Rename flux_kohya_lora_conversion_utils.py

* [MM] add API routes for getting & setting MM cache sizes, and retrieving MM stats

* Update invokeai/app/api/routers/model_manager.py

Co-authored-by: Ryan Dick <ryanjdick3@gmail.com>

* code cleanup after @ryand review

* Update invokeai/app/api/routers/model_manager.py

Co-authored-by: Ryan Dick <ryanjdick3@gmail.com>

* fix merge conflicts; tested and working

---------

Co-authored-by: Lincoln Stein <lstein@gmail.com>
Co-authored-by: Ryan Dick <ryanjdick3@gmail.com>

.github/workflows/python-tests.yml

@@ -60,7 +60,7 @@ jobs:
             extra-index-url: 'https://download.pytorch.org/whl/cpu'
             github-env: $GITHUB_ENV
           - platform: macos-default
-            os: macOS-12
+            os: macOS-14
             github-env: $GITHUB_ENV
           - platform: windows-cpu
             os: windows-2022

docs/help/FAQ.md

@@ -196,6 +196,22 @@ tips to reduce the problem:
     === "12GB VRAM GPU"
 
         This should be sufficient to generate larger images up to about 1280x1280.
+		
+## Checkpoint Models Load Slowly or Use Too Much RAM
+
+The difference between diffusers models (a folder containing multiple
+subfolders) and checkpoint models (a file ending with .safetensors or
+.ckpt) is that InvokeAI is able to load diffusers models into memory
+incrementally, while checkpoint models must be loaded all at
+once. With very large models, or systems with limited RAM, you may
+experience slowdowns and other memory-related issues when loading
+checkpoint models.
+
+To solve this, go to the Model Manager tab (the cube), select the
+checkpoint model that's giving you trouble, and press the "Convert"
+button in the upper right of your browser window. This will conver the
+checkpoint into a diffusers model, after which loading should be
+faster and less memory-intensive.
 
 ## Memory Leak (Linux)
 

invokeai/app/api/routers/model_manager.py

@@ -3,8 +3,10 @@
 
 import io
 import pathlib
+import shutil
 import traceback
 from copy import deepcopy
+from enum import Enum
 from tempfile import TemporaryDirectory
 from typing import List, Optional, Type
 
@@ -17,6 +19,7 @@ from starlette.exceptions import HTTPException
 from typing_extensions import Annotated
 
 from invokeai.app.api.dependencies import ApiDependencies
+from invokeai.app.services.config import get_config
 from invokeai.app.services.model_images.model_images_common import ModelImageFileNotFoundException
 from invokeai.app.services.model_install.model_install_common import ModelInstallJob
 from invokeai.app.services.model_records import (
@@ -31,6 +34,7 @@ from invokeai.backend.model_manager.config import (
     ModelFormat,
     ModelType,
 )
+from invokeai.backend.model_manager.load.model_cache.model_cache_base import CacheStats
 from invokeai.backend.model_manager.metadata.fetch.huggingface import HuggingFaceMetadataFetch
 from invokeai.backend.model_manager.metadata.metadata_base import ModelMetadataWithFiles, UnknownMetadataException
 from invokeai.backend.model_manager.search import ModelSearch
@@ -50,6 +54,13 @@ class ModelsList(BaseModel):
     model_config = ConfigDict(use_enum_values=True)
 
 
+class CacheType(str, Enum):
+    """Cache type - one of vram or ram."""
+
+    RAM = "RAM"
+    VRAM = "VRAM"
+
+
 def add_cover_image_to_model_config(config: AnyModelConfig, dependencies: Type[ApiDependencies]) -> AnyModelConfig:
     """Add a cover image URL to a model configuration."""
     cover_image = dependencies.invoker.services.model_images.get_url(config.key)
@@ -797,3 +808,83 @@ async def get_starter_models() -> list[StarterModel]:
         model.dependencies = missing_deps
 
     return starter_models
+
+
+@model_manager_router.get(
+    "/model_cache",
+    operation_id="get_cache_size",
+    response_model=float,
+    summary="Get maximum size of model manager RAM or VRAM cache.",
+)
+async def get_cache_size(cache_type: CacheType = Query(description="The cache type", default=CacheType.RAM)) -> float:
+    """Return the current RAM or VRAM cache size setting (in GB)."""
+    cache = ApiDependencies.invoker.services.model_manager.load.ram_cache
+    value = 0.0
+    if cache_type == CacheType.RAM:
+        value = cache.max_cache_size
+    elif cache_type == CacheType.VRAM:
+        value = cache.max_vram_cache_size
+    return value
+
+
+@model_manager_router.put(
+    "/model_cache",
+    operation_id="set_cache_size",
+    response_model=float,
+    summary="Set maximum size of model manager RAM or VRAM cache, optionally writing new value out to invokeai.yaml config file.",
+)
+async def set_cache_size(
+    value: float = Query(description="The new value for the maximum cache size"),
+    cache_type: CacheType = Query(description="The cache type", default=CacheType.RAM),
+    persist: bool = Query(description="Write new value out to invokeai.yaml", default=False),
+) -> float:
+    """Set the current RAM or VRAM cache size setting (in GB). ."""
+    cache = ApiDependencies.invoker.services.model_manager.load.ram_cache
+    app_config = get_config()
+    # Record initial state.
+    vram_old = app_config.vram
+    ram_old = app_config.ram
+
+    # Prepare target state.
+    vram_new = vram_old
+    ram_new = ram_old
+    if cache_type == CacheType.RAM:
+        ram_new = value
+    elif cache_type == CacheType.VRAM:
+        vram_new = value
+    else:
+        raise ValueError(f"Unexpected {cache_type=}.")
+
+    config_path = app_config.config_file_path
+    new_config_path = config_path.with_suffix(".yaml.new")
+
+    try:
+        # Try to apply the target state.
+        cache.max_vram_cache_size = vram_new
+        cache.max_cache_size = ram_new
+        app_config.ram = ram_new
+        app_config.vram = vram_new
+        if persist:
+            app_config.write_file(new_config_path)
+            shutil.move(new_config_path, config_path)
+    except Exception as e:
+        # If there was a failure, restore the initial state.
+        cache.max_cache_size = ram_old
+        cache.max_vram_cache_size = vram_old
+        app_config.ram = ram_old
+        app_config.vram = vram_old
+
+        raise RuntimeError("Failed to update cache size") from e
+    return value
+
+
+@model_manager_router.get(
+    "/stats",
+    operation_id="get_stats",
+    response_model=Optional[CacheStats],
+    summary="Get model manager RAM cache performance statistics.",
+)
+async def get_stats() -> Optional[CacheStats]:
+    """Return performance statistics on the model manager's RAM cache. Will return null if no models have been loaded."""
+
+    return ApiDependencies.invoker.services.model_manager.load.ram_cache.stats

invokeai/app/api/routers/session_queue.py

@@ -11,7 +11,6 @@ from invokeai.app.services.session_queue.session_queue_common import (
     Batch,
     BatchStatus,
     CancelByBatchIDsResult,
-    CancelByOriginResult,
     ClearResult,
     EnqueueBatchResult,
     PruneResult,
@@ -106,19 +105,6 @@ async def cancel_by_batch_ids(
     return ApiDependencies.invoker.services.session_queue.cancel_by_batch_ids(queue_id=queue_id, batch_ids=batch_ids)
 
 
-@session_queue_router.put(
-    "/{queue_id}/cancel_by_origin",
-    operation_id="cancel_by_origin",
-    responses={200: {"model": CancelByBatchIDsResult}},
-)
-async def cancel_by_origin(
-    queue_id: str = Path(description="The queue id to perform this operation on"),
-    origin: str = Query(description="The origin to cancel all queue items for"),
-) -> CancelByOriginResult:
-    """Immediately cancels all queue items with the given origin"""
-    return ApiDependencies.invoker.services.session_queue.cancel_by_origin(queue_id=queue_id, origin=origin)
-
-
 @session_queue_router.put(
     "/{queue_id}/clear",
     operation_id="clear",

invokeai/app/invocations/compel.py

@@ -19,8 +19,8 @@ from invokeai.app.invocations.model import CLIPField
 from invokeai.app.invocations.primitives import ConditioningOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.ti_utils import generate_ti_list
-from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_patcher import ModelPatcher
+from invokeai.backend.peft.lora import LoRAModelRaw
 from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
     BasicConditioningInfo,
     ConditioningFieldData,

invokeai/app/invocations/denoise_latents.py

@@ -36,9 +36,9 @@ from invokeai.app.invocations.t2i_adapter import T2IAdapterField
 from invokeai.app.services.shared.invocation_context import InvocationContext
 from invokeai.app.util.controlnet_utils import prepare_control_image
 from invokeai.backend.ip_adapter.ip_adapter import IPAdapter
-from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_manager import BaseModelType, ModelVariantType
 from invokeai.backend.model_patcher import ModelPatcher
+from invokeai.backend.peft.lora import LoRAModelRaw
 from invokeai.backend.stable_diffusion import PipelineIntermediateState
 from invokeai.backend.stable_diffusion.denoise_context import DenoiseContext, DenoiseInputs
 from invokeai.backend.stable_diffusion.diffusers_pipeline import (
@@ -185,7 +185,7 @@ class DenoiseLatentsInvocation(BaseInvocation):
     )
     denoise_mask: Optional[DenoiseMaskField] = InputField(
         default=None,
-        description=FieldDescriptions.mask,
+        description=FieldDescriptions.denoise_mask,
         input=Input.Connection,
         ui_order=8,
     )

invokeai/app/invocations/fields.py

@@ -40,14 +40,18 @@ class UIType(str, Enum, metaclass=MetaEnum):
 
     # region Model Field Types
     MainModel = "MainModelField"
+    FluxMainModel = "FluxMainModelField"
     SDXLMainModel = "SDXLMainModelField"
     SDXLRefinerModel = "SDXLRefinerModelField"
     ONNXModel = "ONNXModelField"
     VAEModel = "VAEModelField"
+    FluxVAEModel = "FluxVAEModelField"
     LoRAModel = "LoRAModelField"
     ControlNetModel = "ControlNetModelField"
     IPAdapterModel = "IPAdapterModelField"
     T2IAdapterModel = "T2IAdapterModelField"
+    T5EncoderModel = "T5EncoderModelField"
+    CLIPEmbedModel = "CLIPEmbedModelField"
     SpandrelImageToImageModel = "SpandrelImageToImageModelField"
     # endregion
 
@@ -125,13 +129,17 @@ class FieldDescriptions:
     negative_cond = "Negative conditioning tensor"
     noise = "Noise tensor"
     clip = "CLIP (tokenizer, text encoder, LoRAs) and skipped layer count"
+    t5_encoder = "T5 tokenizer and text encoder"
+    clip_embed_model = "CLIP Embed loader"
     unet = "UNet (scheduler, LoRAs)"
+    transformer = "Transformer"
     vae = "VAE"
     cond = "Conditioning tensor"
     controlnet_model = "ControlNet model to load"
     vae_model = "VAE model to load"
     lora_model = "LoRA model to load"
     main_model = "Main model (UNet, VAE, CLIP) to load"
+    flux_model = "Flux model (Transformer) 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"
@@ -173,7 +181,7 @@ class FieldDescriptions:
     )
     num_1 = "The first number"
     num_2 = "The second number"
-    mask = "The mask to use for the operation"
+    denoise_mask = "A mask of the region to apply the denoising process to."
     board = "The board to save the image to"
     image = "The image to process"
     tile_size = "Tile size"
@@ -231,6 +239,12 @@ class ColorField(BaseModel):
         return (self.r, self.g, self.b, self.a)
 
 
+class FluxConditioningField(BaseModel):
+    """A conditioning tensor primitive value"""
+
+    conditioning_name: str = Field(description="The name of conditioning tensor")
+
+
 class ConditioningField(BaseModel):
     """A conditioning tensor primitive value"""
 

invokeai/app/invocations/flux_denoise.py

@@ -0,0 +1,292 @@
+from typing import Callable, Iterator, Optional, Tuple
+
+import torch
+import torchvision.transforms as tv_transforms
+from torchvision.transforms.functional import resize as tv_resize
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.fields import (
+    DenoiseMaskField,
+    FieldDescriptions,
+    FluxConditioningField,
+    Input,
+    InputField,
+    LatentsField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import TransformerField
+from invokeai.app.invocations.primitives import LatentsOutput
+from invokeai.app.services.session_processor.session_processor_common import CanceledException
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.denoise import denoise
+from invokeai.backend.flux.inpaint_extension import InpaintExtension
+from invokeai.backend.flux.model import Flux
+from invokeai.backend.flux.sampling_utils import (
+    clip_timestep_schedule,
+    generate_img_ids,
+    get_noise,
+    get_schedule,
+    pack,
+    unpack,
+)
+from invokeai.backend.peft.lora import LoRAModelRaw
+from invokeai.backend.peft.peft_patcher import PeftPatcher
+from invokeai.backend.stable_diffusion.diffusion.conditioning_data import FLUXConditioningInfo
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "flux_denoise",
+    title="FLUX Denoise",
+    tags=["image", "flux"],
+    category="image",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class FluxDenoiseInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Run denoising process with a FLUX transformer model."""
+
+    # If latents is provided, this means we are doing image-to-image.
+    latents: Optional[LatentsField] = InputField(
+        default=None,
+        description=FieldDescriptions.latents,
+        input=Input.Connection,
+    )
+    # denoise_mask is used for image-to-image inpainting. Only the masked region is modified.
+    denoise_mask: Optional[DenoiseMaskField] = InputField(
+        default=None,
+        description=FieldDescriptions.denoise_mask,
+        input=Input.Connection,
+    )
+    denoising_start: float = InputField(
+        default=0.0,
+        ge=0,
+        le=1,
+        description=FieldDescriptions.denoising_start,
+    )
+    denoising_end: float = InputField(default=1.0, ge=0, le=1, description=FieldDescriptions.denoising_end)
+    transformer: TransformerField = InputField(
+        description=FieldDescriptions.flux_model,
+        input=Input.Connection,
+        title="Transformer",
+    )
+    positive_text_conditioning: FluxConditioningField = InputField(
+        description=FieldDescriptions.positive_cond, input=Input.Connection
+    )
+    width: int = InputField(default=1024, multiple_of=16, description="Width of the generated image.")
+    height: int = InputField(default=1024, multiple_of=16, description="Height of the generated image.")
+    num_steps: int = InputField(
+        default=4, description="Number of diffusion steps. Recommended values are schnell: 4, dev: 50."
+    )
+    guidance: float = InputField(
+        default=4.0,
+        description="The guidance strength. Higher values adhere more strictly to the prompt, and will produce less diverse images. FLUX dev only, ignored for schnell.",
+    )
+    seed: int = InputField(default=0, description="Randomness seed for reproducibility.")
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        latents = self._run_diffusion(context)
+        latents = latents.detach().to("cpu")
+
+        name = context.tensors.save(tensor=latents)
+        return LatentsOutput.build(latents_name=name, latents=latents, seed=None)
+
+    def _run_diffusion(
+        self,
+        context: InvocationContext,
+    ):
+        inference_dtype = torch.bfloat16
+
+        # Load the conditioning data.
+        cond_data = context.conditioning.load(self.positive_text_conditioning.conditioning_name)
+        assert len(cond_data.conditionings) == 1
+        flux_conditioning = cond_data.conditionings[0]
+        assert isinstance(flux_conditioning, FLUXConditioningInfo)
+        flux_conditioning = flux_conditioning.to(dtype=inference_dtype)
+        t5_embeddings = flux_conditioning.t5_embeds
+        clip_embeddings = flux_conditioning.clip_embeds
+
+        # Load the input latents, if provided.
+        init_latents = context.tensors.load(self.latents.latents_name) if self.latents else None
+        if init_latents is not None:
+            init_latents = init_latents.to(device=TorchDevice.choose_torch_device(), dtype=inference_dtype)
+
+        # Prepare input noise.
+        noise = get_noise(
+            num_samples=1,
+            height=self.height,
+            width=self.width,
+            device=TorchDevice.choose_torch_device(),
+            dtype=inference_dtype,
+            seed=self.seed,
+        )
+
+        transformer_info = context.models.load(self.transformer.transformer)
+        is_schnell = "schnell" in transformer_info.config.config_path
+
+        # Calculate the timestep schedule.
+        image_seq_len = noise.shape[-1] * noise.shape[-2] // 4
+        timesteps = get_schedule(
+            num_steps=self.num_steps,
+            image_seq_len=image_seq_len,
+            shift=not is_schnell,
+        )
+
+        # Clip the timesteps schedule based on denoising_start and denoising_end.
+        timesteps = clip_timestep_schedule(timesteps, self.denoising_start, self.denoising_end)
+
+        # Prepare input latent image.
+        if init_latents is not None:
+            # If init_latents is provided, we are doing image-to-image.
+
+            if is_schnell:
+                context.logger.warning(
+                    "Running image-to-image with a FLUX schnell model. This is not recommended. The results are likely "
+                    "to be poor. Consider using a FLUX dev model instead."
+                )
+
+            # Noise the orig_latents by the appropriate amount for the first timestep.
+            t_0 = timesteps[0]
+            x = t_0 * noise + (1.0 - t_0) * init_latents
+        else:
+            # init_latents are not provided, so we are not doing image-to-image (i.e. we are starting from pure noise).
+            if self.denoising_start > 1e-5:
+                raise ValueError("denoising_start should be 0 when initial latents are not provided.")
+
+            x = noise
+
+        # If len(timesteps) == 1, then short-circuit. We are just noising the input latents, but not taking any
+        # denoising steps.
+        if len(timesteps) <= 1:
+            return x
+
+        inpaint_mask = self._prep_inpaint_mask(context, x)
+
+        b, _c, h, w = x.shape
+        img_ids = generate_img_ids(h=h, w=w, batch_size=b, device=x.device, dtype=x.dtype)
+
+        bs, t5_seq_len, _ = t5_embeddings.shape
+        txt_ids = torch.zeros(bs, t5_seq_len, 3, dtype=inference_dtype, device=TorchDevice.choose_torch_device())
+
+        # Pack all latent tensors.
+        init_latents = pack(init_latents) if init_latents is not None else None
+        inpaint_mask = pack(inpaint_mask) if inpaint_mask is not None else None
+        noise = pack(noise)
+        x = pack(x)
+
+        # Now that we have 'packed' the latent tensors, verify that we calculated the image_seq_len correctly.
+        assert image_seq_len == x.shape[1]
+
+        # Prepare inpaint extension.
+        inpaint_extension: InpaintExtension | None = None
+        if inpaint_mask is not None:
+            assert init_latents is not None
+            inpaint_extension = InpaintExtension(
+                init_latents=init_latents,
+                inpaint_mask=inpaint_mask,
+                noise=noise,
+            )
+
+        with (
+            transformer_info.model_on_device() as (cached_weights, transformer),
+            # Apply the LoRA after transformer has been moved to its target device for faster patching.
+            PeftPatcher.apply_peft_patches(
+                model=transformer,
+                patches=self._lora_iterator(context),
+                prefix="",
+                cached_weights=cached_weights,
+            ),
+        ):
+            assert isinstance(transformer, Flux)
+
+            x = denoise(
+                model=transformer,
+                img=x,
+                img_ids=img_ids,
+                txt=t5_embeddings,
+                txt_ids=txt_ids,
+                vec=clip_embeddings,
+                timesteps=timesteps,
+                step_callback=self._build_step_callback(context),
+                guidance=self.guidance,
+                inpaint_extension=inpaint_extension,
+            )
+
+        x = unpack(x.float(), self.height, self.width)
+        return x
+
+    def _prep_inpaint_mask(self, context: InvocationContext, latents: torch.Tensor) -> torch.Tensor | None:
+        """Prepare the inpaint mask.
+
+        - Loads the mask
+        - Resizes if necessary
+        - Casts to same device/dtype as latents
+        - Expands mask to the same shape as latents so that they line up after 'packing'
+
+        Args:
+            context (InvocationContext): The invocation context, for loading the inpaint mask.
+            latents (torch.Tensor): A latent image tensor. In 'unpacked' format. Used to determine the target shape,
+                device, and dtype for the inpaint mask.
+
+        Returns:
+            torch.Tensor | None: Inpaint mask.
+        """
+        if self.denoise_mask is None:
+            return None
+
+        mask = context.tensors.load(self.denoise_mask.mask_name)
+
+        _, _, latent_height, latent_width = latents.shape
+        mask = tv_resize(
+            img=mask,
+            size=[latent_height, latent_width],
+            interpolation=tv_transforms.InterpolationMode.BILINEAR,
+            antialias=False,
+        )
+
+        mask = mask.to(device=latents.device, dtype=latents.dtype)
+
+        # Expand the inpaint mask to the same shape as `latents` so that when we 'pack' `mask` it lines up with
+        # `latents`.
+        return mask.expand_as(latents)
+
+    def _lora_iterator(self, context: InvocationContext) -> Iterator[Tuple[LoRAModelRaw, float]]:
+        for lora in self.transformer.loras:
+            lora_info = context.models.load(lora.lora)
+            assert isinstance(lora_info.model, LoRAModelRaw)
+            yield (lora_info.model, lora.weight)
+            del lora_info
+
+    def _build_step_callback(self, context: InvocationContext) -> Callable[[], None]:
+        def step_callback() -> None:
+            if context.util.is_canceled():
+                raise CanceledException
+
+            # TODO: Make this look like the image before re-enabling
+            # latent_image = unpack(img.float(), self.height, self.width)
+            # latent_image = latent_image.squeeze()  # Remove unnecessary dimensions
+            # flattened_tensor = latent_image.reshape(-1)  # Flatten to shape [48*128*128]
+
+            # # Create a new tensor of the required shape [255, 255, 3]
+            # latent_image = flattened_tensor[: 255 * 255 * 3].reshape(255, 255, 3)  # Reshape to RGB format
+
+            # # Convert to a NumPy array and then to a PIL Image
+            # image = Image.fromarray(latent_image.cpu().numpy().astype(np.uint8))
+
+            # (width, height) = image.size
+            # width *= 8
+            # height *= 8
+
+            # dataURL = image_to_dataURL(image, image_format="JPEG")
+
+            # # TODO: move this whole function to invocation context to properly reference these variables
+            # context._services.events.emit_invocation_denoise_progress(
+            #     context._data.queue_item,
+            #     context._data.invocation,
+            #     state,
+            #     ProgressImage(dataURL=dataURL, width=width, height=height),
+            # )
+
+        return step_callback

