Konva.Circle({
       name: this.KONVA_CIRCLE_NAME,
-      x: Math.round(coord.x),
-      y: Math.round(coord.y),
+      x: roundedCoord.x,
+      y: roundedCoord.y,
       radius: this.manager.stage.unscale(this.config.SAM_POINT_RADIUS), // We will scale this as the stage scale changes
       fill: rgbaColorToString(this.getSAMPointColor(label)),
       stroke: rgbaColorToString(this.config.SAM_POINT_BORDER_COLOR),
@@ -273,11 +296,12 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
       // This event should not bubble up to the parent, stage or any other nodes
       e.cancelBubble = true;
       circle.destroy();
-      this.$points.set(this.$points.get().filter((point) => point.id !== id));
-      if (this.$points.get().length === 0) {
+
+      const newPoints = this.$points.get().filter((point) => point.id !== id);
+      if (newPoints.length === 0) {
         this.resetEphemeralState();
       } else {
-        this.$hasProcessed.set(false);
+        this.$points.set(newPoints);
       }
     });
 
@@ -286,25 +310,28 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     });
 
     circle.on('dragend', () => {
+      const roundedCoord = roundCoord(circle.position());
+
+      this.log.trace({ ...roundedCoord, label: SAM_POINT_LABEL_NUMBER_TO_STRING[label] }, 'Moved SAM point');
       this.$isDraggingPoint.set(false);
-      // Point has changed!
-      this.$hasProcessed.set(false);
-      this.$points.notify();
-      this.log.trace(
-        { x: Math.round(circle.x()), y: Math.round(circle.y()), label: SAM_POINT_LABEL_NUMBER_TO_STRING[label] },
-        'Moved SAM point'
-      );
+
+      const newPoints = this.$points.get().map((point) => {
+        if (point.id === id) {
+          return { ...point, coord: roundedCoord };
+        }
+        return point;
+      });
+
+      this.$points.set(newPoints);
     });
 
     this.konva.pointGroup.add(circle);
 
-    this.log.trace(
-      { x: Math.round(circle.x()), y: Math.round(circle.y()), label: SAM_POINT_LABEL_NUMBER_TO_STRING[label] },
-      'Created SAM point'
-    );
+    this.log.trace({ ...roundedCoord, label: SAM_POINT_LABEL_NUMBER_TO_STRING[label] }, 'Created SAM point');
 
     return {
       id,
+      coord: roundedCoord,
       label,
       konva: { circle },
     };
@@ -327,14 +354,14 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
   /**
    * Gets the SAM points in the format expected by the segment-anything API. The x and y values are rounded to integers.
    */
-  getSAMPoints = (): SAMPoint[] => {
-    const points: SAMPoint[] = [];
+  getSAMPoints = (): SAMPointWithId[] => {
+    const points: SAMPointWithId[] = [];
 
-    for (const { konva, label } of this.$points.get()) {
+    for (const { id, coord, label } of this.$points.get()) {
       points.push({
-        // Pull out and round the x and y values from Konva
-        x: Math.round(konva.circle.x()),
-        y: Math.round(konva.circle.y()),
+        id,
+        x: coord.x,
+        y: coord.y,
         label,
       });
     }
@@ -381,10 +408,8 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
 
     // Create a SAM point at the normalized position
     const point = this.createPoint(normalizedPoint, this.$pointType.get());
-    this.$points.set([...this.$points.get(), point]);
-
-    // Mark the module as having _not_ processed the points now that they have changed
-    this.$hasProcessed.set(false);
+    const newPoints = [...this.$points.get(), point];
+    this.$points.set(newPoints);
   };
 
   /**
@@ -421,6 +446,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
         if (points.length === 0) {
           return;
         }
+
         if (this.manager.stateApi.getSettings().autoProcess) {
           this.process();
         }
@@ -433,7 +459,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
         if (this.$points.get().length === 0) {
           return;
         }
-        if (autoProcess && !this.$hasProcessed.get()) {
+        if (autoProcess) {
           this.process();
         }
       })
@@ -500,6 +526,12 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
       return;
     }
 
+    const hash = stableHash(points);
+    if (hash === this.$lastProcessedHash.get()) {
+      this.log.trace('Already processed points');
+      return;
+    }
+
     this.$isProcessing.set(true);
 
     this.log.trace({ points }, 'Segmenting');
@@ -521,7 +553,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     this.abortController = controller;
 
     // Build the graph for segmenting the image, using the rasterized image DTO
-    const { graph, outputNodeId } = this.buildGraph(rasterizeResult.value);
+    const { graph, outputNodeId } = this.buildGraph(rasterizeResult.value, points);
 
     // Run the graph and get the segmented image output
     const segmentResult = await withResultAsync(() =>
@@ -548,21 +580,27 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     this.log.trace({ imageDTO: segmentResult.value }, 'Segmented');
 
     // Prepare the ephemeral image state
-    this.imageState = imageDTOToImageObject(segmentResult.value);
+    const imageState = imageDTOToImageObject(segmentResult.value);
+    this.$imageState.set(imageState);
 
     // Destroy any existing masked image and create a new one
     if (this.maskedImage) {
       this.maskedImage.destroy();
     }
-    this.maskedImage = new CanvasObjectImage(this.imageState, this);
+    if (this.konva.maskTween) {
+      this.konva.maskTween.destroy();
+      this.konva.maskTween = null;
+    }
+
+    this.maskedImage = new CanvasObjectImage(imageState, this);
 
     // Force update the masked image - after awaiting, the image will be rendered (in memory)
-    await this.maskedImage.update(this.imageState, true);
+    await this.maskedImage.update(imageState, true);
 