invokeai/app/invocations/flux_lora_loader.py

@@ -0,0 +1,53 @@
+from invokeai.app.invocations.baseinvocation import BaseInvocation, BaseInvocationOutput, invocation, invocation_output
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField, OutputField, UIType
+from invokeai.app.invocations.model import LoRAField, ModelIdentifierField, TransformerField
+from invokeai.app.services.shared.invocation_context import InvocationContext
+
+
+@invocation_output("flux_lora_loader_output")
+class FluxLoRALoaderOutput(BaseInvocationOutput):
+    """FLUX LoRA Loader Output"""
+
+    transformer: TransformerField = OutputField(
+        default=None, description=FieldDescriptions.transformer, title="FLUX Transformer"
+    )
+
+
+@invocation(
+    "flux_lora_loader",
+    title="FLUX LoRA",
+    tags=["lora", "model", "flux"],
+    category="model",
+    version="1.0.0",
+)
+class FluxLoRALoaderInvocation(BaseInvocation):
+    """Apply a LoRA model to a FLUX transformer."""
+
+    lora: ModelIdentifierField = InputField(
+        description=FieldDescriptions.lora_model, title="LoRA", ui_type=UIType.LoRAModel
+    )
+    weight: float = InputField(default=0.75, description=FieldDescriptions.lora_weight)
+    transformer: TransformerField = InputField(
+        description=FieldDescriptions.transformer,
+        input=Input.Connection,
+        title="FLUX Transformer",
+    )
+
+    def invoke(self, context: InvocationContext) -> FluxLoRALoaderOutput:
+        lora_key = self.lora.key
+
+        if not context.models.exists(lora_key):
+            raise ValueError(f"Unknown lora: {lora_key}!")
+
+        if any(lora.lora.key == lora_key for lora in self.transformer.loras):
+            raise Exception(f'LoRA "{lora_key}" already applied to transformer.')
+
+        transformer = self.transformer.model_copy(deep=True)
+        transformer.loras.append(
+            LoRAField(
+                lora=self.lora,
+                weight=self.weight,
+            )
+        )
+
+        return FluxLoRALoaderOutput(transformer=transformer)

invokeai/app/invocations/flux_text_encoder.py

@@ -0,0 +1,92 @@
+from typing import Literal
+
+import torch
+from transformers import CLIPTextModel, CLIPTokenizer, T5EncoderModel, T5Tokenizer
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
+from invokeai.app.invocations.fields import FieldDescriptions, Input, InputField
+from invokeai.app.invocations.model import CLIPField, T5EncoderField
+from invokeai.app.invocations.primitives import FluxConditioningOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.modules.conditioner import HFEncoder
+from invokeai.backend.stable_diffusion.diffusion.conditioning_data import ConditioningFieldData, FLUXConditioningInfo
+
+
+@invocation(
+    "flux_text_encoder",
+    title="FLUX Text Encoding",
+    tags=["prompt", "conditioning", "flux"],
+    category="conditioning",
+    version="1.0.0",
+    classification=Classification.Prototype,
+)
+class FluxTextEncoderInvocation(BaseInvocation):
+    """Encodes and preps a prompt for a flux image."""
+
+    clip: CLIPField = InputField(
+        title="CLIP",
+        description=FieldDescriptions.clip,
+        input=Input.Connection,
+    )
+    t5_encoder: T5EncoderField = InputField(
+        title="T5Encoder",
+        description=FieldDescriptions.t5_encoder,
+        input=Input.Connection,
+    )
+    t5_max_seq_len: Literal[256, 512] = InputField(
+        description="Max sequence length for the T5 encoder. Expected to be 256 for FLUX schnell models and 512 for FLUX dev models."
+    )
+    prompt: str = InputField(description="Text prompt to encode.")
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> FluxConditioningOutput:
+        # Note: The T5 and CLIP encoding are done in separate functions to ensure that all model references are locally
+        # scoped. This ensures that the T5 model can be freed and gc'd before loading the CLIP model (if necessary).
+        t5_embeddings = self._t5_encode(context)
+        clip_embeddings = self._clip_encode(context)
+        conditioning_data = ConditioningFieldData(
+            conditionings=[FLUXConditioningInfo(clip_embeds=clip_embeddings, t5_embeds=t5_embeddings)]
+        )
+
+        conditioning_name = context.conditioning.save(conditioning_data)
+        return FluxConditioningOutput.build(conditioning_name)
+
+    def _t5_encode(self, context: InvocationContext) -> torch.Tensor:
+        t5_tokenizer_info = context.models.load(self.t5_encoder.tokenizer)
+        t5_text_encoder_info = context.models.load(self.t5_encoder.text_encoder)
+
+        prompt = [self.prompt]
+
+        with (
+            t5_text_encoder_info as t5_text_encoder,
+            t5_tokenizer_info as t5_tokenizer,
+        ):
+            assert isinstance(t5_text_encoder, T5EncoderModel)
+            assert isinstance(t5_tokenizer, T5Tokenizer)
+
+            t5_encoder = HFEncoder(t5_text_encoder, t5_tokenizer, False, self.t5_max_seq_len)
+
+            prompt_embeds = t5_encoder(prompt)
+
+        assert isinstance(prompt_embeds, torch.Tensor)
+        return prompt_embeds
+
+    def _clip_encode(self, context: InvocationContext) -> torch.Tensor:
+        clip_tokenizer_info = context.models.load(self.clip.tokenizer)
+        clip_text_encoder_info = context.models.load(self.clip.text_encoder)
+
+        prompt = [self.prompt]
+
+        with (
+            clip_text_encoder_info as clip_text_encoder,
+            clip_tokenizer_info as clip_tokenizer,
+        ):
+            assert isinstance(clip_text_encoder, CLIPTextModel)
+            assert isinstance(clip_tokenizer, CLIPTokenizer)
+
+            clip_encoder = HFEncoder(clip_text_encoder, clip_tokenizer, True, 77)
+
+            pooled_prompt_embeds = clip_encoder(prompt)
+
+        assert isinstance(pooled_prompt_embeds, torch.Tensor)
+        return pooled_prompt_embeds

invokeai/app/invocations/flux_vae_decode.py

@@ -0,0 +1,60 @@
+import torch
+from einops import rearrange
+from PIL import Image
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    Input,
+    InputField,
+    LatentsField,
+    WithBoard,
+    WithMetadata,
+)
+from invokeai.app.invocations.model import VAEField
+from invokeai.app.invocations.primitives import ImageOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.modules.autoencoder import AutoEncoder
+from invokeai.backend.model_manager.load.load_base import LoadedModel
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "flux_vae_decode",
+    title="FLUX Latents to Image",
+    tags=["latents", "image", "vae", "l2i", "flux"],
+    category="latents",
+    version="1.0.0",
+)
+class FluxVaeDecodeInvocation(BaseInvocation, WithMetadata, WithBoard):
+    """Generates an image from latents."""
+
+    latents: LatentsField = InputField(
+        description=FieldDescriptions.latents,
+        input=Input.Connection,
+    )
+    vae: VAEField = InputField(
+        description=FieldDescriptions.vae,
+        input=Input.Connection,
+    )
+
+    def _vae_decode(self, vae_info: LoadedModel, latents: torch.Tensor) -> Image.Image:
+        with vae_info as vae:
+            assert isinstance(vae, AutoEncoder)
+            latents = latents.to(device=TorchDevice.choose_torch_device(), dtype=TorchDevice.choose_torch_dtype())
+            img = vae.decode(latents)
+
+        img = img.clamp(-1, 1)
+        img = rearrange(img[0], "c h w -> h w c")  # noqa: F821
+        img_pil = Image.fromarray((127.5 * (img + 1.0)).byte().cpu().numpy())
+        return img_pil
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> ImageOutput:
+        latents = context.tensors.load(self.latents.latents_name)
+        vae_info = context.models.load(self.vae.vae)
+        image = self._vae_decode(vae_info=vae_info, latents=latents)
+
+        TorchDevice.empty_cache()
+        image_dto = context.images.save(image=image)
+        return ImageOutput.build(image_dto)

invokeai/app/invocations/flux_vae_encode.py

@@ -0,0 +1,67 @@
+import einops
+import torch
+
+from invokeai.app.invocations.baseinvocation import BaseInvocation, invocation
+from invokeai.app.invocations.fields import (
+    FieldDescriptions,
+    ImageField,
+    Input,
+    InputField,
+)
+from invokeai.app.invocations.model import VAEField
+from invokeai.app.invocations.primitives import LatentsOutput
+from invokeai.app.services.shared.invocation_context import InvocationContext
+from invokeai.backend.flux.modules.autoencoder import AutoEncoder
+from invokeai.backend.model_manager import LoadedModel
+from invokeai.backend.stable_diffusion.diffusers_pipeline import image_resized_to_grid_as_tensor
+from invokeai.backend.util.devices import TorchDevice
+
+
+@invocation(
+    "flux_vae_encode",
+    title="FLUX Image to Latents",
+    tags=["latents", "image", "vae", "i2l", "flux"],
+    category="latents",
+    version="1.0.0",
+)
+class FluxVaeEncodeInvocation(BaseInvocation):
+    """Encodes an image into latents."""
+
+    image: ImageField = InputField(
+        description="The image to encode.",
+    )
+    vae: VAEField = InputField(
+        description=FieldDescriptions.vae,
+        input=Input.Connection,
+    )
+
+    @staticmethod
+    def vae_encode(vae_info: LoadedModel, image_tensor: torch.Tensor) -> torch.Tensor:
+        # TODO(ryand): Expose seed parameter at the invocation level.
+        # TODO(ryand): Write a util function for generating random tensors that is consistent across devices / dtypes.
+        # There's a starting point in get_noise(...), but it needs to be extracted and generalized. This function
+        # should be used for VAE encode sampling.
+        generator = torch.Generator(device=TorchDevice.choose_torch_device()).manual_seed(0)
+        with vae_info as vae:
+            assert isinstance(vae, AutoEncoder)
+            image_tensor = image_tensor.to(
+                device=TorchDevice.choose_torch_device(), dtype=TorchDevice.choose_torch_dtype()
+            )
+            latents = vae.encode(image_tensor, sample=True, generator=generator)
+            return latents
+
+    @torch.no_grad()
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        image = context.images.get_pil(self.image.image_name)
+
+        vae_info = context.models.load(self.vae.vae)
+
+        image_tensor = image_resized_to_grid_as_tensor(image.convert("RGB"))
+        if image_tensor.dim() == 3:
+            image_tensor = einops.rearrange(image_tensor, "c h w -> 1 c h w")
+
+        latents = self.vae_encode(vae_info=vae_info, image_tensor=image_tensor)
+
+        latents = latents.to("cpu")
+        name = context.tensors.save(tensor=latents)
+        return LatentsOutput.build(latents_name=name, latents=latents, seed=None)

invokeai/app/invocations/image.py

@@ -6,19 +6,13 @@ import cv2
 import numpy
 from PIL import Image, ImageChops, ImageFilter, ImageOps
 
-from invokeai.app.invocations.baseinvocation import (
-    BaseInvocation,
-    Classification,
-    invocation,
-    invocation_output,
-)
+from invokeai.app.invocations.baseinvocation import BaseInvocation, Classification, invocation
 from invokeai.app.invocations.constants import IMAGE_MODES
 from invokeai.app.invocations.fields import (
     ColorField,
     FieldDescriptions,
     ImageField,
     InputField,
-    OutputField,
     WithBoard,
     WithMetadata,
 )
@@ -1013,62 +1007,3 @@ class MaskFromIDInvocation(BaseInvocation, WithMetadata, WithBoard):
         image_dto = context.images.save(image=mask, image_category=ImageCategory.MASK)
 
         return ImageOutput.build(image_dto)
-
-
-@invocation_output("canvas_v2_mask_and_crop_output")
-class CanvasV2MaskAndCropOutput(ImageOutput):
-    offset_x: int = OutputField(description="The x offset of the image, after cropping")
-    offset_y: int = OutputField(description="The y offset of the image, after cropping")
-
-
-@invocation(
-    "canvas_v2_mask_and_crop",
-    title="Canvas V2 Mask and Crop",
-    tags=["image", "mask", "id"],
-    category="image",
-    version="1.0.0",
-    classification=Classification.Prototype,
-)
-class CanvasV2MaskAndCropInvocation(BaseInvocation, WithMetadata, WithBoard):
-    """Handles Canvas V2 image output masking and cropping"""
-
-    source_image: ImageField | None = InputField(
-        default=None,
-        description="The source image onto which the masked generated image is pasted. If omitted, the masked generated image is returned with transparency.",
-    )
-    generated_image: ImageField = InputField(description="The image to apply the mask to")
-    mask: ImageField = InputField(description="The mask to apply")
-    mask_blur: int = InputField(default=0, ge=0, description="The amount to blur the mask by")
-
-    def _prepare_mask(self, mask: Image.Image) -> Image.Image:
-        mask_array = numpy.array(mask)
-        kernel = numpy.ones((self.mask_blur, self.mask_blur), numpy.uint8)
-        dilated_mask_array = cv2.erode(mask_array, kernel, iterations=3)
-        dilated_mask = Image.fromarray(dilated_mask_array)
-        if self.mask_blur > 0:
-            mask = dilated_mask.filter(ImageFilter.GaussianBlur(self.mask_blur))
-        return ImageOps.invert(mask.convert("L"))
-
-    def invoke(self, context: InvocationContext) -> CanvasV2MaskAndCropOutput:
-        mask = self._prepare_mask(context.images.get_pil(self.mask.image_name))
-
-        if self.source_image:
-            generated_image = context.images.get_pil(self.generated_image.image_name)
-            source_image = context.images.get_pil(self.source_image.image_name)
-            source_image.paste(generated_image, (0, 0), mask)
-            image_dto = context.images.save(image=source_image)
-        else:
-            generated_image = context.images.get_pil(self.generated_image.image_name)
-            generated_image.putalpha(mask)
-            image_dto = context.images.save(image=generated_image)
-
-        # bbox = image.getbbox()
-        # image = image.crop(bbox)
-
-        return CanvasV2MaskAndCropOutput(
-            image=ImageField(image_name=image_dto.image_name),
-            offset_x=0,
-            offset_y=0,
-            width=image_dto.width,
-            height=image_dto.height,
-        )

invokeai/app/invocations/mask.py

@@ -126,7 +126,7 @@ class ImageMaskToTensorInvocation(BaseInvocation, WithMetadata):
     title="Tensor Mask to Image",
     tags=["mask"],
     category="mask",
-    version="1.0.0",
+    version="1.1.0",
 )
 class MaskTensorToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
     """Convert a mask tensor to an image."""
@@ -135,6 +135,11 @@ class MaskTensorToImageInvocation(BaseInvocation, WithMetadata, WithBoard):
 
     def invoke(self, context: InvocationContext) -> ImageOutput:
         mask = context.tensors.load(self.mask.tensor_name)
+
+        # Squeeze the channel dimension if it exists.
+        if mask.dim() == 3:
+            mask = mask.squeeze(0)
+
         # Ensure that the mask is binary.
         if mask.dtype != torch.bool:
             mask = mask > 0.5

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
 
@@ -13,7 +13,14 @@ 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.model_manager.config import AnyModelConfig, BaseModelType, ModelType, SubModelType
+from invokeai.backend.flux.util import max_seq_lengths
+from invokeai.backend.model_manager.config import (
+    AnyModelConfig,
+    BaseModelType,
+    CheckpointConfigBase,
+    ModelType,
+    SubModelType,
+)
 
 
 class ModelIdentifierField(BaseModel):
@@ -60,6 +67,16 @@ class CLIPField(BaseModel):
     loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
 
 
+class TransformerField(BaseModel):
+    transformer: ModelIdentifierField = Field(description="Info to load Transformer submodel")
+    loras: List[LoRAField] = Field(description="LoRAs to apply on model loading")
+
+
+class T5EncoderField(BaseModel):
+    tokenizer: ModelIdentifierField = Field(description="Info to load tokenizer submodel")
+    text_encoder: ModelIdentifierField = Field(description="Info to load text_encoder submodel")
+
+
 class VAEField(BaseModel):
     vae: ModelIdentifierField = Field(description="Info to load vae submodel")
     seamless_axes: List[str] = Field(default_factory=list, description='Axes("x" and "y") to which apply seamless')
@@ -122,6 +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/primitives.py

@@ -12,6 +12,7 @@ from invokeai.app.invocations.fields import (
     ConditioningField,
     DenoiseMaskField,
     FieldDescriptions,
+    FluxConditioningField,
     ImageField,
     Input,
     InputField,
@@ -414,6 +415,17 @@ class MaskOutput(BaseInvocationOutput):
     height: int = OutputField(description="The height of the mask in pixels.")
 
 
+@invocation_output("flux_conditioning_output")
+class FluxConditioningOutput(BaseInvocationOutput):
+    """Base class for nodes that output a single conditioning tensor"""
+
+    conditioning: FluxConditioningField = OutputField(description=FieldDescriptions.cond)
+
+    @classmethod
+    def build(cls, conditioning_name: str) -> "FluxConditioningOutput":
+        return cls(conditioning=FluxConditioningField(conditioning_name=conditioning_name))
+
+
 @invocation_output("conditioning_output")
 class ConditioningOutput(BaseInvocationOutput):
     """Base class for nodes that output a single conditioning tensor"""

invokeai/app/invocations/tiled_multi_diffusion_denoise_latents.py

@@ -22,8 +22,8 @@ from invokeai.app.invocations.fields import (
 from invokeai.app.invocations.model import UNetField
 from invokeai.app.invocations.primitives import LatentsOutput
 from invokeai.app.services.shared.invocation_context import InvocationContext
-from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_patcher import ModelPatcher
+from invokeai.backend.peft.lora import LoRAModelRaw
 from invokeai.backend.stable_diffusion.diffusers_pipeline import ControlNetData, PipelineIntermediateState
 from invokeai.backend.stable_diffusion.multi_diffusion_pipeline import (
     MultiDiffusionPipeline,

invokeai/app/services/events/events_common.py

@@ -88,7 +88,6 @@ class QueueItemEventBase(QueueEventBase):
 
     item_id: int = Field(description="The ID of the queue item")
     batch_id: str = Field(description="The ID of the queue batch")
-    origin: str | None = Field(default=None, description="The origin of the batch")
 
 
 class InvocationEventBase(QueueItemEventBase):
@@ -96,6 +95,8 @@ class InvocationEventBase(QueueItemEventBase):
 
     session_id: str = Field(description="The ID of the session (aka graph execution state)")
     queue_id: str = Field(description="The ID of the queue")
+    item_id: int = Field(description="The ID of the queue item")
+    batch_id: str = Field(description="The ID of the queue batch")
     session_id: str = Field(description="The ID of the session (aka graph execution state)")
     invocation: AnyInvocation = Field(description="The ID of the invocation")
     invocation_source_id: str = Field(description="The ID of the prepared invocation's source node")
@@ -113,7 +114,6 @@ class InvocationStartedEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
-            origin=queue_item.origin,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -147,7 +147,6 @@ class InvocationDenoiseProgressEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
-            origin=queue_item.origin,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -185,7 +184,6 @@ class InvocationCompleteEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
-            origin=queue_item.origin,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -218,7 +216,6 @@ class InvocationErrorEvent(InvocationEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
-            origin=queue_item.origin,
             session_id=queue_item.session_id,
             invocation=invocation,
             invocation_source_id=queue_item.session.prepared_source_mapping[invocation.id],
@@ -256,7 +253,6 @@ class QueueItemStatusChangedEvent(QueueItemEventBase):
             queue_id=queue_item.queue_id,
             item_id=queue_item.item_id,
             batch_id=queue_item.batch_id,
-            origin=queue_item.origin,
             session_id=queue_item.session_id,
             status=queue_item.status,
             error_type=queue_item.error_type,
@@ -283,14 +279,12 @@ class BatchEnqueuedEvent(QueueEventBase):
         description="The number of invocations initially requested to be enqueued (may be less than enqueued if queue was full)"
     )
     priority: int = Field(description="The priority of the batch")
-    origin: str | None = Field(default=None, description="The origin of the batch")
 
     @classmethod
     def build(cls, enqueue_result: EnqueueBatchResult) -> "BatchEnqueuedEvent":
         return cls(
             queue_id=enqueue_result.queue_id,
             batch_id=enqueue_result.batch.batch_id,
-            origin=enqueue_result.batch.origin,
             enqueued=enqueue_result.enqueued,
             requested=enqueue_result.requested,
             priority=enqueue_result.priority,

invokeai/app/services/model_install/model_install_common.py

@@ -103,7 +103,7 @@ class HFModelSource(StringLikeSource):
         if self.variant:
             base += f":{self.variant or ''}"
         if self.subfolder:
-            base += f":{self.subfolder}"
+            base += f"::{self.subfolder.as_posix()}"
         return base
 
 

invokeai/app/services/model_install/model_install_default.py

@@ -783,8 +783,9 @@ class ModelInstallService(ModelInstallServiceBase):
         # So what we do is to synthesize a folder named "sdxl-turbo_vae" here.
         if subfolder:
             top = Path(remote_files[0].path.parts[0])  # e.g. "sdxl-turbo/"
-            path_to_remove = top / subfolder.parts[-1]  # sdxl-turbo/vae/
-            path_to_add = Path(f"{top}_{subfolder}")
+            path_to_remove = top / subfolder  # sdxl-turbo/vae/
+            subfolder_rename = subfolder.name.replace("/", "_").replace("\\", "_")
+            path_to_add = Path(f"{top}_{subfolder_rename}")
         else:
             path_to_remove = Path(".")
             path_to_add = Path(".")

invokeai/app/services/model_records/model_records_base.py

@@ -77,6 +77,7 @@ class ModelRecordChanges(BaseModelExcludeNull):
     type: Optional[ModelType] = Field(description="Type of model", default=None)
     key: Optional[str] = Field(description="Database ID for this model", default=None)
     hash: Optional[str] = Field(description="hash of model file", default=None)
+    format: Optional[str] = Field(description="format of model file", default=None)
     trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
     default_settings: Optional[MainModelDefaultSettings | ControlAdapterDefaultSettings] = Field(
         description="Default settings for this model", default=None

invokeai/app/services/session_queue/session_queue_base.py

@@ -6,7 +6,6 @@ from invokeai.app.services.session_queue.session_queue_common import (
     Batch,
     BatchStatus,
     CancelByBatchIDsResult,
-    CancelByOriginResult,
     CancelByQueueIDResult,
     ClearResult,
     EnqueueBatchResult,
@@ -96,11 +95,6 @@ class SessionQueueBase(ABC):
         """Cancels all queue items with matching batch IDs"""
         pass
 
-    @abstractmethod
-    def cancel_by_origin(self, queue_id: str, origin: str) -> CancelByOriginResult:
-        """Cancels all queue items with the given batch origin"""
-        pass
-
     @abstractmethod
     def cancel_by_queue_id(self, queue_id: str) -> CancelByQueueIDResult:
         """Cancels all queue items with matching queue ID"""

invokeai/app/services/session_queue/session_queue_common.py

@@ -77,7 +77,6 @@ BatchDataCollection: TypeAlias = list[list[BatchDatum]]
 
 class Batch(BaseModel):
     batch_id: str = Field(default_factory=uuid_string, description="The ID of the batch")
-    origin: str | None = Field(default=None, description="The origin of this batch.")
     data: Optional[BatchDataCollection] = Field(default=None, description="The batch data collection.")
     graph: Graph = Field(description="The graph to initialize the session with")
     workflow: Optional[WorkflowWithoutID] = Field(
@@ -196,7 +195,6 @@ class SessionQueueItemWithoutGraph(BaseModel):
     status: QUEUE_ITEM_STATUS = Field(default="pending", description="The status of this queue item")
     priority: int = Field(default=0, description="The priority of this queue item")
     batch_id: str = Field(description="The ID of the batch associated with this queue item")
-    origin: str | None = Field(default=None, description="The origin of this queue item. ")
     session_id: str = Field(
         description="The ID of the session associated with this queue item. The session doesn't exist in graph_executions until the queue item is executed."
     )
@@ -296,7 +294,6 @@ class SessionQueueStatus(BaseModel):
 class BatchStatus(BaseModel):
     queue_id: str = Field(..., description="The ID of the queue")
     batch_id: str = Field(..., description="The ID of the batch")
-    origin: str | None = Field(..., description="The origin of the batch")
     pending: int = Field(..., description="Number of queue items with status 'pending'")
     in_progress: int = Field(..., description="Number of queue items with status 'in_progress'")
     completed: int = Field(..., description="Number of queue items with status 'complete'")
@@ -331,12 +328,6 @@ class CancelByBatchIDsResult(BaseModel):
     canceled: int = Field(..., description="Number of queue items canceled")
 
 
-class CancelByOriginResult(BaseModel):
-    """Result of canceling by list of batch ids"""
-
-    canceled: int = Field(..., description="Number of queue items canceled")
-
-
 class CancelByQueueIDResult(CancelByBatchIDsResult):
     """Result of canceling by queue id"""
 
@@ -442,7 +433,6 @@ class SessionQueueValueToInsert(NamedTuple):
     field_values: Optional[str]  # field_values json
     priority: int  # priority
     workflow: Optional[str]  # workflow json
-    origin: str | None
 
 
 ValuesToInsert: TypeAlias = list[SessionQueueValueToInsert]
@@ -463,7 +453,6 @@ def prepare_values_to_insert(queue_id: str, batch: Batch, priority: int, max_new
                 json.dumps(field_values, default=to_jsonable_python) if field_values else None,  # field_values (json)
                 priority,  # priority
                 json.dumps(workflow, default=to_jsonable_python) if workflow else None,  # workflow (json)
-                batch.origin,  # origin
             )
         )
     return values_to_insert

invokeai/app/services/session_queue/session_queue_sqlite.py

@@ -10,7 +10,6 @@ from invokeai.app.services.session_queue.session_queue_common import (
     Batch,
     BatchStatus,
     CancelByBatchIDsResult,
-    CancelByOriginResult,
     CancelByQueueIDResult,
     ClearResult,
     EnqueueBatchResult,
@@ -128,8 +127,8 @@ class SqliteSessionQueue(SessionQueueBase):
 
             self.__cursor.executemany(
                 """--sql
-                INSERT INTO session_queue (queue_id, session, session_id, batch_id, field_values, priority, workflow, origin)
-                VALUES (?, ?, ?, ?, ?, ?, ?, ?)
+                INSERT INTO session_queue (queue_id, session, session_id, batch_id, field_values, priority, workflow)
+                VALUES (?, ?, ?, ?, ?, ?, ?)
                 """,
                 values_to_insert,
             )
@@ -418,7 +417,11 @@ class SqliteSessionQueue(SessionQueueBase):
             )
             self.__conn.commit()
             if current_queue_item is not None and current_queue_item.batch_id in batch_ids:
-                self._set_queue_item_status(current_queue_item.item_id, "canceled")
+                batch_status = self.get_batch_status(queue_id=queue_id, batch_id=current_queue_item.batch_id)
+                queue_status = self.get_queue_status(queue_id=queue_id)
+                self.__invoker.services.events.emit_queue_item_status_changed(
+                    current_queue_item, batch_status, queue_status
+                )
         except Exception:
             self.__conn.rollback()
             raise
@@ -426,46 +429,6 @@ class SqliteSessionQueue(SessionQueueBase):
             self.__lock.release()
         return CancelByBatchIDsResult(canceled=count)
 
-    def cancel_by_origin(self, queue_id: str, origin: str) -> CancelByOriginResult:
-        try:
-            current_queue_item = self.get_current(queue_id)
-            self.__lock.acquire()
-            where = """--sql
-                WHERE
-                  queue_id == ?
-                  AND origin == ?
-                  AND status != 'canceled'
-                  AND status != 'completed'
-                  AND status != 'failed'
-                """
-            params = (queue_id, origin)
-            self.__cursor.execute(
-                f"""--sql
-                SELECT COUNT(*)
-                FROM session_queue
-                {where};
-                """,
-                params,
-            )
-            count = self.__cursor.fetchone()[0]
-            self.__cursor.execute(
-                f"""--sql
-                UPDATE session_queue
-                SET status = 'canceled'
-                {where};
-                """,
-                params,
-            )
-            self.__conn.commit()
-            if current_queue_item is not None and current_queue_item.origin == origin:
-                self._set_queue_item_status(current_queue_item.item_id, "canceled")
-        except Exception:
-            self.__conn.rollback()
-            raise
-        finally:
-            self.__lock.release()
-        return CancelByOriginResult(canceled=count)
-
     def cancel_by_queue_id(self, queue_id: str) -> CancelByQueueIDResult:
         try:
             current_queue_item = self.get_current(queue_id)
@@ -578,8 +541,7 @@ class SqliteSessionQueue(SessionQueueBase):
                     started_at,
                     session_id,
                     batch_id,
-                    queue_id,
-                    origin
+                    queue_id
                 FROM session_queue
                 WHERE queue_id = ?
             """
@@ -659,7 +621,7 @@ class SqliteSessionQueue(SessionQueueBase):
             self.__lock.acquire()
             self.__cursor.execute(
                 """--sql
-                SELECT status, count(*), origin
+                SELECT status, count(*)
                 FROM session_queue
                 WHERE
                   queue_id = ?
@@ -671,7 +633,6 @@ class SqliteSessionQueue(SessionQueueBase):
             result = cast(list[sqlite3.Row], self.__cursor.fetchall())
             total = sum(row[1] for row in result)
             counts: dict[str, int] = {row[0]: row[1] for row in result}
-            origin = result[0]["origin"] if result else None
         except Exception:
             self.__conn.rollback()
             raise
@@ -680,7 +641,6 @@ class SqliteSessionQueue(SessionQueueBase):
 
         return BatchStatus(
             batch_id=batch_id,
-            origin=origin,
             queue_id=queue_id,
             pending=counts.get("pending", 0),
             in_progress=counts.get("in_progress", 0),

invokeai/app/services/shared/sqlite/sqlite_util.py

@@ -17,7 +17,6 @@ from invokeai.app.services.shared.sqlite_migrator.migrations.migration_11 import
 from invokeai.app.services.shared.sqlite_migrator.migrations.migration_12 import build_migration_12
 from invokeai.app.services.shared.sqlite_migrator.migrations.migration_13 import build_migration_13
 from invokeai.app.services.shared.sqlite_migrator.migrations.migration_14 import build_migration_14
-from invokeai.app.services.shared.sqlite_migrator.migrations.migration_15 import build_migration_15
 from invokeai.app.services.shared.sqlite_migrator.sqlite_migrator_impl import SqliteMigrator
 
 
@@ -52,7 +51,6 @@ def init_db(config: InvokeAIAppConfig, logger: Logger, image_files: ImageFileSto
     migrator.register_migration(build_migration_12(app_config=config))
     migrator.register_migration(build_migration_13())
     migrator.register_migration(build_migration_14())
-    migrator.register_migration(build_migration_15())
     migrator.run_migrations()
 
     return db

invokeai/app/services/shared/sqlite_migrator/migrations/migration_15.py

@@ -1,31 +0,0 @@
-import sqlite3
-
-from invokeai.app.services.shared.sqlite_migrator.sqlite_migrator_common import Migration
-
-
-class Migration15Callback:
-    def __call__(self, cursor: sqlite3.Cursor) -> None:
-        self._add_origin_col(cursor)
-
-    def _add_origin_col(self, cursor: sqlite3.Cursor) -> None:
-        """
-        - Adds `origin` column to the session queue table.
-        """
-
-        cursor.execute("ALTER TABLE session_queue ADD COLUMN origin TEXT;")
-
-
-def build_migration_15() -> Migration:
-    """
-    Build the migration from database version 14 to 15.
-
-    This migration does the following:
-        - Adds `origin` column to the session queue table.
-    """
-    migration_15 = Migration(
-        from_version=14,
-        to_version=15,
-        callback=Migration15Callback(),
-    )
-
-    return migration_15

invokeai/app/services/workflow_records/default_workflows/FLUX Image to Image.json