     // Update the compositing rect to match the image size
     this.konva.compositingRect.setAttrs({
-      width: this.imageState.image.width,
-      height: this.imageState.image.height,
+      width: imageState.image.width,
+      height: imageState.image.height,
       visible: true,
     });
 
@@ -574,12 +612,24 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     // Cache the group to ensure the mask is rendered correctly w/ opacity
     this.konva.maskGroup.cache();
 
+    // Create a pulsing tween
+    this.konva.maskTween = new Konva.Tween({
+      node: this.konva.maskGroup,
+      duration: 1,
+      opacity: 0.4, // oscillate between this value and pre-tween opacity
+      yoyo: true,
+      repeat: Infinity,
+      easing: Konva.Easings.EaseOut,
+    });
+
+    // Start the pulsing effect
+    this.konva.maskTween.play();
+
+    this.$lastProcessedHash.set(hash);
+
     // We are done processing (still segmenting though!)
     this.$isProcessing.set(false);
 
-    // The current points have been processed
-    this.$hasProcessed.set(true);
-
     // Clean up the abort controller as needed
     if (!this.abortController.signal.aborted) {
       this.abortController.abort();
@@ -596,11 +646,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
    * Applies the segmented image to the entity.
    */
   apply = () => {
-    if (!this.$hasProcessed.get()) {
-      this.log.error('Cannot apply unprocessed points');
-      return;
-    }
-    const imageState = this.imageState;
+    const imageState = this.$imageState.get();
     if (!imageState) {
       this.log.error('No image state to apply');
       return;
@@ -627,6 +673,55 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     this.teardown();
   };
 
+  /**
+   * Applies the segmented image to the entity.
+   */
+  saveAs = (type: Exclude<CanvasEntityType, 'reference_image'>) => {
+    const imageState = this.$imageState.get();
+    if (!imageState) {
+      this.log.error('No image state to save as');
+      return;
+    }
+    this.log.trace(`Saving as ${type}`);
+
+    // Clear the buffer - we are creating a new entity, so we don't want to keep the old one
+    this.parent.bufferRenderer.clearBuffer();
+
+    // Create the new entity with the masked image as its only object
+    const rect = this.parent.transformer.getRelativeRect();
+    const arg = {
+      overrides: {
+        objects: [imageState],
+        position: {
+          x: Math.round(rect.x),
+          y: Math.round(rect.y),
+        },
+      },
+      isSelected: true,
+    };
+
+    switch (type) {
+      case 'raster_layer':
+        this.manager.stateApi.addRasterLayer(arg);
+        break;
+      case 'control_layer':
+        this.manager.stateApi.addControlLayer(arg);
+        break;
+      case 'inpaint_mask':
+        this.manager.stateApi.addInpaintMask(arg);
+        break;
+      case 'regional_guidance':
+        this.manager.stateApi.addRegionalGuidance(arg);
+        break;
+      default:
+        assert<Equals<typeof type, never>>(false);
+    }
+
+    // Final cleanup and teardown, returning user to main canvas UI
+    this.resetEphemeralState();
+    this.teardown();
+  };
+
   /**
    * Resets the module (e.g. remove all points and the mask image).
    *
@@ -686,12 +781,16 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
     if (this.maskedImage) {
       this.maskedImage.destroy();
     }
+    if (this.konva.maskTween) {
+      this.konva.maskTween.destroy();
+      this.konva.maskTween = null;
+    }
 
     // Empty internal module state
     this.$points.set([]);
-    this.imageState = null;
+    this.$imageState.set(null);
     this.$pointType.set(1);
-    this.$hasProcessed.set(false);
+    this.$lastProcessedHash.set('');
     this.$isProcessing.set(false);
 
     // Reset non-ephemeral konva nodes
@@ -706,7 +805,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
   /**
    * Builds a graph for segmenting an image with the given image DTO.
    */
-  buildGraph = ({ image_name }: ImageDTO): { graph: Graph; outputNodeId: string } => {
+  buildGraph = ({ image_name }: ImageDTO, points: SAMPointWithId[]): { graph: Graph; outputNodeId: string } => {
     const graph = new Graph(getPrefixedId('canvas_segment_anything'));
 
     // TODO(psyche): When SAM2 is available in transformers, use it here
@@ -716,7 +815,7 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
       type: 'segment_anything',
       model: 'segment-anything-huge',
       image: { image_name },
-      point_lists: [{ points: this.getSAMPoints() }],
+      point_lists: [{ points: points.map(({ x, y, label }) => ({ x, y, label })) }],
       mask_filter: 'largest',
     });
 
@@ -759,11 +858,11 @@ export class CanvasSegmentAnythingModule extends CanvasModuleBase {
         label,
         circle: getKonvaNodeDebugAttrs(konva.circle),
       })),
-      imageState: deepClone(this.imageState),
+      imageState: deepClone(this.$imageState.get()),
       maskedImage: this.maskedImage?.repr(),
       config: deepClone(this.config),
       $isSegmenting: this.$isSegmenting.get(),
-      $hasProcessed: this.$hasProcessed.get(),
+      $lastProcessedHash: this.$lastProcessedHash.get(),
       $isProcessing: this.$isProcessing.get(),
       $pointType: this.$pointType.get(),
       $pointTypeString: this.$pointTypeString.get(),