@@ -0,0 +1,407 @@
+{
+  "name": "FLUX Image to Image",
+  "author": "InvokeAI",
+  "description": "A simple image-to-image workflow using a FLUX dev model. ",
+  "version": "1.0.4",
+  "contact": "",
+  "tags": "image2image, flux, image-to-image",
+  "notes": "Prerequisite model downloads: T5 Encoder, CLIP-L Encoder, and FLUX VAE. Quantized and un-quantized versions can be found in the starter models tab within your Model Manager. We recommend using FLUX dev models for image-to-image workflows. The image-to-image performance with FLUX schnell models is poor.",
+  "exposedFields": [
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "t5_encoder_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "clip_embed_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "vae_model"
+    },
+    {
+      "nodeId": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "fieldName": "denoising_start"
+    },
+    {
+      "nodeId": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "fieldName": "prompt"
+    },
+    {
+      "nodeId": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "fieldName": "num_steps"
+    }
+  ],
+  "meta": {
+    "version": "3.0.0",
+    "category": "default"
+  },
+  "nodes": [
+    {
+      "id": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "type": "invocation",
+      "data": {
+        "id": "2981a67c-480f-4237-9384-26b68dbf912b",
+        "type": "flux_vae_encode",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "image": {
+            "name": "image",
+            "label": "",
+            "value": {
+              "image_name": "8a5c62aa-9335-45d2-9c71-89af9fc1f8d4.png"
+            }
+          },
+          "vae": {
+            "name": "vae",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 732.7680166609682,
+        "y": -24.37398171806909
+      }
+    },
+    {
+      "id": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "type": "invocation",
+      "data": {
+        "id": "ace0258f-67d7-4eee-a218-6fff27065214",
+        "type": "flux_denoise",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "denoise_mask": {
+            "name": "denoise_mask",
+            "label": ""
+          },
+          "denoising_start": {
+            "name": "denoising_start",
+            "label": "",
+            "value": 0.04
+          },
+          "denoising_end": {
+            "name": "denoising_end",
+            "label": "",
+            "value": 1
+          },
+          "transformer": {
+            "name": "transformer",
+            "label": ""
+          },
+          "positive_text_conditioning": {
+            "name": "positive_text_conditioning",
+            "label": ""
+          },
+          "width": {
+            "name": "width",
+            "label": "",
+            "value": 1024
+          },
+          "height": {
+            "name": "height",
+            "label": "",
+            "value": 1024
+          },
+          "num_steps": {
+            "name": "num_steps",
+            "label": "Steps (Recommend 30 for Dev, 4 for Schnell)",
+            "value": 30
+          },
+          "guidance": {
+            "name": "guidance",
+            "label": "",
+            "value": 4
+          },
+          "seed": {
+            "name": "seed",
+            "label": "",
+            "value": 0
+          }
+        }
+      },
+      "position": {
+        "x": 1182.8836633018684,
+        "y": -251.38882958913183
+      }
+    },
+    {
+      "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "type": "invocation",
+      "data": {
+        "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+        "type": "flux_vae_decode",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": false,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "vae": {
+            "name": "vae",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 1575.5797431839133,
+        "y": -209.00150975507415
+      }
+    },
+    {
+      "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "type": "invocation",
+      "data": {
+        "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+        "type": "flux_model_loader",
+        "version": "1.0.4",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "model": {
+            "name": "model",
+            "label": "Model (dev variant recommended for Image-to-Image)"
+          },
+          "t5_encoder_model": {
+            "name": "t5_encoder_model",
+            "label": ""
+          },
+          "clip_embed_model": {
+            "name": "clip_embed_model",
+            "label": "",
+            "value": {
+              "key": "fa23a584-b623-415d-832a-21b5098ff1a1",
+              "hash": "blake3:17c19f0ef941c3b7609a9c94a659ca5364de0be364a91d4179f0e39ba17c3b70",
+              "name": "clip-vit-large-patch14",
+              "base": "any",
+              "type": "clip_embed"
+            }
+          },
+          "vae_model": {
+            "name": "vae_model",
+            "label": "",
+            "value": {
+              "key": "74fc82ba-c0a8-479d-a890-2126f82da758",
+              "hash": "blake3:ce21cb76364aa6e2421311cf4a4b5eb052a76c4f1cd207b50703d8978198a068",
+              "name": "FLUX.1-schnell_ae",
+              "base": "flux",
+              "type": "vae"
+            }
+          }
+        }
+      },
+      "position": {
+        "x": 328.1809894659957,
+        "y": -90.2241133566946
+      }
+    },
+    {
+      "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "type": "invocation",
+      "data": {
+        "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+        "type": "flux_text_encoder",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "clip": {
+            "name": "clip",
+            "label": ""
+          },
+          "t5_encoder": {
+            "name": "t5_encoder",
+            "label": ""
+          },
+          "t5_max_seq_len": {
+            "name": "t5_max_seq_len",
+            "label": "T5 Max Seq Len",
+            "value": 256
+          },
+          "prompt": {
+            "name": "prompt",
+            "label": "",
+            "value": "a cat wearing a birthday hat"
+          }
+        }
+      },
+      "position": {
+        "x": 745.8823365057267,
+        "y": -299.60249175851914
+      }
+    },
+    {
+      "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "type": "invocation",
+      "data": {
+        "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+        "type": "rand_int",
+        "version": "1.0.1",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "low": {
+            "name": "low",
+            "label": "",
+            "value": 0
+          },
+          "high": {
+            "name": "high",
+            "label": "",
+            "value": 2147483647
+          }
+        }
+      },
+      "position": {
+        "x": 725.834098928012,
+        "y": 496.2710031089931
+      }
+    }
+  ],
+  "edges": [
+    {
+      "id": "reactflow__edge-2981a67c-480f-4237-9384-26b68dbf912bheight-ace0258f-67d7-4eee-a218-6fff27065214height",
+      "type": "default",
+      "source": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "height",
+      "targetHandle": "height"
+    },
+    {
+      "id": "reactflow__edge-2981a67c-480f-4237-9384-26b68dbf912bwidth-ace0258f-67d7-4eee-a218-6fff27065214width",
+      "type": "default",
+      "source": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "width",
+      "targetHandle": "width"
+    },
+    {
+      "id": "reactflow__edge-2981a67c-480f-4237-9384-26b68dbf912blatents-ace0258f-67d7-4eee-a218-6fff27065214latents",
+      "type": "default",
+      "source": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "latents",
+      "targetHandle": "latents"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90vae-2981a67c-480f-4237-9384-26b68dbf912bvae",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "2981a67c-480f-4237-9384-26b68dbf912b",
+      "sourceHandle": "vae",
+      "targetHandle": "vae"
+    },
+    {
+      "id": "reactflow__edge-ace0258f-67d7-4eee-a218-6fff27065214latents-7e5172eb-48c1-44db-a770-8fd83e1435d1latents",
+      "type": "default",
+      "source": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "latents",
+      "targetHandle": "latents"
+    },
+    {
+      "id": "reactflow__edge-4754c534-a5f3-4ad0-9382-7887985e668cvalue-ace0258f-67d7-4eee-a218-6fff27065214seed",
+      "type": "default",
+      "source": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "value",
+      "targetHandle": "seed"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90transformer-ace0258f-67d7-4eee-a218-6fff27065214transformer",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "transformer",
+      "targetHandle": "transformer"
+    },
+    {
+      "id": "reactflow__edge-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cconditioning-ace0258f-67d7-4eee-a218-6fff27065214positive_text_conditioning",
+      "type": "default",
+      "source": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "target": "ace0258f-67d7-4eee-a218-6fff27065214",
+      "sourceHandle": "conditioning",
+      "targetHandle": "positive_text_conditioning"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90vae-7e5172eb-48c1-44db-a770-8fd83e1435d1vae",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "vae",
+      "targetHandle": "vae"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90max_seq_len-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_max_seq_len",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "max_seq_len",
+      "targetHandle": "t5_max_seq_len"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90t5_encoder-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_encoder",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "t5_encoder",
+      "targetHandle": "t5_encoder"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90clip-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cclip",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "clip",
+      "targetHandle": "clip"
+    }
+  ]
+}

invokeai/app/services/workflow_records/default_workflows/Flux Text to Image.json

@@ -0,0 +1,326 @@
+{
+  "name": "FLUX Text to Image",
+  "author": "InvokeAI",
+  "description": "A simple text-to-image workflow using FLUX dev or schnell models.",
+  "version": "1.0.4",
+  "contact": "",
+  "tags": "text2image, flux",
+  "notes": "Prerequisite model downloads: T5 Encoder, CLIP-L Encoder, and FLUX VAE. Quantized and un-quantized versions can be found in the starter models tab within your Model Manager. We recommend 4 steps for FLUX schnell models and 30 steps for FLUX dev models.",
+  "exposedFields": [
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "t5_encoder_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "clip_embed_model"
+    },
+    {
+      "nodeId": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "fieldName": "vae_model"
+    },
+    {
+      "nodeId": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "fieldName": "prompt"
+    },
+    {
+      "nodeId": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "fieldName": "num_steps"
+    }
+  ],
+  "meta": {
+    "version": "3.0.0",
+    "category": "default"
+  },
+  "nodes": [
+    {
+      "id": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "type": "invocation",
+      "data": {
+        "id": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+        "type": "flux_denoise",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "denoise_mask": {
+            "name": "denoise_mask",
+            "label": ""
+          },
+          "denoising_start": {
+            "name": "denoising_start",
+            "label": "",
+            "value": 0
+          },
+          "denoising_end": {
+            "name": "denoising_end",
+            "label": "",
+            "value": 1
+          },
+          "transformer": {
+            "name": "transformer",
+            "label": ""
+          },
+          "positive_text_conditioning": {
+            "name": "positive_text_conditioning",
+            "label": ""
+          },
+          "width": {
+            "name": "width",
+            "label": "",
+            "value": 1024
+          },
+          "height": {
+            "name": "height",
+            "label": "",
+            "value": 1024
+          },
+          "num_steps": {
+            "name": "num_steps",
+            "label": "Steps (Recommend 30 for Dev, 4 for Schnell)",
+            "value": 30
+          },
+          "guidance": {
+            "name": "guidance",
+            "label": "",
+            "value": 4
+          },
+          "seed": {
+            "name": "seed",
+            "label": "",
+            "value": 0
+          }
+        }
+      },
+      "position": {
+        "x": 1186.1868226120378,
+        "y": -214.9459927686657
+      }
+    },
+    {
+      "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "type": "invocation",
+      "data": {
+        "id": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+        "type": "flux_vae_decode",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": false,
+        "useCache": true,
+        "inputs": {
+          "board": {
+            "name": "board",
+            "label": ""
+          },
+          "metadata": {
+            "name": "metadata",
+            "label": ""
+          },
+          "latents": {
+            "name": "latents",
+            "label": ""
+          },
+          "vae": {
+            "name": "vae",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 1575.5797431839133,
+        "y": -209.00150975507415
+      }
+    },
+    {
+      "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "type": "invocation",
+      "data": {
+        "id": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+        "type": "flux_model_loader",
+        "version": "1.0.4",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "model": {
+            "name": "model",
+            "label": ""
+          },
+          "t5_encoder_model": {
+            "name": "t5_encoder_model",
+            "label": ""
+          },
+          "clip_embed_model": {
+            "name": "clip_embed_model",
+            "label": ""
+          },
+          "vae_model": {
+            "name": "vae_model",
+            "label": ""
+          }
+        }
+      },
+      "position": {
+        "x": 381.1882713063478,
+        "y": -95.89663532854017
+      }
+    },
+    {
+      "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "type": "invocation",
+      "data": {
+        "id": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+        "type": "flux_text_encoder",
+        "version": "1.0.0",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": true,
+        "inputs": {
+          "clip": {
+            "name": "clip",
+            "label": ""
+          },
+          "t5_encoder": {
+            "name": "t5_encoder",
+            "label": ""
+          },
+          "t5_max_seq_len": {
+            "name": "t5_max_seq_len",
+            "label": "T5 Max Seq Len",
+            "value": 256
+          },
+          "prompt": {
+            "name": "prompt",
+            "label": "",
+            "value": "a cat"
+          }
+        }
+      },
+      "position": {
+        "x": 778.4899149328337,
+        "y": -100.36469216659502
+      }
+    },
+    {
+      "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "type": "invocation",
+      "data": {
+        "id": "4754c534-a5f3-4ad0-9382-7887985e668c",
+        "type": "rand_int",
+        "version": "1.0.1",
+        "label": "",
+        "notes": "",
+        "isOpen": true,
+        "isIntermediate": true,
+        "useCache": false,
+        "inputs": {
+          "low": {
+            "name": "low",
+            "label": "",
+            "value": 0
+          },
+          "high": {
+            "name": "high",
+            "label": "",
+            "value": 2147483647
+          }
+        }
+      },
+      "position": {
+        "x": 800.9667463219505,
+        "y": 285.8297267547506
+      }
+    }
+  ],
+  "edges": [
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90transformer-4fe24f07-f906-4f55-ab2c-9beee56ef5bdtransformer",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "sourceHandle": "transformer",
+      "targetHandle": "transformer"
+    },
+    {
+      "id": "reactflow__edge-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cconditioning-4fe24f07-f906-4f55-ab2c-9beee56ef5bdpositive_text_conditioning",
+      "type": "default",
+      "source": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "target": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "sourceHandle": "conditioning",
+      "targetHandle": "positive_text_conditioning"
+    },
+    {
+      "id": "reactflow__edge-4754c534-a5f3-4ad0-9382-7887985e668cvalue-4fe24f07-f906-4f55-ab2c-9beee56ef5bdseed",
+      "type": "default",
+      "source": "4754c534-a5f3-4ad0-9382-7887985e668c",
+      "target": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "sourceHandle": "value",
+      "targetHandle": "seed"
+    },
+    {
+      "id": "reactflow__edge-4fe24f07-f906-4f55-ab2c-9beee56ef5bdlatents-7e5172eb-48c1-44db-a770-8fd83e1435d1latents",
+      "type": "default",
+      "source": "4fe24f07-f906-4f55-ab2c-9beee56ef5bd",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "latents",
+      "targetHandle": "latents"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90vae-7e5172eb-48c1-44db-a770-8fd83e1435d1vae",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "7e5172eb-48c1-44db-a770-8fd83e1435d1",
+      "sourceHandle": "vae",
+      "targetHandle": "vae"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90max_seq_len-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_max_seq_len",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "max_seq_len",
+      "targetHandle": "t5_max_seq_len"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90t5_encoder-01f674f8-b3d1-4df1-acac-6cb8e0bfb63ct5_encoder",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "t5_encoder",
+      "targetHandle": "t5_encoder"
+    },
+    {
+      "id": "reactflow__edge-f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90clip-01f674f8-b3d1-4df1-acac-6cb8e0bfb63cclip",
+      "type": "default",
+      "source": "f8d9d7c8-9ed7-4bd7-9e42-ab0e89bfac90",
+      "target": "01f674f8-b3d1-4df1-acac-6cb8e0bfb63c",
+      "sourceHandle": "clip",
+      "targetHandle": "clip"
+    }
+  ]
+}

invokeai/backend/flux/denoise.py

@@ -0,0 +1,45 @@
+from typing import Callable
+
+import torch
+from tqdm import tqdm
+
+from invokeai.backend.flux.inpaint_extension import InpaintExtension
+from invokeai.backend.flux.model import Flux
+
+
+def denoise(
+    model: Flux,
+    # model input
+    img: torch.Tensor,
+    img_ids: torch.Tensor,
+    txt: torch.Tensor,
+    txt_ids: torch.Tensor,
+    vec: torch.Tensor,
+    # sampling parameters
+    timesteps: list[float],
+    step_callback: Callable[[], None],
+    guidance: float,
+    inpaint_extension: InpaintExtension | None,
+):
+    # guidance_vec is ignored for schnell.
+    guidance_vec = torch.full((img.shape[0],), guidance, device=img.device, dtype=img.dtype)
+    for t_curr, t_prev in tqdm(list(zip(timesteps[:-1], timesteps[1:], strict=True))):
+        t_vec = torch.full((img.shape[0],), t_curr, dtype=img.dtype, device=img.device)
+        pred = model(
+            img=img,
+            img_ids=img_ids,
+            txt=txt,
+            txt_ids=txt_ids,
+            y=vec,
+            timesteps=t_vec,
+            guidance=guidance_vec,
+        )
+
+        img = img + (t_prev - t_curr) * pred
+
+        if inpaint_extension is not None:
+            img = inpaint_extension.merge_intermediate_latents_with_init_latents(img, t_prev)
+
+        step_callback()
+
+    return img

invokeai/backend/flux/inpaint_extension.py

@@ -0,0 +1,35 @@
+import torch
+
+
+class InpaintExtension:
+    """A class for managing inpainting with FLUX."""
+
+    def __init__(self, init_latents: torch.Tensor, inpaint_mask: torch.Tensor, noise: torch.Tensor):
+        """Initialize InpaintExtension.
+
+        Args:
+            init_latents (torch.Tensor): The initial latents (i.e. un-noised at timestep 0). In 'packed' format.
+            inpaint_mask (torch.Tensor): A mask specifying which elements to inpaint. Range [0, 1]. Values of 1 will be
+                re-generated. Values of 0 will remain unchanged. Values between 0 and 1 can be used to blend the
+                inpainted region with the background. In 'packed' format.
+            noise (torch.Tensor): The noise tensor used to noise the init_latents. In 'packed' format.
+        """
+        assert init_latents.shape == inpaint_mask.shape == noise.shape
+        self._init_latents = init_latents
+        self._inpaint_mask = inpaint_mask
+        self._noise = noise
+
+    def merge_intermediate_latents_with_init_latents(
+        self, intermediate_latents: torch.Tensor, timestep: float
+    ) -> torch.Tensor:
+        """Merge the intermediate latents with the initial latents for the current timestep using the inpaint mask. I.e.
+        update the intermediate latents to keep the regions that are not being inpainted on the correct noise
+        trajectory.
+
+        This function should be called after each denoising step.
+        """
+        # Noise the init latents for the current timestep.
+        noised_init_latents = self._noise * timestep + (1.0 - timestep) * self._init_latents
+
+        # Merge the intermediate latents with the noised_init_latents using the inpaint_mask.
+        return intermediate_latents * self._inpaint_mask + noised_init_latents * (1.0 - self._inpaint_mask)

invokeai/backend/flux/math.py

@@ -0,0 +1,32 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+import torch
+from einops import rearrange
+from torch import Tensor
+
+
+def attention(q: Tensor, k: Tensor, v: Tensor, pe: Tensor) -> Tensor:
+    q, k = apply_rope(q, k, pe)
+
+    x = torch.nn.functional.scaled_dot_product_attention(q, k, v)
+    x = rearrange(x, "B H L D -> B L (H D)")
+
+    return x
+
+
+def rope(pos: Tensor, dim: int, theta: int) -> Tensor:
+    assert dim % 2 == 0
+    scale = torch.arange(0, dim, 2, dtype=torch.float64, device=pos.device) / dim
+    omega = 1.0 / (theta**scale)
+    out = torch.einsum("...n,d->...nd", pos, omega)
+    out = torch.stack([torch.cos(out), -torch.sin(out), torch.sin(out), torch.cos(out)], dim=-1)
+    out = rearrange(out, "b n d (i j) -> b n d i j", i=2, j=2)
+    return out.float()
+
+
+def apply_rope(xq: Tensor, xk: Tensor, freqs_cis: Tensor) -> tuple[Tensor, Tensor]:
+    xq_ = xq.float().reshape(*xq.shape[:-1], -1, 1, 2)
+    xk_ = xk.float().reshape(*xk.shape[:-1], -1, 1, 2)
+    xq_out = freqs_cis[..., 0] * xq_[..., 0] + freqs_cis[..., 1] * xq_[..., 1]
+    xk_out = freqs_cis[..., 0] * xk_[..., 0] + freqs_cis[..., 1] * xk_[..., 1]
+    return xq_out.reshape(*xq.shape).type_as(xq), xk_out.reshape(*xk.shape).type_as(xk)

invokeai/backend/flux/model.py

@@ -0,0 +1,117 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from dataclasses import dataclass
+
+import torch
+from torch import Tensor, nn
+
+from invokeai.backend.flux.modules.layers import (
+    DoubleStreamBlock,
+    EmbedND,
+    LastLayer,
+    MLPEmbedder,
+    SingleStreamBlock,
+    timestep_embedding,
+)
+
+
+@dataclass
+class FluxParams:
+    in_channels: int
+    vec_in_dim: int
+    context_in_dim: int
+    hidden_size: int
+    mlp_ratio: float
+    num_heads: int
+    depth: int
+    depth_single_blocks: int
+    axes_dim: list[int]
+    theta: int
+    qkv_bias: bool
+    guidance_embed: bool
+
+
+class Flux(nn.Module):
+    """
+    Transformer model for flow matching on sequences.
+    """
+
+    def __init__(self, params: FluxParams):
+        super().__init__()
+
+        self.params = params
+        self.in_channels = params.in_channels
+        self.out_channels = self.in_channels
+        if params.hidden_size % params.num_heads != 0:
+            raise ValueError(f"Hidden size {params.hidden_size} must be divisible by num_heads {params.num_heads}")
+        pe_dim = params.hidden_size // params.num_heads
+        if sum(params.axes_dim) != pe_dim:
+            raise ValueError(f"Got {params.axes_dim} but expected positional dim {pe_dim}")
+        self.hidden_size = params.hidden_size
+        self.num_heads = params.num_heads
+        self.pe_embedder = EmbedND(dim=pe_dim, theta=params.theta, axes_dim=params.axes_dim)
+        self.img_in = nn.Linear(self.in_channels, self.hidden_size, bias=True)
+        self.time_in = MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size)
+        self.vector_in = MLPEmbedder(params.vec_in_dim, self.hidden_size)
+        self.guidance_in = (
+            MLPEmbedder(in_dim=256, hidden_dim=self.hidden_size) if params.guidance_embed else nn.Identity()
+        )
+        self.txt_in = nn.Linear(params.context_in_dim, self.hidden_size)
+
+        self.double_blocks = nn.ModuleList(
+            [
+                DoubleStreamBlock(
+                    self.hidden_size,
+                    self.num_heads,
+                    mlp_ratio=params.mlp_ratio,
+                    qkv_bias=params.qkv_bias,
+                )
+                for _ in range(params.depth)
+            ]
+        )
+
+        self.single_blocks = nn.ModuleList(
+            [
+                SingleStreamBlock(self.hidden_size, self.num_heads, mlp_ratio=params.mlp_ratio)
+                for _ in range(params.depth_single_blocks)
+            ]
+        )
+
+        self.final_layer = LastLayer(self.hidden_size, 1, self.out_channels)
+
+    def forward(
+        self,
+        img: Tensor,
+        img_ids: Tensor,
+        txt: Tensor,
+        txt_ids: Tensor,
+        timesteps: Tensor,
+        y: Tensor,
+        guidance: Tensor | None = None,
+    ) -> Tensor:
+        if img.ndim != 3 or txt.ndim != 3:
+            raise ValueError("Input img and txt tensors must have 3 dimensions.")
+
+        # running on sequences img
+        img = self.img_in(img)
+        vec = self.time_in(timestep_embedding(timesteps, 256))
+        if self.params.guidance_embed:
+            if guidance is None:
+                raise ValueError("Didn't get guidance strength for guidance distilled model.")
+            vec = vec + self.guidance_in(timestep_embedding(guidance, 256))
+        vec = vec + self.vector_in(y)
+        txt = self.txt_in(txt)
+
+        ids = torch.cat((txt_ids, img_ids), dim=1)
+        pe = self.pe_embedder(ids)
+
+        for block in self.double_blocks:
+            img, txt = block(img=img, txt=txt, vec=vec, pe=pe)
+
+        img = torch.cat((txt, img), 1)
+        for block in self.single_blocks:
+            img = block(img, vec=vec, pe=pe)
+        img = img[:, txt.shape[1] :, ...]
+
+        img = self.final_layer(img, vec)  # (N, T, patch_size ** 2 * out_channels)
+        return img

invokeai/backend/flux/modules/autoencoder.py

@@ -0,0 +1,324 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from dataclasses import dataclass
+
+import torch
+from einops import rearrange
+from torch import Tensor, nn
+
+
+@dataclass
+class AutoEncoderParams:
+    resolution: int
+    in_channels: int
+    ch: int
+    out_ch: int
+    ch_mult: list[int]
+    num_res_blocks: int
+    z_channels: int
+    scale_factor: float
+    shift_factor: float
+
+
+class AttnBlock(nn.Module):
+    def __init__(self, in_channels: int):
+        super().__init__()
+        self.in_channels = in_channels
+
+        self.norm = nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
+
+        self.q = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+        self.k = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+        self.v = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+        self.proj_out = nn.Conv2d(in_channels, in_channels, kernel_size=1)
+
+    def attention(self, h_: Tensor) -> Tensor:
+        h_ = self.norm(h_)
+        q = self.q(h_)
+        k = self.k(h_)
+        v = self.v(h_)
+
+        b, c, h, w = q.shape
+        q = rearrange(q, "b c h w -> b 1 (h w) c").contiguous()
+        k = rearrange(k, "b c h w -> b 1 (h w) c").contiguous()
+        v = rearrange(v, "b c h w -> b 1 (h w) c").contiguous()
+        h_ = nn.functional.scaled_dot_product_attention(q, k, v)
+
+        return rearrange(h_, "b 1 (h w) c -> b c h w", h=h, w=w, c=c, b=b)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return x + self.proj_out(self.attention(x))
+
+
+class ResnetBlock(nn.Module):
+    def __init__(self, in_channels: int, out_channels: int):
+        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 = nn.GroupNorm(num_groups=32, num_channels=in_channels, eps=1e-6, affine=True)
+        self.conv1 = nn.Conv2d(in_channels, out_channels, kernel_size=3, stride=1, padding=1)
+        self.norm2 = nn.GroupNorm(num_groups=32, num_channels=out_channels, eps=1e-6, affine=True)
+        self.conv2 = nn.Conv2d(out_channels, out_channels, kernel_size=3, stride=1, padding=1)
+        if self.in_channels != self.out_channels:
+            self.nin_shortcut = nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0)
+
+    def forward(self, x):
+        h = x
+        h = self.norm1(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv1(h)
+
+        h = self.norm2(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv2(h)
+
+        if self.in_channels != self.out_channels:
+            x = self.nin_shortcut(x)
+
+        return x + h
+
+
+class Downsample(nn.Module):
+    def __init__(self, in_channels: int):
+        super().__init__()
+        # no asymmetric padding in torch conv, must do it ourselves
+        self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=2, padding=0)
+
+    def forward(self, x: Tensor):
+        pad = (0, 1, 0, 1)
+        x = nn.functional.pad(x, pad, mode="constant", value=0)
+        x = self.conv(x)
+        return x
+
+
+class Upsample(nn.Module):
+    def __init__(self, in_channels: int):
+        super().__init__()
+        self.conv = nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=1, padding=1)
+
+    def forward(self, x: Tensor):
+        x = nn.functional.interpolate(x, scale_factor=2.0, mode="nearest")
+        x = self.conv(x)
+        return x
+
+
+class Encoder(nn.Module):
+    def __init__(
+        self,
+        resolution: int,
+        in_channels: int,
+        ch: int,
+        ch_mult: list[int],
+        num_res_blocks: int,
+        z_channels: int,
+    ):
+        super().__init__()
+        self.ch = ch
+        self.num_resolutions = len(ch_mult)
+        self.num_res_blocks = num_res_blocks
+        self.resolution = resolution
+        self.in_channels = in_channels
+        # downsampling
+        self.conv_in = nn.Conv2d(in_channels, self.ch, kernel_size=3, stride=1, padding=1)
+
+        curr_res = resolution
+        in_ch_mult = (1,) + tuple(ch_mult)
+        self.in_ch_mult = in_ch_mult
+        self.down = nn.ModuleList()
+        block_in = self.ch
+        for i_level in range(self.num_resolutions):
+            block = nn.ModuleList()
+            attn = nn.ModuleList()
+            block_in = ch * in_ch_mult[i_level]
+            block_out = ch * ch_mult[i_level]
+            for _ in range(self.num_res_blocks):
+                block.append(ResnetBlock(in_channels=block_in, out_channels=block_out))
+                block_in = block_out
+            down = nn.Module()
+            down.block = block
+            down.attn = attn
+            if i_level != self.num_resolutions - 1:
+                down.downsample = Downsample(block_in)
+                curr_res = curr_res // 2
+            self.down.append(down)
+
+        # middle
+        self.mid = nn.Module()
+        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+        self.mid.attn_1 = AttnBlock(block_in)
+        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+
+        # end
+        self.norm_out = nn.GroupNorm(num_groups=32, num_channels=block_in, eps=1e-6, affine=True)
+        self.conv_out = nn.Conv2d(block_in, 2 * z_channels, kernel_size=3, stride=1, padding=1)
+
+    def forward(self, x: Tensor) -> Tensor:
+        # 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])
+                if len(self.down[i_level].attn) > 0:
+                    h = self.down[i_level].attn[i_block](h)
+                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 = torch.nn.functional.silu(h)
+        h = self.conv_out(h)
+        return h
+
+
+class Decoder(nn.Module):
+    def __init__(
+        self,
+        ch: int,
+        out_ch: int,
+        ch_mult: list[int],
+        num_res_blocks: int,
+        in_channels: int,
+        resolution: int,
+        z_channels: int,
+    ):
+        super().__init__()
+        self.ch = ch
+        self.num_resolutions = len(ch_mult)
+        self.num_res_blocks = num_res_blocks
+        self.resolution = resolution
+        self.in_channels = in_channels
+        self.ffactor = 2 ** (self.num_resolutions - 1)
+
+        # compute in_ch_mult, block_in and curr_res at lowest res
+        block_in = ch * ch_mult[self.num_resolutions - 1]
+        curr_res = resolution // 2 ** (self.num_resolutions - 1)
+        self.z_shape = (1, z_channels, curr_res, curr_res)
+
+        # z to block_in
+        self.conv_in = nn.Conv2d(z_channels, block_in, kernel_size=3, stride=1, padding=1)
+
+        # middle
+        self.mid = nn.Module()
+        self.mid.block_1 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+        self.mid.attn_1 = AttnBlock(block_in)
+        self.mid.block_2 = ResnetBlock(in_channels=block_in, out_channels=block_in)
+
+        # upsampling
+        self.up = nn.ModuleList()
+        for i_level in reversed(range(self.num_resolutions)):
+            block = nn.ModuleList()
+            attn = nn.ModuleList()
+            block_out = ch * ch_mult[i_level]
+            for _ in range(self.num_res_blocks + 1):
+                block.append(ResnetBlock(in_channels=block_in, out_channels=block_out))
+                block_in = block_out
+            up = nn.Module()
+            up.block = block
+            up.attn = attn
+            if i_level != 0:
+                up.upsample = Upsample(block_in)
+                curr_res = curr_res * 2
+            self.up.insert(0, up)  # prepend to get consistent order
+
+        # end
+        self.norm_out = nn.GroupNorm(num_groups=32, num_channels=block_in, eps=1e-6, affine=True)
+        self.conv_out = nn.Conv2d(block_in, out_ch, kernel_size=3, stride=1, padding=1)
+
+    def forward(self, z: Tensor) -> Tensor:
+        # z to block_in
+        h = self.conv_in(z)
+
+        # middle
+        h = self.mid.block_1(h)
+        h = self.mid.attn_1(h)
+        h = self.mid.block_2(h)
+
+        # upsampling
+        for i_level in reversed(range(self.num_resolutions)):
+            for i_block in range(self.num_res_blocks + 1):
+                h = self.up[i_level].block[i_block](h)
+                if len(self.up[i_level].attn) > 0:
+                    h = self.up[i_level].attn[i_block](h)
+            if i_level != 0:
+                h = self.up[i_level].upsample(h)
+
+        # end
+        h = self.norm_out(h)
+        h = torch.nn.functional.silu(h)
+        h = self.conv_out(h)
+        return h
+
+
+class DiagonalGaussian(nn.Module):
+    def __init__(self, chunk_dim: int = 1):
+        super().__init__()
+        self.chunk_dim = chunk_dim
+
+    def forward(self, z: Tensor, sample: bool = True, generator: torch.Generator | None = None) -> Tensor:
+        mean, logvar = torch.chunk(z, 2, dim=self.chunk_dim)
+        if sample:
+            std = torch.exp(0.5 * logvar)
+            # Unfortunately, torch.randn_like(...) does not accept a generator argument at the time of writing, so we
+            # have to use torch.randn(...) instead.
+            return mean + std * torch.randn(size=mean.size(), generator=generator, dtype=mean.dtype, device=mean.device)
+        else:
+            return mean
+
+
+class AutoEncoder(nn.Module):
+    def __init__(self, params: AutoEncoderParams):
+        super().__init__()
+        self.encoder = Encoder(
+            resolution=params.resolution,
+            in_channels=params.in_channels,
+            ch=params.ch,
+            ch_mult=params.ch_mult,
+            num_res_blocks=params.num_res_blocks,
+            z_channels=params.z_channels,
+        )
+        self.decoder = Decoder(
+            resolution=params.resolution,
+            in_channels=params.in_channels,
+            ch=params.ch,
+            out_ch=params.out_ch,
+            ch_mult=params.ch_mult,
+            num_res_blocks=params.num_res_blocks,
+            z_channels=params.z_channels,
+        )
+        self.reg = DiagonalGaussian()
+
+        self.scale_factor = params.scale_factor
+        self.shift_factor = params.shift_factor
+
+    def encode(self, x: Tensor, sample: bool = True, generator: torch.Generator | None = None) -> Tensor:
+        """Run VAE encoding on input tensor x.
+
+        Args:
+            x (Tensor): Input image tensor. Shape: (batch_size, in_channels, height, width).
+            sample (bool, optional): If True, sample from the encoded distribution, else, return the distribution mean.
+                Defaults to True.
+            generator (torch.Generator | None, optional): Optional random number generator for reproducibility.
+                Defaults to None.
+
+        Returns:
+            Tensor: Encoded latent tensor. Shape: (batch_size, z_channels, latent_height, latent_width).
+        """
+
+        z = self.reg(self.encoder(x), sample=sample, generator=generator)
+        z = self.scale_factor * (z - self.shift_factor)
+        return z
+
+    def decode(self, z: Tensor) -> Tensor:
+        z = z / self.scale_factor + self.shift_factor
+        return self.decoder(z)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return self.decode(self.encode(x))

invokeai/backend/flux/modules/conditioner.py

@@ -0,0 +1,33 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from torch import Tensor, nn
+from transformers import PreTrainedModel, PreTrainedTokenizer
+
+
+class HFEncoder(nn.Module):
+    def __init__(self, encoder: PreTrainedModel, tokenizer: PreTrainedTokenizer, is_clip: bool, max_length: int):
+        super().__init__()
+        self.max_length = max_length
+        self.is_clip = is_clip
+        self.output_key = "pooler_output" if self.is_clip else "last_hidden_state"
+        self.tokenizer = tokenizer
+        self.hf_module = encoder
+        self.hf_module = self.hf_module.eval().requires_grad_(False)
+
+    def forward(self, text: list[str]) -> Tensor:
+        batch_encoding = self.tokenizer(
+            text,
+            truncation=True,
+            max_length=self.max_length,
+            return_length=False,
+            return_overflowing_tokens=False,
+            padding="max_length",
+            return_tensors="pt",
+        )
+
+        outputs = self.hf_module(
+            input_ids=batch_encoding["input_ids"].to(self.hf_module.device),
+            attention_mask=None,
+            output_hidden_states=False,
+        )
+        return outputs[self.output_key]