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

@@ -17,12 +17,16 @@ import {
 } from 'features/controlLayers/store/canvasSettingsSlice';
 import {
   bboxChangedFromCanvas,
+  controlLayerAdded,
   entityBrushLineAdded,
   entityEraserLineAdded,
   entityMoved,
   entityRasterized,
   entityRectAdded,
   entityReset,
+  inpaintMaskAdded,
+  rasterLayerAdded,
+  rgAdded,
 } from 'features/controlLayers/store/canvasSlice';
 import { selectCanvasStagingAreaSlice } from 'features/controlLayers/store/canvasStagingAreaSlice';
 import {
@@ -51,6 +55,7 @@ import { getImageDTO } from 'services/api/endpoints/images';
 import { queueApi } from 'services/api/endpoints/queue';
 import type { BatchConfig, ImageDTO, S } from 'services/api/types';
 import { QueueError } from 'services/events/errors';
+import type { Param0 } from 'tsafe';
 import { assert } from 'tsafe';
 
 import type { CanvasEntityAdapter } from './CanvasEntity/types';
@@ -160,6 +165,34 @@ export class CanvasStateApiModule extends CanvasModuleBase {
     this.store.dispatch(entityRectAdded(arg));
   };
 
+  /**
+   * Adds a raster layer to the canvas, pushing state to redux.
+   */
+  addRasterLayer = (arg: Param0<typeof rasterLayerAdded>) => {
+    this.store.dispatch(rasterLayerAdded(arg));
+  };
+
+  /**
+   * Adds a control layer to the canvas, pushing state to redux.
+   */
+  addControlLayer = (arg: Param0<typeof controlLayerAdded>) => {
+    this.store.dispatch(controlLayerAdded(arg));
+  };
+
+  /**
+   * Adds an inpaint mask to the canvas, pushing state to redux.
+   */
+  addInpaintMask = (arg: Param0<typeof inpaintMaskAdded>) => {
+    this.store.dispatch(inpaintMaskAdded(arg));
+  };
+
+  /**
+   * Adds regional guidance to the canvas, pushing state to redux.
+   */
+  addRegionalGuidance = (arg: Param0<typeof rgAdded>) => {
+    this.store.dispatch(rgAdded(arg));
+  };
+
   /**
    * Rasterizes an entity, pushing state to redux.
    */

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

@@ -126,6 +126,13 @@ export const floorCoord = (coord: Coordinate): Coordinate => {
   };
 };
 
+export const roundCoord = (coord: Coordinate): Coordinate => {
+  return {
+    x: Math.round(coord.x),
+    y: Math.round(coord.y),
+  };
+};
+
 /**
  * Snaps a position to the edge of the given rect if within a threshold of the edge
  * @param pos The position to snap

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

@@ -29,7 +29,7 @@ import { isMainModelBase, zModelIdentifierField } from 'features/nodes/types/com
 import { ASPECT_RATIO_MAP } from 'features/parameters/components/Bbox/constants';
 import { getGridSize, getIsSizeOptimal, getOptimalDimension } from 'features/parameters/util/optimalDimension';
 import type { IRect } from 'konva/lib/types';
-import { merge, omit } from 'lodash-es';
+import { merge } from 'lodash-es';
 import type { UndoableOptions } from 'redux-undo';
 import type { ControlNetModelConfig, ImageDTO, IPAdapterModelConfig, T2IAdapterModelConfig } from 'services/api/types';
 import { assert } from 'tsafe';
@@ -57,13 +57,13 @@ import type {
 } from './types';
 import { getEntityIdentifier, isRenderableEntity } from './types';
 import {
+  converters,
   getControlLayerState,
   getInpaintMaskState,
   getRasterLayerState,
   getReferenceImageState,
   getRegionalGuidanceState,
   imageDTOToImageWithDims,
-  initialControlNet,
   initialIPAdapter,
 } from './util';
 
@@ -157,28 +157,25 @@ export const canvasSlice = createSlice({
       reducer: (
         state,
         action: PayloadAction<
-          EntityIdentifierPayload<{ newId: string; overrides?: Partial<CanvasControlLayerState> }, 'raster_layer'>
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasControlLayerState>; replace?: boolean },
+            'raster_layer'
+          >
         >
       ) => {
-        const { entityIdentifier, newId, overrides } = action.payload;
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
         const layer = selectEntity(state, entityIdentifier);
         if (!layer) {
           return;
         }
 
         // Convert the raster layer to control layer
-        const controlLayerState: CanvasControlLayerState = {
-          ...deepClone(layer),
-          id: newId,
-          type: 'control_layer',
-          controlAdapter: deepClone(initialControlNet),
-          withTransparencyEffect: true,
-        };
+        const controlLayerState = converters.rasterLayer.toControlLayer(newId, layer, overrides);
 
-        merge(controlLayerState, overrides);
-
-        // Remove the raster layer
-        state.rasterLayers.entities = state.rasterLayers.entities.filter((layer) => layer.id !== entityIdentifier.id);
+        if (replace) {
+          // Remove the raster layer
+          state.rasterLayers.entities = state.rasterLayers.entities.filter((layer) => layer.id !== entityIdentifier.id);
+        }
 
         // Add the converted control layer
         state.controlLayers.entities.push(controlLayerState);
@@ -186,11 +183,90 @@ export const canvasSlice = createSlice({
         state.selectedEntityIdentifier = { type: controlLayerState.type, id: controlLayerState.id };
       },
       prepare: (
-        payload: EntityIdentifierPayload<{ overrides?: Partial<CanvasControlLayerState> } | undefined, 'raster_layer'>
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasControlLayerState>; replace?: boolean } | undefined,
+          'raster_layer'
+        >
       ) => ({
         payload: { ...payload, newId: getPrefixedId('control_layer') },
       }),
     },
+    rasterLayerConvertedToInpaintMask: {
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasInpaintMaskState>; replace?: boolean },
+            'raster_layer'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
+        const layer = selectEntity(state, entityIdentifier);
+        if (!layer) {
+          return;
+        }
+
+        // Convert the raster layer to inpaint mask
+        const inpaintMaskState = converters.rasterLayer.toInpaintMask(newId, layer, overrides);
+
+        if (replace) {
+          // Remove the raster layer
+          state.rasterLayers.entities = state.rasterLayers.entities.filter((layer) => layer.id !== entityIdentifier.id);
+        }
+
+        // Add the converted inpaint mask
+        state.inpaintMasks.entities.push(inpaintMaskState);
+
+        state.selectedEntityIdentifier = { type: inpaintMaskState.type, id: inpaintMaskState.id };
+      },
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasInpaintMaskState>; replace?: boolean } | undefined,
+          'raster_layer'
+        >
+      ) => ({
+        payload: { ...payload, newId: getPrefixedId('inpaint_mask') },
+      }),
+    },
+    rasterLayerConvertedToRegionalGuidance: {
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasRegionalGuidanceState>; replace?: boolean },
+            'raster_layer'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
+        const layer = selectEntity(state, entityIdentifier);
+        if (!layer) {
+          return;
+        }
+
+        // Convert the raster layer to inpaint mask
+        const regionalGuidanceState = converters.rasterLayer.toRegionalGuidance(newId, layer, overrides);
+
+        if (replace) {
+          // Remove the raster layer
+          state.rasterLayers.entities = state.rasterLayers.entities.filter((layer) => layer.id !== entityIdentifier.id);
+        }
+
+        // Add the converted inpaint mask
+        state.regionalGuidance.entities.push(regionalGuidanceState);
+
+        state.selectedEntityIdentifier = { type: regionalGuidanceState.type, id: regionalGuidanceState.id };
+      },
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasRegionalGuidanceState>; replace?: boolean } | undefined,
+          'raster_layer'
+        >
+      ) => ({
+        payload: { ...payload, newId: getPrefixedId('regional_guidance') },
+      }),
+    },
     //#region Control layers
     controlLayerAdded: {
       reducer: (
@@ -217,32 +293,125 @@ export const canvasSlice = createSlice({
       state.selectedEntityIdentifier = { type: 'control_layer', id: data.id };
     },
     controlLayerConvertedToRasterLayer: {
-      reducer: (state, action: PayloadAction<EntityIdentifierPayload<{ newId: string }, 'control_layer'>>) => {
-        const { entityIdentifier, newId } = action.payload;
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasRasterLayerState>; replace?: boolean },
+            'control_layer'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
         const layer = selectEntity(state, entityIdentifier);
         if (!layer) {
           return;
         }
 
         // Convert the raster layer to control layer
-        const rasterLayerState: CanvasRasterLayerState = {
-          ...omit(deepClone(layer), ['type', 'controlAdapter', 'withTransparencyEffect']),
-          id: newId,
-          type: 'raster_layer',
-        };
+        const rasterLayerState = converters.controlLayer.toRasterLayer(newId, layer, overrides);
 
-        // Remove the control layer
-        state.controlLayers.entities = state.controlLayers.entities.filter((layer) => layer.id !== entityIdentifier.id);
+        if (replace) {
+          // Remove the control layer
+          state.controlLayers.entities = state.controlLayers.entities.filter(
+            (layer) => layer.id !== entityIdentifier.id
+          );
+        }
 
         // Add the new raster layer
         state.rasterLayers.entities.push(rasterLayerState);
 