invokeai/backend/flux/modules/layers.py

@@ -0,0 +1,253 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+import math
+from dataclasses import dataclass
+
+import torch
+from einops import rearrange
+from torch import Tensor, nn
+
+from invokeai.backend.flux.math import attention, rope
+
+
+class EmbedND(nn.Module):
+    def __init__(self, dim: int, theta: int, axes_dim: list[int]):
+        super().__init__()
+        self.dim = dim
+        self.theta = theta
+        self.axes_dim = axes_dim
+
+    def forward(self, ids: Tensor) -> Tensor:
+        n_axes = ids.shape[-1]
+        emb = torch.cat(
+            [rope(ids[..., i], self.axes_dim[i], self.theta) for i in range(n_axes)],
+            dim=-3,
+        )
+
+        return emb.unsqueeze(1)
+
+
+def timestep_embedding(t: Tensor, dim, max_period=10000, time_factor: float = 1000.0):
+    """
+    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.
+    """
+    t = time_factor * t
+    half = dim // 2
+    freqs = torch.exp(-math.log(max_period) * torch.arange(start=0, end=half, dtype=torch.float32) / half).to(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(t)
+    return embedding
+
+
+class MLPEmbedder(nn.Module):
+    def __init__(self, in_dim: int, hidden_dim: int):
+        super().__init__()
+        self.in_layer = nn.Linear(in_dim, hidden_dim, bias=True)
+        self.silu = nn.SiLU()
+        self.out_layer = nn.Linear(hidden_dim, hidden_dim, bias=True)
+
+    def forward(self, x: Tensor) -> Tensor:
+        return self.out_layer(self.silu(self.in_layer(x)))
+
+
+class RMSNorm(torch.nn.Module):
+    def __init__(self, dim: int):
+        super().__init__()
+        self.scale = nn.Parameter(torch.ones(dim))
+
+    def forward(self, x: Tensor):
+        x_dtype = x.dtype
+        x = x.float()
+        rrms = torch.rsqrt(torch.mean(x**2, dim=-1, keepdim=True) + 1e-6)
+        return (x * rrms).to(dtype=x_dtype) * self.scale
+
+
+class QKNorm(torch.nn.Module):
+    def __init__(self, dim: int):
+        super().__init__()
+        self.query_norm = RMSNorm(dim)
+        self.key_norm = RMSNorm(dim)
+
+    def forward(self, q: Tensor, k: Tensor, v: Tensor) -> tuple[Tensor, Tensor]:
+        q = self.query_norm(q)
+        k = self.key_norm(k)
+        return q.to(v), k.to(v)
+
+
+class SelfAttention(nn.Module):
+    def __init__(self, dim: int, num_heads: int = 8, qkv_bias: bool = False):
+        super().__init__()
+        self.num_heads = num_heads
+        head_dim = dim // num_heads
+
+        self.qkv = nn.Linear(dim, dim * 3, bias=qkv_bias)
+        self.norm = QKNorm(head_dim)
+        self.proj = nn.Linear(dim, dim)
+
+    def forward(self, x: Tensor, pe: Tensor) -> Tensor:
+        qkv = self.qkv(x)
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+        x = attention(q, k, v, pe=pe)
+        x = self.proj(x)
+        return x
+
+
+@dataclass
+class ModulationOut:
+    shift: Tensor
+    scale: Tensor
+    gate: Tensor
+
+
+class Modulation(nn.Module):
+    def __init__(self, dim: int, double: bool):
+        super().__init__()
+        self.is_double = double
+        self.multiplier = 6 if double else 3
+        self.lin = nn.Linear(dim, self.multiplier * dim, bias=True)
+
+    def forward(self, vec: Tensor) -> tuple[ModulationOut, ModulationOut | None]:
+        out = self.lin(nn.functional.silu(vec))[:, None, :].chunk(self.multiplier, dim=-1)
+
+        return (
+            ModulationOut(*out[:3]),
+            ModulationOut(*out[3:]) if self.is_double else None,
+        )
+
+
+class DoubleStreamBlock(nn.Module):
+    def __init__(self, hidden_size: int, num_heads: int, mlp_ratio: float, qkv_bias: bool = False):
+        super().__init__()
+
+        mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        self.num_heads = num_heads
+        self.hidden_size = hidden_size
+        self.img_mod = Modulation(hidden_size, double=True)
+        self.img_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.img_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
+
+        self.img_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.img_mlp = nn.Sequential(
+            nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
+            nn.GELU(approximate="tanh"),
+            nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
+        )
+
+        self.txt_mod = Modulation(hidden_size, double=True)
+        self.txt_norm1 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.txt_attn = SelfAttention(dim=hidden_size, num_heads=num_heads, qkv_bias=qkv_bias)
+
+        self.txt_norm2 = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.txt_mlp = nn.Sequential(
+            nn.Linear(hidden_size, mlp_hidden_dim, bias=True),
+            nn.GELU(approximate="tanh"),
+            nn.Linear(mlp_hidden_dim, hidden_size, bias=True),
+        )
+
+    def forward(self, img: Tensor, txt: Tensor, vec: Tensor, pe: Tensor) -> tuple[Tensor, Tensor]:
+        img_mod1, img_mod2 = self.img_mod(vec)
+        txt_mod1, txt_mod2 = self.txt_mod(vec)
+
+        # prepare image for attention
+        img_modulated = self.img_norm1(img)
+        img_modulated = (1 + img_mod1.scale) * img_modulated + img_mod1.shift
+        img_qkv = self.img_attn.qkv(img_modulated)
+        img_q, img_k, img_v = rearrange(img_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        img_q, img_k = self.img_attn.norm(img_q, img_k, img_v)
+
+        # prepare txt for attention
+        txt_modulated = self.txt_norm1(txt)
+        txt_modulated = (1 + txt_mod1.scale) * txt_modulated + txt_mod1.shift
+        txt_qkv = self.txt_attn.qkv(txt_modulated)
+        txt_q, txt_k, txt_v = rearrange(txt_qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        txt_q, txt_k = self.txt_attn.norm(txt_q, txt_k, txt_v)
+
+        # run actual attention
+        q = torch.cat((txt_q, img_q), dim=2)
+        k = torch.cat((txt_k, img_k), dim=2)
+        v = torch.cat((txt_v, img_v), dim=2)
+
+        attn = attention(q, k, v, pe=pe)
+        txt_attn, img_attn = attn[:, : txt.shape[1]], attn[:, txt.shape[1] :]
+
+        # calculate the img bloks
+        img = img + img_mod1.gate * self.img_attn.proj(img_attn)
+        img = img + img_mod2.gate * self.img_mlp((1 + img_mod2.scale) * self.img_norm2(img) + img_mod2.shift)
+
+        # calculate the txt bloks
+        txt = txt + txt_mod1.gate * self.txt_attn.proj(txt_attn)
+        txt = txt + txt_mod2.gate * self.txt_mlp((1 + txt_mod2.scale) * self.txt_norm2(txt) + txt_mod2.shift)
+        return img, txt
+
+
+class SingleStreamBlock(nn.Module):
+    """
+    A DiT block with parallel linear layers as described in
+    https://arxiv.org/abs/2302.05442 and adapted modulation interface.
+    """
+
+    def __init__(
+        self,
+        hidden_size: int,
+        num_heads: int,
+        mlp_ratio: float = 4.0,
+        qk_scale: float | None = None,
+    ):
+        super().__init__()
+        self.hidden_dim = hidden_size
+        self.num_heads = num_heads
+        head_dim = hidden_size // num_heads
+        self.scale = qk_scale or head_dim**-0.5
+
+        self.mlp_hidden_dim = int(hidden_size * mlp_ratio)
+        # qkv and mlp_in
+        self.linear1 = nn.Linear(hidden_size, hidden_size * 3 + self.mlp_hidden_dim)
+        # proj and mlp_out
+        self.linear2 = nn.Linear(hidden_size + self.mlp_hidden_dim, hidden_size)
+
+        self.norm = QKNorm(head_dim)
+
+        self.hidden_size = hidden_size
+        self.pre_norm = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+
+        self.mlp_act = nn.GELU(approximate="tanh")
+        self.modulation = Modulation(hidden_size, double=False)
+
+    def forward(self, x: Tensor, vec: Tensor, pe: Tensor) -> Tensor:
+        mod, _ = self.modulation(vec)
+        x_mod = (1 + mod.scale) * self.pre_norm(x) + mod.shift
+        qkv, mlp = torch.split(self.linear1(x_mod), [3 * self.hidden_size, self.mlp_hidden_dim], dim=-1)
+
+        q, k, v = rearrange(qkv, "B L (K H D) -> K B H L D", K=3, H=self.num_heads)
+        q, k = self.norm(q, k, v)
+
+        # compute attention
+        attn = attention(q, k, v, pe=pe)
+        # compute activation in mlp stream, cat again and run second linear layer
+        output = self.linear2(torch.cat((attn, self.mlp_act(mlp)), 2))
+        return x + mod.gate * output
+
+
+class LastLayer(nn.Module):
+    def __init__(self, hidden_size: int, patch_size: int, out_channels: int):
+        super().__init__()
+        self.norm_final = nn.LayerNorm(hidden_size, elementwise_affine=False, eps=1e-6)
+        self.linear = nn.Linear(hidden_size, patch_size * patch_size * out_channels, bias=True)
+        self.adaLN_modulation = nn.Sequential(nn.SiLU(), nn.Linear(hidden_size, 2 * hidden_size, bias=True))
+
+    def forward(self, x: Tensor, vec: Tensor) -> Tensor:
+        shift, scale = self.adaLN_modulation(vec).chunk(2, dim=1)
+        x = (1 + scale[:, None, :]) * self.norm_final(x) + shift[:, None, :]
+        x = self.linear(x)
+        return x

invokeai/backend/flux/sampling_utils.py

@@ -0,0 +1,135 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+import math
+from typing import Callable
+
+import torch
+from einops import rearrange, repeat
+
+
+def get_noise(
+    num_samples: int,
+    height: int,
+    width: int,
+    device: torch.device,
+    dtype: torch.dtype,
+    seed: int,
+):
+    # We always generate noise on the same device and dtype then cast to ensure consistency across devices/dtypes.
+    rand_device = "cpu"
+    rand_dtype = torch.float16
+    return torch.randn(
+        num_samples,
+        16,
+        # allow for packing
+        2 * math.ceil(height / 16),
+        2 * math.ceil(width / 16),
+        device=rand_device,
+        dtype=rand_dtype,
+        generator=torch.Generator(device=rand_device).manual_seed(seed),
+    ).to(device=device, dtype=dtype)
+
+
+def time_shift(mu: float, sigma: float, t: torch.Tensor) -> torch.Tensor:
+    return math.exp(mu) / (math.exp(mu) + (1 / t - 1) ** sigma)
+
+
+def get_lin_function(x1: float = 256, y1: float = 0.5, x2: float = 4096, y2: float = 1.15) -> Callable[[float], float]:
+    m = (y2 - y1) / (x2 - x1)
+    b = y1 - m * x1
+    return lambda x: m * x + b
+
+
+def get_schedule(
+    num_steps: int,
+    image_seq_len: int,
+    base_shift: float = 0.5,
+    max_shift: float = 1.15,
+    shift: bool = True,
+) -> list[float]:
+    # extra step for zero
+    timesteps = torch.linspace(1, 0, num_steps + 1)
+
+    # shifting the schedule to favor high timesteps for higher signal images
+    if shift:
+        # estimate mu based on linear estimation between two points
+        mu = get_lin_function(y1=base_shift, y2=max_shift)(image_seq_len)
+        timesteps = time_shift(mu, 1.0, timesteps)
+
+    return timesteps.tolist()
+
+
+def _find_last_index_ge_val(timesteps: list[float], val: float, eps: float = 1e-6) -> int:
+    """Find the last index in timesteps that is >= val.
+
+    We use epsilon-close equality to avoid potential floating point errors.
+    """
+    idx = len(list(filter(lambda t: t >= (val - eps), timesteps))) - 1
+    assert idx >= 0
+    return idx
+
+
+def clip_timestep_schedule(timesteps: list[float], denoising_start: float, denoising_end: float) -> list[float]:
+    """Clip the timestep schedule to the denoising range.
+
+    Args:
+        timesteps (list[float]): The original timestep schedule: [1.0, ..., 0.0].
+        denoising_start (float): A value in [0, 1] specifying the start of the denoising process. E.g. a value of 0.2
+            would mean that the denoising process start at the last timestep in the schedule >= 0.8.
+        denoising_end (float): A value in [0, 1] specifying the end of the denoising process. E.g. a value of 0.8 would
+            mean that the denoising process end at the last timestep in the schedule >= 0.2.
+
+    Returns:
+        list[float]: The clipped timestep schedule.
+    """
+    assert 0.0 <= denoising_start <= 1.0
+    assert 0.0 <= denoising_end <= 1.0
+    assert denoising_start <= denoising_end
+
+    t_start_val = 1.0 - denoising_start
+    t_end_val = 1.0 - denoising_end
+
+    t_start_idx = _find_last_index_ge_val(timesteps, t_start_val)
+    t_end_idx = _find_last_index_ge_val(timesteps, t_end_val)
+
+    clipped_timesteps = timesteps[t_start_idx : t_end_idx + 1]
+
+    return clipped_timesteps
+
+
+def unpack(x: torch.Tensor, height: int, width: int) -> torch.Tensor:
+    """Unpack flat array of patch embeddings to latent image."""
+    return rearrange(
+        x,
+        "b (h w) (c ph pw) -> b c (h ph) (w pw)",
+        h=math.ceil(height / 16),
+        w=math.ceil(width / 16),
+        ph=2,
+        pw=2,
+    )
+
+
+def pack(x: torch.Tensor) -> torch.Tensor:
+    """Pack latent image to flattented array of patch embeddings."""
+    # Pixel unshuffle with a scale of 2, and flatten the height/width dimensions to get an array of patches.
+    return rearrange(x, "b c (h ph) (w pw) -> b (h w) (c ph pw)", ph=2, pw=2)
+
+
+def generate_img_ids(h: int, w: int, batch_size: int, device: torch.device, dtype: torch.dtype) -> torch.Tensor:
+    """Generate tensor of image position ids.
+
+    Args:
+        h (int): Height of image in latent space.
+        w (int): Width of image in latent space.
+        batch_size (int): Batch size.
+        device (torch.device): Device.
+        dtype (torch.dtype): dtype.
+
+    Returns:
+        torch.Tensor: Image position ids.
+    """
+    img_ids = torch.zeros(h // 2, w // 2, 3, device=device, dtype=dtype)
+    img_ids[..., 1] = img_ids[..., 1] + torch.arange(h // 2, device=device, dtype=dtype)[:, None]
+    img_ids[..., 2] = img_ids[..., 2] + torch.arange(w // 2, device=device, dtype=dtype)[None, :]
+    img_ids = repeat(img_ids, "h w c -> b (h w) c", b=batch_size)
+    return img_ids

invokeai/backend/flux/util.py

@@ -0,0 +1,71 @@
+# Initially pulled from https://github.com/black-forest-labs/flux
+
+from dataclasses import dataclass
+from typing import Dict, Literal
+
+from invokeai.backend.flux.model import FluxParams
+from invokeai.backend.flux.modules.autoencoder import AutoEncoderParams
+
+
+@dataclass
+class ModelSpec:
+    params: FluxParams
+    ae_params: AutoEncoderParams
+    ckpt_path: str | None
+    ae_path: str | None
+    repo_id: str | None
+    repo_flow: str | None
+    repo_ae: str | None
+
+
+max_seq_lengths: Dict[str, Literal[256, 512]] = {
+    "flux-dev": 512,
+    "flux-schnell": 256,
+}
+
+
+ae_params = {
+    "flux": AutoEncoderParams(
+        resolution=256,
+        in_channels=3,
+        ch=128,
+        out_ch=3,
+        ch_mult=[1, 2, 4, 4],
+        num_res_blocks=2,
+        z_channels=16,
+        scale_factor=0.3611,
+        shift_factor=0.1159,
+    )
+}
+
+
+params = {
+    "flux-dev": FluxParams(
+        in_channels=64,
+        vec_in_dim=768,
+        context_in_dim=4096,
+        hidden_size=3072,
+        mlp_ratio=4.0,
+        num_heads=24,
+        depth=19,
+        depth_single_blocks=38,
+        axes_dim=[16, 56, 56],
+        theta=10_000,
+        qkv_bias=True,
+        guidance_embed=True,
+    ),
+    "flux-schnell": FluxParams(
+        in_channels=64,
+        vec_in_dim=768,
+        context_in_dim=4096,
+        hidden_size=3072,
+        mlp_ratio=4.0,
+        num_heads=24,
+        depth=19,
+        depth_single_blocks=38,
+        axes_dim=[16, 56, 56],
+        theta=10_000,
+        qkv_bias=True,
+        guidance_embed=False,
+    ),
+}