         state.selectedEntityIdentifier = { type: rasterLayerState.type, id: rasterLayerState.id };
       },
-      prepare: (payload: EntityIdentifierPayload<void, 'control_layer'>) => ({
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasRasterLayerState>; replace?: boolean } | undefined,
+          'control_layer'
+        >
+      ) => ({
         payload: { ...payload, newId: getPrefixedId('raster_layer') },
       }),
     },
+    controlLayerConvertedToInpaintMask: {
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasInpaintMaskState>; replace?: boolean },
+            'control_layer'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
+        const layer = selectEntity(state, entityIdentifier);
+        if (!layer) {
+          return;
+        }
+
+        // Convert the control layer to inpaint mask
+        const inpaintMaskState = converters.controlLayer.toInpaintMask(newId, layer, overrides);
+
+        if (replace) {
+          // Remove the control layer
+          state.controlLayers.entities = state.controlLayers.entities.filter(
+            (layer) => layer.id !== entityIdentifier.id
+          );
+        }
+
+        // Add the new inpaint mask
+        state.inpaintMasks.entities.push(inpaintMaskState);
+
+        state.selectedEntityIdentifier = { type: inpaintMaskState.type, id: inpaintMaskState.id };
+      },
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasInpaintMaskState>; replace?: boolean } | undefined,
+          'control_layer'
+        >
+      ) => ({
+        payload: { ...payload, newId: getPrefixedId('inpaint_mask') },
+      }),
+    },
+    controlLayerConvertedToRegionalGuidance: {
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasRegionalGuidanceState>; replace?: boolean },
+            'control_layer'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
+        const layer = selectEntity(state, entityIdentifier);
+        if (!layer) {
+          return;
+        }
+
+        // Convert the control layer to regional guidance
+        const regionalGuidanceState = converters.controlLayer.toRegionalGuidance(newId, layer, overrides);
+
+        if (replace) {
+          // Remove the control layer
+          state.controlLayers.entities = state.controlLayers.entities.filter(
+            (layer) => layer.id !== entityIdentifier.id
+          );
+        }
+
+        // Add the new regional guidance
+        state.regionalGuidance.entities.push(regionalGuidanceState);
+
+        state.selectedEntityIdentifier = { type: regionalGuidanceState.type, id: regionalGuidanceState.id };
+      },
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasRegionalGuidanceState>; replace?: boolean } | undefined,
+          'control_layer'
+        >
+      ) => ({
+        payload: { ...payload, newId: getPrefixedId('regional_guidance') },
+      }),
+    },
     controlLayerModelChanged: (
       state,
       action: PayloadAction<
@@ -447,6 +616,46 @@ export const canvasSlice = createSlice({
       state.regionalGuidance.entities.push(data);
       state.selectedEntityIdentifier = { type: 'regional_guidance', id: data.id };
     },
+    rgConvertedToInpaintMask: {
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasInpaintMaskState>; replace?: boolean },
+            'regional_guidance'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
+        const layer = selectEntity(state, entityIdentifier);
+        if (!layer) {
+          return;
+        }
+
+        // Convert the regional guidance to inpaint mask
+        const inpaintMaskState = converters.regionalGuidance.toInpaintMask(newId, layer, overrides);
+
+        if (replace) {
+          // Remove the regional guidance
+          state.regionalGuidance.entities = state.regionalGuidance.entities.filter(
+            (layer) => layer.id !== entityIdentifier.id
+          );
+        }
+
+        // Add the new inpaint mask
+        state.inpaintMasks.entities.push(inpaintMaskState);
+
+        state.selectedEntityIdentifier = { type: inpaintMaskState.type, id: inpaintMaskState.id };
+      },
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasInpaintMaskState>; replace?: boolean } | undefined,
+          'regional_guidance'
+        >
+      ) => ({
+        payload: { ...payload, newId: getPrefixedId('inpaint_mask') },
+      }),
+    },
     rgPositivePromptChanged: (
       state,
       action: PayloadAction<EntityIdentifierPayload<{ prompt: string | null }, 'regional_guidance'>>
@@ -644,6 +853,44 @@ export const canvasSlice = createSlice({
       state.inpaintMasks.entities = [data];
       state.selectedEntityIdentifier = { type: 'inpaint_mask', id: data.id };
     },
+    inpaintMaskConvertedToRegionalGuidance: {
+      reducer: (
+        state,
+        action: PayloadAction<
+          EntityIdentifierPayload<
+            { newId: string; overrides?: Partial<CanvasRegionalGuidanceState>; replace?: boolean },
+            'inpaint_mask'
+          >
+        >
+      ) => {
+        const { entityIdentifier, newId, overrides, replace } = action.payload;
+        const layer = selectEntity(state, entityIdentifier);
+        if (!layer) {
+          return;
+        }
+
+        // Convert the inpaint mask to regional guidance
+        const regionalGuidanceState = converters.inpaintMask.toRegionalGuidance(newId, layer, overrides);
+
+        if (replace) {
+          // Remove the inpaint mask
+          state.inpaintMasks.entities = state.inpaintMasks.entities.filter((layer) => layer.id !== entityIdentifier.id);
+        }
+
+        // Add the new regional guidance
+        state.regionalGuidance.entities.push(regionalGuidanceState);
+
+        state.selectedEntityIdentifier = { type: regionalGuidanceState.type, id: regionalGuidanceState.id };
+      },
+      prepare: (
+        payload: EntityIdentifierPayload<
+          { overrides?: Partial<CanvasRegionalGuidanceState>; replace?: boolean } | undefined,
+          'inpaint_mask'
+        >
+      ) => ({
+        payload: { ...payload, newId: getPrefixedId('regional_guidance') },
+      }),
+    },
     //#region BBox
     bboxScaledWidthChanged: (state, action: PayloadAction<number>) => {
       const gridSize = getGridSize(state.bbox.modelBase);
@@ -1210,10 +1457,14 @@ export const {
   rasterLayerAdded,
   // rasterLayerRecalled,
   rasterLayerConvertedToControlLayer,
+  rasterLayerConvertedToInpaintMask,
+  rasterLayerConvertedToRegionalGuidance,
   // Control layers
   controlLayerAdded,
   // controlLayerRecalled,
   controlLayerConvertedToRasterLayer,
+  controlLayerConvertedToInpaintMask,
+  controlLayerConvertedToRegionalGuidance,
   controlLayerModelChanged,
   controlLayerControlModeChanged,
   controlLayerWeightChanged,
@@ -1231,6 +1482,7 @@ export const {
   // Regions
   rgAdded,
   // rgRecalled,
+  rgConvertedToInpaintMask,
   rgPositivePromptChanged,
   rgNegativePromptChanged,
   rgAutoNegativeToggled,
@@ -1244,6 +1496,7 @@ export const {
   rgIPAdapterCLIPVisionModelChanged,
   // Inpaint mask
   inpaintMaskAdded,
+  inpaintMaskConvertedToRegionalGuidance,
   // inpaintMaskRecalled,
 } = canvasSlice.actions;
 

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

@@ -131,7 +131,8 @@ const zSAMPoint = z.object({
   y: z.number().int().gte(0),
   label: zSAMPointLabel,
 });
-export type SAMPoint = z.infer<typeof zSAMPoint>;
+type SAMPoint = z.infer<typeof zSAMPoint>;
+export type SAMPointWithId = SAMPoint & { id: string };
 
 const zRect = z.object({
   x: z.number(),

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

@@ -184,3 +184,153 @@ export const getInpaintMaskState = (
   merge(entityState, overrides);
   return entityState;
 };
+
+const convertRasterLayerToControlLayer = (
+  newId: string,
+  rasterLayerState: CanvasRasterLayerState,
+  overrides?: Partial<CanvasControlLayerState>
+): CanvasControlLayerState => {
+  const { name, objects, position } = rasterLayerState;
+  const controlLayerState = getControlLayerState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(controlLayerState, overrides);
+  return controlLayerState;
+};
+
+const convertRasterLayerToInpaintMask = (
+  newId: string,
+  rasterLayerState: CanvasRasterLayerState,
+  overrides?: Partial<CanvasInpaintMaskState>
+): CanvasInpaintMaskState => {
+  const { name, objects, position } = rasterLayerState;
+  const inpaintMaskState = getInpaintMaskState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(inpaintMaskState, overrides);
+  return inpaintMaskState;
+};
+
+const convertRasterLayerToRegionalGuidance = (
+  newId: string,
+  rasterLayerState: CanvasRasterLayerState,
+  overrides?: Partial<CanvasRegionalGuidanceState>
+): CanvasRegionalGuidanceState => {
+  const { name, objects, position } = rasterLayerState;
+  const regionalGuidanceState = getRegionalGuidanceState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(regionalGuidanceState, overrides);
+  return regionalGuidanceState;
+};
+
+const convertControlLayerToRasterLayer = (
+  newId: string,
+  controlLayerState: CanvasControlLayerState,
+  overrides?: Partial<CanvasRasterLayerState>
+): CanvasRasterLayerState => {
+  const { name, objects, position } = controlLayerState;
+  const rasterLayerState = getRasterLayerState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(rasterLayerState, overrides);
+  return rasterLayerState;
+};
+
+const convertControlLayerToInpaintMask = (
+  newId: string,
+  rasterLayerState: CanvasControlLayerState,
+  overrides?: Partial<CanvasInpaintMaskState>
+): CanvasInpaintMaskState => {
+  const { name, objects, position } = rasterLayerState;
+  const inpaintMaskState = getInpaintMaskState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(inpaintMaskState, overrides);
+  return inpaintMaskState;
+};
+
+const convertControlLayerToRegionalGuidance = (
+  newId: string,
+  rasterLayerState: CanvasControlLayerState,
+  overrides?: Partial<CanvasRegionalGuidanceState>
+): CanvasRegionalGuidanceState => {
+  const { name, objects, position } = rasterLayerState;
+  const regionalGuidanceState = getRegionalGuidanceState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(regionalGuidanceState, overrides);
+  return regionalGuidanceState;
+};
+
+const convertInpaintMaskToRegionalGuidance = (
+  newId: string,
+  inpaintMaskState: CanvasInpaintMaskState,
+  overrides?: Partial<CanvasRegionalGuidanceState>
+): CanvasRegionalGuidanceState => {
+  const { name, objects, position } = inpaintMaskState;
+  const regionalGuidanceState = getRegionalGuidanceState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(regionalGuidanceState, overrides);
+  return regionalGuidanceState;
+};
+
+const convertRegionalGuidanceToInpaintMask = (
+  newId: string,
+  regionalGuidanceState: CanvasRegionalGuidanceState,
+  overrides?: Partial<CanvasInpaintMaskState>
+): CanvasInpaintMaskState => {
+  const { name, objects, position } = regionalGuidanceState;
+  const inpaintMaskState = getInpaintMaskState(newId, {
+    name,
+    objects,
+    position,
+  });
+  merge(inpaintMaskState, overrides);
+  return inpaintMaskState;
+};
+
+/**
+ * Supported conversions:
+ * - Raster Layer -> Control Layer
+ * - Raster Layer -> Inpaint Mask
+ * - Raster Layer -> Regional Guidance
+ * - Control Layer -> Control Layer
+ * - Control Layer -> Inpaint Mask
+ * - Control Layer -> Regional Guidance
+ * - Inpaint Mask -> Regional Guidance
+ * - Regional Guidance -> Inpaint Mask
+ */
+export const converters = {
+  rasterLayer: {
+    toControlLayer: convertRasterLayerToControlLayer,
+    toInpaintMask: convertRasterLayerToInpaintMask,
+    toRegionalGuidance: convertRasterLayerToRegionalGuidance,
+  },
+  controlLayer: {
+    toRasterLayer: convertControlLayerToRasterLayer,
+    toInpaintMask: convertControlLayerToInpaintMask,
+    toRegionalGuidance: convertControlLayerToRegionalGuidance,
+  },
+  inpaintMask: {
+    toRegionalGuidance: convertInpaintMaskToRegionalGuidance,
+  },
+  regionalGuidance: {
+    toInpaintMask: convertRegionalGuidanceToInpaintMask,
+  },
+};

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

@@ -1,4 +1,4 @@
-import { Flex, Link, Spacer, Text } from '@invoke-ai/ui-library';
+import { Link } from '@invoke-ai/ui-library';
 import { useStore } from '@nanostores/react';
 import { $projectName, $projectUrl } from 'app/store/nanostores/projectId';
 import { memo } from 'react';
@@ -9,15 +9,13 @@ export const GalleryHeader = memo(() => {
 
   if (projectName && projectUrl) {
     return (
-      <Flex gap={2} alignItems="center" justifyContent="space-evenly" pe={2} w="50%">
-        <Text fontSize="md" fontWeight="semibold" noOfLines={1} wordBreak="break-all" w="full" textAlign="center">
-          <Link href={projectUrl}>{projectName}</Link>
-        </Text>
-      </Flex>
+      <Link fontSize="md" fontWeight="semibold" noOfLines={1} wordBreak="break-all" href={projectUrl}>
+        {projectName}
+      </Link>
     );
   }
 
-  return <Spacer />;
+  return null;
 });
 
 GalleryHeader.displayName = 'GalleryHeader';