invokeai/backend/lora.py

@@ -1,672 +0,0 @@
-# Copyright (c) 2024 The InvokeAI Development team
-"""LoRA model support."""
-
-import bisect
-from pathlib import Path
-from typing import Dict, List, Optional, Set, Tuple, Union
-
-import torch
-from safetensors.torch import load_file
-from typing_extensions import Self
-
-import invokeai.backend.util.logging as logger
-from invokeai.backend.model_manager import BaseModelType
-from invokeai.backend.raw_model import RawModel
-
-
-class LoRALayerBase:
-    # rank: Optional[int]
-    # alpha: Optional[float]
-    # bias: Optional[torch.Tensor]
-    # layer_key: str
-
-    # @property
-    # def scale(self):
-    #    return self.alpha / self.rank if (self.alpha and self.rank) else 1.0
-
-    def __init__(
-        self,
-        layer_key: str,
-        values: Dict[str, torch.Tensor],
-    ):
-        if "alpha" in values:
-            self.alpha = values["alpha"].item()
-        else:
-            self.alpha = None
-
-        if "bias_indices" in values and "bias_values" in values and "bias_size" in values:
-            self.bias: Optional[torch.Tensor] = torch.sparse_coo_tensor(
-                values["bias_indices"],
-                values["bias_values"],
-                tuple(values["bias_size"]),
-            )
-
-        else:
-            self.bias = None
-
-        self.rank = None  # set in layer implementation
-        self.layer_key = layer_key
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        raise NotImplementedError()
-
-    def get_bias(self, orig_bias: torch.Tensor) -> Optional[torch.Tensor]:
-        return self.bias
-
-    def get_parameters(self, orig_module: torch.nn.Module) -> Dict[str, torch.Tensor]:
-        params = {"weight": self.get_weight(orig_module.weight)}
-        bias = self.get_bias(orig_module.bias)
-        if bias is not None:
-            params["bias"] = bias
-        return params
-
-    def calc_size(self) -> int:
-        model_size = 0
-        for val in [self.bias]:
-            if val is not None:
-                model_size += val.nelement() * val.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        if self.bias is not None:
-            self.bias = self.bias.to(device=device, dtype=dtype)
-
-    def check_keys(self, values: Dict[str, torch.Tensor], known_keys: Set[str]):
-        """Log a warning if values contains unhandled keys."""
-        # {"alpha", "bias_indices", "bias_values", "bias_size"} are hard-coded, because they are handled by
-        # `LoRALayerBase`. Sub-classes should provide the known_keys that they handled.
-        all_known_keys = known_keys | {"alpha", "bias_indices", "bias_values", "bias_size"}
-        unknown_keys = set(values.keys()) - all_known_keys
-        if unknown_keys:
-            logger.warning(
-                f"Unexpected keys found in LoRA/LyCORIS layer, model might work incorrectly! Keys: {unknown_keys}"
-            )
-
-
-# TODO: find and debug lora/locon with bias
-class LoRALayer(LoRALayerBase):
-    # up: torch.Tensor
-    # mid: Optional[torch.Tensor]
-    # down: torch.Tensor
-
-    def __init__(
-        self,
-        layer_key: str,
-        values: Dict[str, torch.Tensor],
-    ):
-        super().__init__(layer_key, values)
-
-        self.up = values["lora_up.weight"]
-        self.down = values["lora_down.weight"]
-        self.mid = values.get("lora_mid.weight", None)
-
-        self.rank = self.down.shape[0]
-        self.check_keys(
-            values,
-            {
-                "lora_up.weight",
-                "lora_down.weight",
-                "lora_mid.weight",
-            },
-        )
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        if self.mid is not None:
-            up = self.up.reshape(self.up.shape[0], self.up.shape[1])
-            down = self.down.reshape(self.down.shape[0], self.down.shape[1])
-            weight = torch.einsum("m n w h, i m, n j -> i j w h", self.mid, up, down)
-        else:
-            weight = self.up.reshape(self.up.shape[0], -1) @ self.down.reshape(self.down.shape[0], -1)
-
-        return weight
-
-    def calc_size(self) -> int:
-        model_size = super().calc_size()
-        for val in [self.up, self.mid, self.down]:
-            if val is not None:
-                model_size += val.nelement() * val.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        super().to(device=device, dtype=dtype)
-
-        self.up = self.up.to(device=device, dtype=dtype)
-        self.down = self.down.to(device=device, dtype=dtype)
-
-        if self.mid is not None:
-            self.mid = self.mid.to(device=device, dtype=dtype)
-
-
-class LoHALayer(LoRALayerBase):
-    # w1_a: torch.Tensor
-    # w1_b: torch.Tensor
-    # w2_a: torch.Tensor
-    # w2_b: torch.Tensor
-    # t1: Optional[torch.Tensor] = None
-    # t2: Optional[torch.Tensor] = None
-
-    def __init__(self, layer_key: str, values: Dict[str, torch.Tensor]):
-        super().__init__(layer_key, values)
-
-        self.w1_a = values["hada_w1_a"]
-        self.w1_b = values["hada_w1_b"]
-        self.w2_a = values["hada_w2_a"]
-        self.w2_b = values["hada_w2_b"]
-        self.t1 = values.get("hada_t1", None)
-        self.t2 = values.get("hada_t2", None)
-
-        self.rank = self.w1_b.shape[0]
-        self.check_keys(
-            values,
-            {
-                "hada_w1_a",
-                "hada_w1_b",
-                "hada_w2_a",
-                "hada_w2_b",
-                "hada_t1",
-                "hada_t2",
-            },
-        )
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        if self.t1 is None:
-            weight: torch.Tensor = (self.w1_a @ self.w1_b) * (self.w2_a @ self.w2_b)
-
-        else:
-            rebuild1 = torch.einsum("i j k l, j r, i p -> p r k l", self.t1, self.w1_b, self.w1_a)
-            rebuild2 = torch.einsum("i j k l, j r, i p -> p r k l", self.t2, self.w2_b, self.w2_a)
-            weight = rebuild1 * rebuild2
-
-        return weight
-
-    def calc_size(self) -> int:
-        model_size = super().calc_size()
-        for val in [self.w1_a, self.w1_b, self.w2_a, self.w2_b, self.t1, self.t2]:
-            if val is not None:
-                model_size += val.nelement() * val.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        super().to(device=device, dtype=dtype)
-
-        self.w1_a = self.w1_a.to(device=device, dtype=dtype)
-        self.w1_b = self.w1_b.to(device=device, dtype=dtype)
-        if self.t1 is not None:
-            self.t1 = self.t1.to(device=device, dtype=dtype)
-
-        self.w2_a = self.w2_a.to(device=device, dtype=dtype)
-        self.w2_b = self.w2_b.to(device=device, dtype=dtype)
-        if self.t2 is not None:
-            self.t2 = self.t2.to(device=device, dtype=dtype)
-
-
-class LoKRLayer(LoRALayerBase):
-    # w1: Optional[torch.Tensor] = None
-    # w1_a: Optional[torch.Tensor] = None
-    # w1_b: Optional[torch.Tensor] = None
-    # w2: Optional[torch.Tensor] = None
-    # w2_a: Optional[torch.Tensor] = None
-    # w2_b: Optional[torch.Tensor] = None
-    # t2: Optional[torch.Tensor] = None
-
-    def __init__(
-        self,
-        layer_key: str,
-        values: Dict[str, torch.Tensor],
-    ):
-        super().__init__(layer_key, values)
-
-        self.w1 = values.get("lokr_w1", None)
-        if self.w1 is None:
-            self.w1_a = values["lokr_w1_a"]
-            self.w1_b = values["lokr_w1_b"]
-        else:
-            self.w1_b = None
-            self.w1_a = None
-
-        self.w2 = values.get("lokr_w2", None)
-        if self.w2 is None:
-            self.w2_a = values["lokr_w2_a"]
-            self.w2_b = values["lokr_w2_b"]
-        else:
-            self.w2_a = None
-            self.w2_b = None
-
-        self.t2 = values.get("lokr_t2", None)
-
-        if self.w1_b is not None:
-            self.rank = self.w1_b.shape[0]
-        elif self.w2_b is not None:
-            self.rank = self.w2_b.shape[0]
-        else:
-            self.rank = None  # unscaled
-
-        self.check_keys(
-            values,
-            {
-                "lokr_w1",
-                "lokr_w1_a",
-                "lokr_w1_b",
-                "lokr_w2",
-                "lokr_w2_a",
-                "lokr_w2_b",
-                "lokr_t2",
-            },
-        )
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        w1: Optional[torch.Tensor] = self.w1
-        if w1 is None:
-            assert self.w1_a is not None
-            assert self.w1_b is not None
-            w1 = self.w1_a @ self.w1_b
-
-        w2 = self.w2
-        if w2 is None:
-            if self.t2 is None:
-                assert self.w2_a is not None
-                assert self.w2_b is not None
-                w2 = self.w2_a @ self.w2_b
-            else:
-                w2 = torch.einsum("i j k l, i p, j r -> p r k l", self.t2, self.w2_a, self.w2_b)
-
-        if len(w2.shape) == 4:
-            w1 = w1.unsqueeze(2).unsqueeze(2)
-        w2 = w2.contiguous()
-        assert w1 is not None
-        assert w2 is not None
-        weight = torch.kron(w1, w2)
-
-        return weight
-
-    def calc_size(self) -> int:
-        model_size = super().calc_size()
-        for val in [self.w1, self.w1_a, self.w1_b, self.w2, self.w2_a, self.w2_b, self.t2]:
-            if val is not None:
-                model_size += val.nelement() * val.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        super().to(device=device, dtype=dtype)
-
-        if self.w1 is not None:
-            self.w1 = self.w1.to(device=device, dtype=dtype)
-        else:
-            assert self.w1_a is not None
-            assert self.w1_b is not None
-            self.w1_a = self.w1_a.to(device=device, dtype=dtype)
-            self.w1_b = self.w1_b.to(device=device, dtype=dtype)
-
-        if self.w2 is not None:
-            self.w2 = self.w2.to(device=device, dtype=dtype)
-        else:
-            assert self.w2_a is not None
-            assert self.w2_b is not None
-            self.w2_a = self.w2_a.to(device=device, dtype=dtype)
-            self.w2_b = self.w2_b.to(device=device, dtype=dtype)
-
-        if self.t2 is not None:
-            self.t2 = self.t2.to(device=device, dtype=dtype)
-
-
-class FullLayer(LoRALayerBase):
-    # bias handled in LoRALayerBase(calc_size, to)
-    # weight: torch.Tensor
-    # bias: Optional[torch.Tensor]
-
-    def __init__(
-        self,
-        layer_key: str,
-        values: Dict[str, torch.Tensor],
-    ):
-        super().__init__(layer_key, values)
-
-        self.weight = values["diff"]
-        self.bias = values.get("diff_b", None)
-
-        self.rank = None  # unscaled
-        self.check_keys(values, {"diff", "diff_b"})
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        return self.weight
-
-    def calc_size(self) -> int:
-        model_size = super().calc_size()
-        model_size += self.weight.nelement() * self.weight.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        super().to(device=device, dtype=dtype)
-
-        self.weight = self.weight.to(device=device, dtype=dtype)
-
-
-class IA3Layer(LoRALayerBase):
-    # weight: torch.Tensor
-    # on_input: torch.Tensor
-
-    def __init__(
-        self,
-        layer_key: str,
-        values: Dict[str, torch.Tensor],
-    ):
-        super().__init__(layer_key, values)
-
-        self.weight = values["weight"]
-        self.on_input = values["on_input"]
-
-        self.rank = None  # unscaled
-        self.check_keys(values, {"weight", "on_input"})
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        weight = self.weight
-        if not self.on_input:
-            weight = weight.reshape(-1, 1)
-        assert orig_weight is not None
-        return orig_weight * weight
-
-    def calc_size(self) -> int:
-        model_size = super().calc_size()
-        model_size += self.weight.nelement() * self.weight.element_size()
-        model_size += self.on_input.nelement() * self.on_input.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None):
-        super().to(device=device, dtype=dtype)
-
-        self.weight = self.weight.to(device=device, dtype=dtype)
-        self.on_input = self.on_input.to(device=device, dtype=dtype)
-
-
-class NormLayer(LoRALayerBase):
-    # bias handled in LoRALayerBase(calc_size, to)
-    # weight: torch.Tensor
-    # bias: Optional[torch.Tensor]
-
-    def __init__(
-        self,
-        layer_key: str,
-        values: Dict[str, torch.Tensor],
-    ):
-        super().__init__(layer_key, values)
-
-        self.weight = values["w_norm"]
-        self.bias = values.get("b_norm", None)
-
-        self.rank = None  # unscaled
-        self.check_keys(values, {"w_norm", "b_norm"})
-
-    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
-        return self.weight
-
-    def calc_size(self) -> int:
-        model_size = super().calc_size()
-        model_size += self.weight.nelement() * self.weight.element_size()
-        return model_size
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        super().to(device=device, dtype=dtype)
-
-        self.weight = self.weight.to(device=device, dtype=dtype)
-
-
-AnyLoRALayer = Union[LoRALayer, LoHALayer, LoKRLayer, FullLayer, IA3Layer, NormLayer]
-
-
-class LoRAModelRaw(RawModel):  # (torch.nn.Module):
-    _name: str
-    layers: Dict[str, AnyLoRALayer]
-
-    def __init__(
-        self,
-        name: str,
-        layers: Dict[str, AnyLoRALayer],
-    ):
-        self._name = name
-        self.layers = layers
-
-    @property
-    def name(self) -> str:
-        return self._name
-
-    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
-        # TODO: try revert if exception?
-        for _key, layer in self.layers.items():
-            layer.to(device=device, dtype=dtype)
-
-    def calc_size(self) -> int:
-        model_size = 0
-        for _, layer in self.layers.items():
-            model_size += layer.calc_size()
-        return model_size
-
-    @classmethod
-    def _convert_sdxl_keys_to_diffusers_format(cls, state_dict: Dict[str, torch.Tensor]) -> Dict[str, torch.Tensor]:
-        """Convert the keys of an SDXL LoRA state_dict to diffusers format.
-
-        The input state_dict can be in either Stability AI format or diffusers format. If the state_dict is already in
-        diffusers format, then this function will have no effect.
-
-        This function is adapted from:
-        https://github.com/bmaltais/kohya_ss/blob/2accb1305979ba62f5077a23aabac23b4c37e935/networks/lora_diffusers.py#L385-L409
-
-        Args:
-            state_dict (Dict[str, Tensor]): The SDXL LoRA state_dict.
-
-        Raises:
-            ValueError: If state_dict contains an unrecognized key, or not all keys could be converted.
-
-        Returns:
-            Dict[str, Tensor]: The diffusers-format state_dict.
-        """
-        converted_count = 0  # The number of Stability AI keys converted to diffusers format.
-        not_converted_count = 0  # The number of keys that were not converted.
-
-        # Get a sorted list of Stability AI UNet keys so that we can efficiently search for keys with matching prefixes.
-        # For example, we want to efficiently find `input_blocks_4_1` in the list when searching for
-        # `input_blocks_4_1_proj_in`.
-        stability_unet_keys = list(SDXL_UNET_STABILITY_TO_DIFFUSERS_MAP)
-        stability_unet_keys.sort()
-
-        new_state_dict = {}
-        for full_key, value in state_dict.items():
-            if full_key.startswith("lora_unet_"):
-                search_key = full_key.replace("lora_unet_", "")
-                # Use bisect to find the key in stability_unet_keys that *may* match the search_key's prefix.
-                position = bisect.bisect_right(stability_unet_keys, search_key)
-                map_key = stability_unet_keys[position - 1]
-                # Now, check if the map_key *actually* matches the search_key.
-                if search_key.startswith(map_key):
-                    new_key = full_key.replace(map_key, SDXL_UNET_STABILITY_TO_DIFFUSERS_MAP[map_key])
-                    new_state_dict[new_key] = value
-                    converted_count += 1
-                else:
-                    new_state_dict[full_key] = value
-                    not_converted_count += 1
-            elif full_key.startswith("lora_te1_") or full_key.startswith("lora_te2_"):
-                # The CLIP text encoders have the same keys in both Stability AI and diffusers formats.
-                new_state_dict[full_key] = value
-                continue
-            else:
-                raise ValueError(f"Unrecognized SDXL LoRA key prefix: '{full_key}'.")
-
-        if converted_count > 0 and not_converted_count > 0:
-            raise ValueError(
-                f"The SDXL LoRA could only be partially converted to diffusers format. converted={converted_count},"
-                f" not_converted={not_converted_count}"
-            )
-
-        return new_state_dict
-
-    @classmethod
-    def from_checkpoint(
-        cls,
-        file_path: Union[str, Path],
-        device: Optional[torch.device] = None,
-        dtype: Optional[torch.dtype] = None,
-        base_model: Optional[BaseModelType] = None,
-    ) -> Self:
-        device = device or torch.device("cpu")
-        dtype = dtype or torch.float32
-
-        if isinstance(file_path, str):
-            file_path = Path(file_path)
-
-        model = cls(
-            name=file_path.stem,
-            layers={},
-        )
-
-        if file_path.suffix == ".safetensors":
-            sd = load_file(file_path.absolute().as_posix(), device="cpu")
-        else:
-            sd = torch.load(file_path, map_location="cpu")
-
-        state_dict = cls._group_state(sd)
-
-        if base_model == BaseModelType.StableDiffusionXL:
-            state_dict = cls._convert_sdxl_keys_to_diffusers_format(state_dict)
-
-        for layer_key, values in state_dict.items():
-            # Detect layers according to LyCORIS detection logic(`weight_list_det`)
-            # https://github.com/KohakuBlueleaf/LyCORIS/tree/8ad8000efb79e2b879054da8c9356e6143591bad/lycoris/modules
-
-            # lora and locon
-            if "lora_up.weight" in values:
-                layer: AnyLoRALayer = LoRALayer(layer_key, values)
-
-            # loha
-            elif "hada_w1_a" in values:
-                layer = LoHALayer(layer_key, values)
-
-            # lokr
-            elif "lokr_w1" in values or "lokr_w1_a" in values:
-                layer = LoKRLayer(layer_key, values)
-
-            # diff
-            elif "diff" in values:
-                layer = FullLayer(layer_key, values)
-
-            # ia3
-            elif "on_input" in values:
-                layer = IA3Layer(layer_key, values)
-
-            # norms
-            elif "w_norm" in values:
-                layer = NormLayer(layer_key, values)
-
-            else:
-                print(f">> Encountered unknown lora layer module in {model.name}: {layer_key} - {list(values.keys())}")
-                raise Exception("Unknown lora format!")
-
-            # lower memory consumption by removing already parsed layer values
-            state_dict[layer_key].clear()
-
-            layer.to(device=device, dtype=dtype)
-            model.layers[layer_key] = layer
-
-        return model
-
-    @staticmethod
-    def _group_state(state_dict: Dict[str, torch.Tensor]) -> Dict[str, Dict[str, torch.Tensor]]:
-        state_dict_groupped: Dict[str, Dict[str, torch.Tensor]] = {}
-
-        for key, value in state_dict.items():
-            stem, leaf = key.split(".", 1)
-            if stem not in state_dict_groupped:
-                state_dict_groupped[stem] = {}
-            state_dict_groupped[stem][leaf] = value
-
-        return state_dict_groupped
-
-
-# code from
-# https://github.com/bmaltais/kohya_ss/blob/2accb1305979ba62f5077a23aabac23b4c37e935/networks/lora_diffusers.py#L15C1-L97C32
-def make_sdxl_unet_conversion_map() -> List[Tuple[str, str]]:
-    """Create a dict mapping state_dict keys from Stability AI SDXL format to diffusers SDXL format."""
-    unet_conversion_map_layer = []
-
-    for i in range(3):  # num_blocks is 3 in sdxl
-        # loop over downblocks/upblocks
-        for j in range(2):
-            # loop over resnets/attentions for downblocks
-            hf_down_res_prefix = f"down_blocks.{i}.resnets.{j}."
-            sd_down_res_prefix = f"input_blocks.{3*i + j + 1}.0."
-            unet_conversion_map_layer.append((sd_down_res_prefix, hf_down_res_prefix))
-
-            if i < 3:
-                # no attention layers in down_blocks.3
-                hf_down_atn_prefix = f"down_blocks.{i}.attentions.{j}."
-                sd_down_atn_prefix = f"input_blocks.{3*i + j + 1}.1."
-                unet_conversion_map_layer.append((sd_down_atn_prefix, hf_down_atn_prefix))
-
-        for j in range(3):
-            # loop over resnets/attentions for upblocks
-            hf_up_res_prefix = f"up_blocks.{i}.resnets.{j}."
-            sd_up_res_prefix = f"output_blocks.{3*i + j}.0."
-            unet_conversion_map_layer.append((sd_up_res_prefix, hf_up_res_prefix))
-
-            # if i > 0: commentout for sdxl
-            # no attention layers in up_blocks.0
-            hf_up_atn_prefix = f"up_blocks.{i}.attentions.{j}."
-            sd_up_atn_prefix = f"output_blocks.{3*i + j}.1."
-            unet_conversion_map_layer.append((sd_up_atn_prefix, hf_up_atn_prefix))
-
-        if i < 3:
-            # no downsample in down_blocks.3
-            hf_downsample_prefix = f"down_blocks.{i}.downsamplers.0.conv."
-            sd_downsample_prefix = f"input_blocks.{3*(i+1)}.0.op."
-            unet_conversion_map_layer.append((sd_downsample_prefix, hf_downsample_prefix))
-
-            # no upsample in up_blocks.3
-            hf_upsample_prefix = f"up_blocks.{i}.upsamplers.0."
-            sd_upsample_prefix = f"output_blocks.{3*i + 2}.{2}."  # change for sdxl
-            unet_conversion_map_layer.append((sd_upsample_prefix, hf_upsample_prefix))
-
-    hf_mid_atn_prefix = "mid_block.attentions.0."
-    sd_mid_atn_prefix = "middle_block.1."
-    unet_conversion_map_layer.append((sd_mid_atn_prefix, hf_mid_atn_prefix))
-
-    for j in range(2):
-        hf_mid_res_prefix = f"mid_block.resnets.{j}."
-        sd_mid_res_prefix = f"middle_block.{2*j}."
-        unet_conversion_map_layer.append((sd_mid_res_prefix, hf_mid_res_prefix))
-
-    unet_conversion_map_resnet = [
-        # (stable-diffusion, HF Diffusers)
-        ("in_layers.0.", "norm1."),
-        ("in_layers.2.", "conv1."),
-        ("out_layers.0.", "norm2."),
-        ("out_layers.3.", "conv2."),
-        ("emb_layers.1.", "time_emb_proj."),
-        ("skip_connection.", "conv_shortcut."),
-    ]
-
-    unet_conversion_map = []
-    for sd, hf in unet_conversion_map_layer:
-        if "resnets" in hf:
-            for sd_res, hf_res in unet_conversion_map_resnet:
-                unet_conversion_map.append((sd + sd_res, hf + hf_res))
-        else:
-            unet_conversion_map.append((sd, hf))
-
-    for j in range(2):
-        hf_time_embed_prefix = f"time_embedding.linear_{j+1}."
-        sd_time_embed_prefix = f"time_embed.{j*2}."
-        unet_conversion_map.append((sd_time_embed_prefix, hf_time_embed_prefix))
-
-    for j in range(2):
-        hf_label_embed_prefix = f"add_embedding.linear_{j+1}."
-        sd_label_embed_prefix = f"label_emb.0.{j*2}."
-        unet_conversion_map.append((sd_label_embed_prefix, hf_label_embed_prefix))
-
-    unet_conversion_map.append(("input_blocks.0.0.", "conv_in."))
-    unet_conversion_map.append(("out.0.", "conv_norm_out."))
-    unet_conversion_map.append(("out.2.", "conv_out."))
-
-    return unet_conversion_map
-
-
-SDXL_UNET_STABILITY_TO_DIFFUSERS_MAP = {
-    sd.rstrip(".").replace(".", "_"): hf.rstrip(".").replace(".", "_") for sd, hf in make_sdxl_unet_conversion_map()
-}

invokeai/backend/model_manager/config.py

@@ -52,6 +52,7 @@ class BaseModelType(str, Enum):
     StableDiffusion2 = "sd-2"
     StableDiffusionXL = "sdxl"
     StableDiffusionXLRefiner = "sdxl-refiner"
+    Flux = "flux"
     # Kandinsky2_1 = "kandinsky-2.1"
 
 
@@ -66,7 +67,9 @@ class ModelType(str, Enum):
     TextualInversion = "embedding"
     IPAdapter = "ip_adapter"
     CLIPVision = "clip_vision"
+    CLIPEmbed = "clip_embed"
     T2IAdapter = "t2i_adapter"
+    T5Encoder = "t5_encoder"
     SpandrelImageToImage = "spandrel_image_to_image"
 
 
@@ -74,6 +77,7 @@ class SubModelType(str, Enum):
     """Submodel type."""
 