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

@@ -51,8 +51,8 @@ const GalleryPanelContent = () => {
 
   return (
     <Flex ref={galleryPanelFocusRef} position="relative" flexDirection="column" h="full" w="full" tabIndex={-1}>
-      <Flex alignItems="center" w="full">
-        <Flex w="25%">
+      <Flex alignItems="center" justifyContent="space-between" w="full">
+        <Flex flexGrow={1} flexBasis={0}>
           <Button
             size="sm"
             variant="ghost"
@@ -62,8 +62,10 @@ const GalleryPanelContent = () => {
             {boardsListPanel.isCollapsed ? t('boards.viewBoards') : t('boards.hideBoards')}
           </Button>
         </Flex>
-        <GalleryHeader />
-        <Flex h="full" w="25%" justifyContent="flex-end">
+        <Flex>
+          <GalleryHeader />
+        </Flex>
+        <Flex flexGrow={1} flexBasis={0} justifyContent="flex-end">
           <BoardsSettingsPopover />
           <IconButton
             size="sm"

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

@@ -1,9 +1,11 @@
-import { MenuItem } from '@invoke-ai/ui-library';
+import { Menu, MenuButton, MenuItem, MenuList } from '@invoke-ai/ui-library';
+import { SubMenuButtonContent, useSubMenu } from 'common/hooks/useSubMenu';
 import { useImageDTOContext } from 'features/gallery/contexts/ImageDTOContext';
 import { useImageActions } from 'features/gallery/hooks/useImageActions';
 import { memo } from 'react';
 import { useTranslation } from 'react-i18next';
 import {
+  PiArrowBendUpLeftBold,
   PiArrowsCounterClockwiseBold,
   PiAsteriskBold,
   PiPaintBrushBold,
@@ -14,28 +16,36 @@ import {
 export const ImageMenuItemMetadataRecallActions = memo(() => {
   const { t } = useTranslation();
   const imageDTO = useImageDTOContext();
+  const subMenu = useSubMenu();
 
   const { recallAll, remix, recallSeed, recallPrompts, hasMetadata, hasSeed, hasPrompts, createAsPreset } =
     useImageActions(imageDTO);
 
   return (
-    <>
-      <MenuItem icon={<PiArrowsCounterClockwiseBold />} onClickCapture={remix} isDisabled={!hasMetadata}>
-        {t('parameters.remixImage')}
-      </MenuItem>
-      <MenuItem icon={<PiQuotesBold />} onClickCapture={recallPrompts} isDisabled={!hasPrompts}>
-        {t('parameters.usePrompt')}
-      </MenuItem>
-      <MenuItem icon={<PiPlantBold />} onClickCapture={recallSeed} isDisabled={!hasSeed}>
-        {t('parameters.useSeed')}
-      </MenuItem>
-      <MenuItem icon={<PiAsteriskBold />} onClickCapture={recallAll} isDisabled={!hasMetadata}>
-        {t('parameters.useAll')}
-      </MenuItem>
-      <MenuItem icon={<PiPaintBrushBold />} onClickCapture={createAsPreset} isDisabled={!hasPrompts}>
-        {t('stylePresets.useForTemplate')}
-      </MenuItem>
-    </>
+    <MenuItem {...subMenu.parentMenuItemProps} icon={<PiArrowBendUpLeftBold />}>
+      <Menu {...subMenu.menuProps}>
+        <MenuButton {...subMenu.menuButtonProps}>
+          <SubMenuButtonContent label="Recall Metadata" />
+        </MenuButton>
+        <MenuList {...subMenu.menuListProps}>
+          <MenuItem icon={<PiArrowsCounterClockwiseBold />} onClick={remix} isDisabled={!hasMetadata}>
+            {t('parameters.remixImage')}
+          </MenuItem>
+          <MenuItem icon={<PiQuotesBold />} onClick={recallPrompts} isDisabled={!hasPrompts}>
+            {t('parameters.usePrompt')}
+          </MenuItem>
+          <MenuItem icon={<PiPlantBold />} onClick={recallSeed} isDisabled={!hasSeed}>
+            {t('parameters.useSeed')}
+          </MenuItem>
+          <MenuItem icon={<PiAsteriskBold />} onClick={recallAll} isDisabled={!hasMetadata}>
+            {t('parameters.useAll')}
+          </MenuItem>
+          <MenuItem icon={<PiPaintBrushBold />} onClick={createAsPreset} isDisabled={!hasPrompts}>
+            {t('stylePresets.useForTemplate')}
+          </MenuItem>
+        </MenuList>
+      </Menu>
+    </MenuItem>
   );
 });
 

invokeai/frontend/web/src/features/modelManagerV2/hooks/useBuildModelsToInstall.ts

@@ -21,9 +21,12 @@ export const useBuildModelInstallArg = () => {
   });
 
   const getIsInstalled = useCallback(
-    ({ source, name, base, type, is_installed }: StarterModel): boolean =>
+    ({ source, name, base, type, is_installed, previous_names }: StarterModel): boolean =>
       modelList.some(
-        (mc) => is_installed || source === mc.source || (base === mc.base && name === mc.name && type === mc.type)
+        (mc) =>
+          is_installed ||
+          source === mc.source ||
+          (base === mc.base && (name === mc.name || previous_names?.includes(name)) && type === mc.type)
       ),
     [modelList]
   );

invokeai/frontend/web/src/features/modelManagerV2/subpanels/AddModelPanel/StarterModels/StarterBundle.tsx

@@ -1,4 +1,4 @@
-import { Button, Flex, Text, Tooltip } from '@invoke-ai/ui-library';
+import { Button, Flex, ListItem, Text, Tooltip, UnorderedList } from '@invoke-ai/ui-library';
 import { flattenStarterModel, useBuildModelInstallArg } from 'features/modelManagerV2/hooks/useBuildModelsToInstall';
 import { isMainModelBase } from 'features/nodes/types/common';
 import { MODEL_TYPE_SHORT_MAP } from 'features/parameters/types/constants';
@@ -44,8 +44,15 @@ export const StarterBundle = ({ bundleName, bundle }: { bundleName: string; bund
   return (
     <Tooltip
       label={
-        <Flex flexDir="column">
-          <Text>{t('modelManager.includesNModels', { n: bundle.length })}</Text>
+        <Flex flexDir="column" p={1}>
+          <Text>{t('modelManager.includesNModels', { n: bundle.length })}:</Text>
+          <UnorderedList>
+            {bundle.map((model, index) => (
+              <ListItem key={index} wordBreak="break-all">
+                {model.name}
+              </ListItem>
+            ))}
+          </UnorderedList>
         </Flex>
       }
     >

invokeai/frontend/web/src/features/modelManagerV2/subpanels/AddModelPanel/StarterModels/StarterModelsResults.tsx

@@ -1,14 +1,4 @@
-import {
-  Box,
-  Flex,
-  Icon,
-  IconButton,
-  Input,
-  InputGroup,
-  InputRightElement,
-  Text,
-  Tooltip,
-} from '@invoke-ai/ui-library';
+import { Flex, Icon, IconButton, Input, InputGroup, InputRightElement, Text, Tooltip } from '@invoke-ai/ui-library';
 import ScrollableContent from 'common/components/OverlayScrollbars/ScrollableContent';
 import { map, size } from 'lodash-es';
 import type { ChangeEventHandler } from 'react';
@@ -59,14 +49,14 @@ export const StarterModelsResults = memo(({ results }: StarterModelsResultsProps
       <Flex justifyContent="space-between" alignItems="center">
         {size(results.starter_bundles) > 0 && (
           <Flex gap={4} alignItems="center">
-            <Flex gap={1} alignItems="center">
+            <Flex gap={2} alignItems="center">
               <Text color="base.200" fontWeight="semibold">
                 {t('modelManager.starterBundles')}
               </Text>
               <Tooltip label={t('modelManager.starterBundleHelpText')}>
-                <Box>
+                <Flex alignItems="center">
                   <Icon as={PiInfoBold} color="base.200" />
-                </Box>
+                </Flex>
               </Tooltip>
             </Flex>
             <Flex gap={2}>

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

@@ -106,10 +106,12 @@ export const getInfill = (
   }
 
   if (infillMethod === 'color') {
+    const { a, ...rgb } = infillColorValue;
+    const color = { ...rgb, a: Math.round(a * 255) };
     return g.addNode({
       id: 'infill_rgba',
       type: 'infill_rgba',
-      color: infillColorValue,
+      color,
     });
   }
 

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

@@ -14731,7 +14731,7 @@ export type components = {
             bounding_boxes?: components["schemas"]["BoundingBoxField"][] | null;
             /**
              * Point Lists
-             * @description The points to prompt the SAM model with.
+             * @description The list of point lists to prompt the SAM model with. Each list of points represents a single object.
              * @default null
              */
             point_lists?: components["schemas"]["SAMPointsField"][] | null;
@@ -15347,6 +15347,11 @@ export type components = {
              * @default false
              */
             is_installed?: boolean;
+            /**
+             * Previous Names
+             * @default []
+             */
+            previous_names?: string[];
             /** Dependencies */
             dependencies?: components["schemas"]["StarterModelWithoutDependencies"][] | null;
         };
@@ -15375,6 +15380,11 @@ export type components = {
              * @default false
              */
             is_installed?: boolean;
+            /**
+             * Previous Names
+             * @default []
+             */
+            previous_names?: string[];
         };
         /**
          * Step Param Easing

pyproject.toml

@@ -89,7 +89,7 @@ dependencies = [
   "pypatchmatch",
   'pyperclip',
   "pyreadline3",
-  "python-multipart",
+  "python-multipart==0.0.12",
   "requests~=2.28.2",
   "rich~=13.3",
   "scikit-image~=0.21.0",

tests/backend/sd3/test_sd3_state_dict_utils.py

@@ -1,43 +0,0 @@
-import pytest
-import torch
-
-from invokeai.backend.sd3.sd3_mmditx import Sd3MMDiTX
-from invokeai.backend.sd3.sd3_state_dict_utils import infer_sd3_mmditx_params, is_sd3_checkpoint
-from tests.backend.sd3.sd3_5_mmditx_state_dict import sd3_sd_shapes
-
-
-@pytest.mark.parametrize(
-    ["sd_shapes", "expected"],
-    [
-        (sd3_sd_shapes, True),
-        ({}, False),
-        ({"foo": [1]}, False),
-    ],
-)
-def test_is_sd3_checkpoint(sd_shapes: dict[str, list[int]], expected: bool):
-    # Build mock state dict from the provided shape dict.
-    sd = {k: None for k in sd_shapes}
-    assert is_sd3_checkpoint(sd) == expected
-
-
-def test_infer_sd3_mmditx_params():
-    # Build mock state dict on the meta device.
-    with torch.device("meta"):
-        sd = {k: torch.zeros(shape) for k, shape in sd3_sd_shapes.items()}
-
-    # Filter the MMDiTX parameters from the state dict.
-    sd = {k: v for k, v in sd.items() if k.startswith("model.diffusion_model.")}
-
-    params = infer_sd3_mmditx_params(sd)
-
-    # Construct model from params.
-    with torch.device("meta"):
-        model = Sd3MMDiTX(params=params)
-
-    model_sd = model.state_dict()
-
-    # Assert that the model state dict is compatible with the original state dict.
-    sd_without_prefix = {k.split("model.diffusion_model.")[-1]: v for k, v in model_sd.items()}
-    assert set(model_sd.keys()) == set(sd_without_prefix.keys())
-    for k in model_sd:
-        assert model_sd[k].shape == sd_without_prefix[k].shape