     UNet = "unet"
+    Transformer = "transformer"
     TextEncoder = "text_encoder"
     TextEncoder2 = "text_encoder_2"
     Tokenizer = "tokenizer"
@@ -104,6 +108,9 @@ class ModelFormat(str, Enum):
     EmbeddingFile = "embedding_file"
     EmbeddingFolder = "embedding_folder"
     InvokeAI = "invokeai"
+    T5Encoder = "t5_encoder"
+    BnbQuantizedLlmInt8b = "bnb_quantized_int8b"
+    BnbQuantizednf4b = "bnb_quantized_nf4b"
 
 
 class SchedulerPredictionType(str, Enum):
@@ -186,7 +193,9 @@ class ModelConfigBase(BaseModel):
 class CheckpointConfigBase(ModelConfigBase):
     """Model config for checkpoint-style models."""
 
-    format: Literal[ModelFormat.Checkpoint] = ModelFormat.Checkpoint
+    format: Literal[ModelFormat.Checkpoint, ModelFormat.BnbQuantizednf4b] = Field(
+        description="Format of the provided checkpoint model", default=ModelFormat.Checkpoint
+    )
     config_path: str = Field(description="path to the checkpoint model config file")
     converted_at: Optional[float] = Field(
         description="When this model was last converted to diffusers", default_factory=time.time
@@ -205,6 +214,26 @@ class LoRAConfigBase(ModelConfigBase):
     trigger_phrases: Optional[set[str]] = Field(description="Set of trigger phrases for this model", default=None)
 
 
+class T5EncoderConfigBase(ModelConfigBase):
+    type: Literal[ModelType.T5Encoder] = ModelType.T5Encoder
+
+
+class T5EncoderConfig(T5EncoderConfigBase):
+    format: Literal[ModelFormat.T5Encoder] = ModelFormat.T5Encoder
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.T5Encoder.value}.{ModelFormat.T5Encoder.value}")
+
+
+class T5EncoderBnbQuantizedLlmInt8bConfig(T5EncoderConfigBase):
+    format: Literal[ModelFormat.BnbQuantizedLlmInt8b] = ModelFormat.BnbQuantizedLlmInt8b
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.T5Encoder.value}.{ModelFormat.BnbQuantizedLlmInt8b.value}")
+
+
 class LoRALyCORISConfig(LoRAConfigBase):
     """Model config for LoRA/Lycoris models."""
 
@@ -229,7 +258,6 @@ class VAECheckpointConfig(CheckpointConfigBase):
     """Model config for standalone VAE models."""
 
     type: Literal[ModelType.VAE] = ModelType.VAE
-    format: Literal[ModelFormat.Checkpoint] = ModelFormat.Checkpoint
 
     @staticmethod
     def get_tag() -> Tag:
@@ -268,7 +296,6 @@ class ControlNetCheckpointConfig(CheckpointConfigBase, ControlAdapterConfigBase)
     """Model config for ControlNet models (diffusers version)."""
 
     type: Literal[ModelType.ControlNet] = ModelType.ControlNet
-    format: Literal[ModelFormat.Checkpoint] = ModelFormat.Checkpoint
 
     @staticmethod
     def get_tag() -> Tag:
@@ -317,6 +344,21 @@ class MainCheckpointConfig(CheckpointConfigBase, MainConfigBase):
         return Tag(f"{ModelType.Main.value}.{ModelFormat.Checkpoint.value}")
 
 
+class MainBnbQuantized4bCheckpointConfig(CheckpointConfigBase, MainConfigBase):
+    """Model config for main checkpoint models."""
+
+    prediction_type: SchedulerPredictionType = SchedulerPredictionType.Epsilon
+    upcast_attention: bool = False
+
+    def __init__(self, *args, **kwargs):
+        super().__init__(*args, **kwargs)
+        self.format = ModelFormat.BnbQuantizednf4b
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.Main.value}.{ModelFormat.BnbQuantizednf4b.value}")
+
+
 class MainDiffusersConfig(DiffusersConfigBase, MainConfigBase):
     """Model config for main diffusers models."""
 
@@ -350,6 +392,17 @@ class IPAdapterCheckpointConfig(IPAdapterBaseConfig):
         return Tag(f"{ModelType.IPAdapter.value}.{ModelFormat.Checkpoint.value}")
 
 
+class CLIPEmbedDiffusersConfig(DiffusersConfigBase):
+    """Model config for Clip Embeddings."""
+
+    type: Literal[ModelType.CLIPEmbed] = ModelType.CLIPEmbed
+    format: Literal[ModelFormat.Diffusers] = ModelFormat.Diffusers
+
+    @staticmethod
+    def get_tag() -> Tag:
+        return Tag(f"{ModelType.CLIPEmbed.value}.{ModelFormat.Diffusers.value}")
+
+
 class CLIPVisionDiffusersConfig(DiffusersConfigBase):
     """Model config for CLIPVision."""
 
@@ -408,12 +461,15 @@ AnyModelConfig = Annotated[
     Union[
         Annotated[MainDiffusersConfig, MainDiffusersConfig.get_tag()],
         Annotated[MainCheckpointConfig, MainCheckpointConfig.get_tag()],
+        Annotated[MainBnbQuantized4bCheckpointConfig, MainBnbQuantized4bCheckpointConfig.get_tag()],
         Annotated[VAEDiffusersConfig, VAEDiffusersConfig.get_tag()],
         Annotated[VAECheckpointConfig, VAECheckpointConfig.get_tag()],
         Annotated[ControlNetDiffusersConfig, ControlNetDiffusersConfig.get_tag()],
         Annotated[ControlNetCheckpointConfig, ControlNetCheckpointConfig.get_tag()],
         Annotated[LoRALyCORISConfig, LoRALyCORISConfig.get_tag()],
         Annotated[LoRADiffusersConfig, LoRADiffusersConfig.get_tag()],
+        Annotated[T5EncoderConfig, T5EncoderConfig.get_tag()],
+        Annotated[T5EncoderBnbQuantizedLlmInt8bConfig, T5EncoderBnbQuantizedLlmInt8bConfig.get_tag()],
         Annotated[TextualInversionFileConfig, TextualInversionFileConfig.get_tag()],
         Annotated[TextualInversionFolderConfig, TextualInversionFolderConfig.get_tag()],
         Annotated[IPAdapterInvokeAIConfig, IPAdapterInvokeAIConfig.get_tag()],
@@ -421,6 +477,7 @@ AnyModelConfig = Annotated[
         Annotated[T2IAdapterConfig, T2IAdapterConfig.get_tag()],
         Annotated[SpandrelImageToImageConfig, SpandrelImageToImageConfig.get_tag()],
         Annotated[CLIPVisionDiffusersConfig, CLIPVisionDiffusersConfig.get_tag()],
+        Annotated[CLIPEmbedDiffusersConfig, CLIPEmbedDiffusersConfig.get_tag()],
     ],
     Discriminator(get_model_discriminator_value),
 ]

invokeai/backend/model_manager/load/load_default.py

@@ -66,12 +66,14 @@ class ModelLoader(ModelLoaderBase):
         return (model_base / config.path).resolve()
 
     def _load_and_cache(self, config: AnyModelConfig, submodel_type: Optional[SubModelType] = None) -> ModelLockerBase:
+        stats_name = ":".join([config.base, config.type, config.name, (submodel_type or "")])
         try:
-            return self._ram_cache.get(config.key, submodel_type)
+            return self._ram_cache.get(config.key, submodel_type, stats_name=stats_name)
         except IndexError:
             pass
 
         config.path = str(self._get_model_path(config))
+        self._ram_cache.make_room(self.get_size_fs(config, Path(config.path), submodel_type))
         loaded_model = self._load_model(config, submodel_type)
 
         self._ram_cache.put(
@@ -83,7 +85,7 @@ class ModelLoader(ModelLoaderBase):
         return self._ram_cache.get(
             key=config.key,
             submodel_type=submodel_type,
-            stats_name=":".join([config.base, config.type, config.name, (submodel_type or "")]),
+            stats_name=stats_name,
         )
 
     def get_size_fs(

invokeai/backend/model_manager/load/model_cache/model_cache_base.py

@@ -128,7 +128,24 @@ class ModelCacheBase(ABC, Generic[T]):
     @property
     @abstractmethod
     def max_cache_size(self) -> float:
-        """Return true if the cache is configured to lazily offload models in VRAM."""
+        """Return the maximum size the RAM cache can grow to."""
+        pass
+
+    @max_cache_size.setter
+    @abstractmethod
+    def max_cache_size(self, value: float) -> None:
+        """Set the cap on vram cache size."""
+
+    @property
+    @abstractmethod
+    def max_vram_cache_size(self) -> float:
+        """Return the maximum size the VRAM cache can grow to."""
+        pass
+
+    @max_vram_cache_size.setter
+    @abstractmethod
+    def max_vram_cache_size(self, value: float) -> float:
+        """Set the maximum size the VRAM cache can grow to."""
         pass
 
     @abstractmethod
@@ -193,15 +210,6 @@ class ModelCacheBase(ABC, Generic[T]):
         """
         pass
 
-    @abstractmethod
-    def exists(
-        self,
-        key: str,
-        submodel_type: Optional[SubModelType] = None,
-    ) -> bool:
-        """Return true if the model identified by key and submodel_type is in the cache."""
-        pass
-
     @abstractmethod
     def cache_size(self) -> int:
         """Get the total size of the models currently cached."""

invokeai/backend/model_manager/load/model_cache/model_cache_default.py

@@ -1,22 +1,6 @@
 # Copyright (c) 2024 Lincoln D. Stein and the InvokeAI Development team
 # TODO: Add Stalker's proper name to copyright
-"""
-Manage a RAM cache of diffusion/transformer models for fast switching.
-They are moved between GPU VRAM and CPU RAM as necessary. If the cache
-grows larger than a preset maximum, then the least recently used
-model will be cleared and (re)loaded from disk when next needed.
-
-The cache returns context manager generators designed to load the
-model into the GPU within the context, and unload outside the
-context. Use like this:
-
-   cache = ModelCache(max_cache_size=7.5)
-   with cache.get_model('runwayml/stable-diffusion-1-5') as SD1,
-          cache.get_model('stabilityai/stable-diffusion-2') as SD2:
-       do_something_in_GPU(SD1,SD2)
-
-
-"""
+""" """
 
 import gc
 import math
@@ -40,53 +24,74 @@ from invokeai.backend.model_manager.load.model_util import calc_model_size_by_da
 from invokeai.backend.util.devices import TorchDevice
 from invokeai.backend.util.logging import InvokeAILogger
 
-# Maximum size of the cache, in gigs
-# Default is roughly enough to hold three fp16 diffusers models in RAM simultaneously
-DEFAULT_MAX_CACHE_SIZE = 6.0
-
-# amount of GPU memory to hold in reserve for use by generations (GB)
-DEFAULT_MAX_VRAM_CACHE_SIZE = 2.75
-
-# actual size of a gig
-GIG = 1073741824
+# Size of a GB in bytes.
+GB = 2**30
 
 # Size of a MB in bytes.
 MB = 2**20
 
 
 class ModelCache(ModelCacheBase[AnyModel]):
-    """Implementation of ModelCacheBase."""
+    """A cache for managing models in memory.
+
+    The cache is based on two levels of model storage:
+    - execution_device: The device where most models are executed (typically "cuda", "mps", or "cpu").
+    - storage_device: The device where models are offloaded when not in active use (typically "cpu").
+
+    The model cache is based on the following assumptions:
+    - storage_device_mem_size > execution_device_mem_size
+    - disk_to_storage_device_transfer_time >> storage_device_to_execution_device_transfer_time
+
+    A copy of all models in the cache is always kept on the storage_device. A subset of the models also have a copy on
+    the execution_device.
+
+    Models are moved between the storage_device and the execution_device as necessary. Cache size limits are enforced
+    on both the storage_device and the execution_device. The execution_device cache uses a smallest-first offload
+    policy. The storage_device cache uses a least-recently-used (LRU) offload policy.
+
+    Note: Neither of these offload policies has really been compared against alternatives. It's likely that different
+    policies would be better, although the optimal policies are likely heavily dependent on usage patterns and HW
+    configuration.
+
+    The cache returns context manager generators designed to load the model into the execution device (often GPU) within
+    the context, and unload outside the context.
+
+    Example usage:
+    ```
+    cache = ModelCache(max_cache_size=7.5, max_vram_cache_size=6.0)
+    with cache.get_model('runwayml/stable-diffusion-1-5') as SD1:
+        do_something_on_gpu(SD1)
+    ```
+    """
 
     def __init__(
         self,
-        max_cache_size: float = DEFAULT_MAX_CACHE_SIZE,
-        max_vram_cache_size: float = DEFAULT_MAX_VRAM_CACHE_SIZE,
+        max_cache_size: float,
+        max_vram_cache_size: float,
         execution_device: torch.device = torch.device("cuda"),
         storage_device: torch.device = torch.device("cpu"),
         precision: torch.dtype = torch.float16,
-        sequential_offload: bool = False,
         lazy_offloading: bool = True,
-        sha_chunksize: int = 16777216,
         log_memory_usage: bool = False,
         logger: Optional[Logger] = None,
     ):
         """
         Initialize the model RAM cache.
 
-        :param max_cache_size: Maximum size of the RAM cache [6.0 GB]
+        :param max_cache_size: Maximum size of the storage_device cache in GBs.
+        :param max_vram_cache_size: Maximum size of the execution_device cache in GBs.
         :param execution_device: Torch device to load active model into [torch.device('cuda')]
         :param storage_device: Torch device to save inactive model in [torch.device('cpu')]
         :param precision: Precision for loaded models [torch.float16]
         :param lazy_offloading: Keep model in VRAM until another model needs to be loaded
-        :param sequential_offload: Conserve VRAM by loading and unloading each stage of the pipeline sequentially
         :param log_memory_usage: If True, a memory snapshot will be captured before and after every model cache
             operation, and the result will be logged (at debug level). There is a time cost to capturing the memory
             snapshots, so it is recommended to disable this feature unless you are actively inspecting the model cache's
             behaviour.
+        :param logger: InvokeAILogger to use (otherwise creates one)
         """
         # allow lazy offloading only when vram cache enabled
         self._lazy_offloading = lazy_offloading and max_vram_cache_size > 0
-        self._precision: torch.dtype = precision
         self._max_cache_size: float = max_cache_size
         self._max_vram_cache_size: float = max_vram_cache_size
         self._execution_device: torch.device = execution_device
@@ -128,6 +133,16 @@ class ModelCache(ModelCacheBase[AnyModel]):
         """Set the cap on cache size."""
         self._max_cache_size = value
 
+    @property
+    def max_vram_cache_size(self) -> float:
+        """Return the cap on vram cache size."""
+        return self._max_vram_cache_size
+
+    @max_vram_cache_size.setter
+    def max_vram_cache_size(self, value: float) -> None:
+        """Set the cap on vram cache size."""
+        self._max_vram_cache_size = value
+
     @property
     def stats(self) -> Optional[CacheStats]:
         """Return collected CacheStats object."""
@@ -145,15 +160,6 @@ class ModelCache(ModelCacheBase[AnyModel]):
             total += cache_record.size
         return total
 
-    def exists(
-        self,
-        key: str,
-        submodel_type: Optional[SubModelType] = None,
-    ) -> bool:
-        """Return true if the model identified by key and submodel_type is in the cache."""
-        key = self._make_cache_key(key, submodel_type)
-        return key in self._cached_models
-
     def put(
         self,
         key: str,
@@ -203,7 +209,7 @@ class ModelCache(ModelCacheBase[AnyModel]):
         # more stats
         if self.stats:
             stats_name = stats_name or key
-            self.stats.cache_size = int(self._max_cache_size * GIG)
+            self.stats.cache_size = int(self._max_cache_size * GB)
             self.stats.high_watermark = max(self.stats.high_watermark, self.cache_size())
             self.stats.in_cache = len(self._cached_models)
             self.stats.loaded_model_sizes[stats_name] = max(
@@ -231,10 +237,13 @@ class ModelCache(ModelCacheBase[AnyModel]):
             return model_key
 
     def offload_unlocked_models(self, size_required: int) -> None:
-        """Move any unused models from VRAM."""
-        reserved = self._max_vram_cache_size * GIG
+        """Offload models from the execution_device to make room for size_required.
+
+        :param size_required: The amount of space to clear in the execution_device cache, in bytes.
+        """
+        reserved = self._max_vram_cache_size * GB
         vram_in_use = torch.cuda.memory_allocated() + size_required
-        self.logger.debug(f"{(vram_in_use/GIG):.2f}GB VRAM needed for models; max allowed={(reserved/GIG):.2f}GB")
+        self.logger.debug(f"{(vram_in_use/GB):.2f}GB VRAM needed for models; max allowed={(reserved/GB):.2f}GB")
         for _, cache_entry in sorted(self._cached_models.items(), key=lambda x: x[1].size):
             if vram_in_use <= reserved:
                 break
@@ -245,7 +254,7 @@ class ModelCache(ModelCacheBase[AnyModel]):
                 cache_entry.loaded = False
                 vram_in_use = torch.cuda.memory_allocated() + size_required
                 self.logger.debug(
-                    f"Removing {cache_entry.key} from VRAM to free {(cache_entry.size/GIG):.2f}GB; vram free = {(torch.cuda.memory_allocated()/GIG):.2f}GB"
+                    f"Removing {cache_entry.key} from VRAM to free {(cache_entry.size/GB):.2f}GB; vram free = {(torch.cuda.memory_allocated()/GB):.2f}GB"
                 )
 
         TorchDevice.empty_cache()
@@ -303,7 +312,7 @@ class ModelCache(ModelCacheBase[AnyModel]):
         self.logger.debug(
             f"Moved model '{cache_entry.key}' from {source_device} to"
             f" {target_device} in {(end_model_to_time-start_model_to_time):.2f}s."
-            f"Estimated model size: {(cache_entry.size/GIG):.3f} GB."
+            f"Estimated model size: {(cache_entry.size/GB):.3f} GB."
             f"{get_pretty_snapshot_diff(snapshot_before, snapshot_after)}"
         )
 
@@ -326,14 +335,14 @@ class ModelCache(ModelCacheBase[AnyModel]):
                     f"Moving model '{cache_entry.key}' from {source_device} to"
                     f" {target_device} caused an unexpected change in VRAM usage. The model's"
                     " estimated size may be incorrect. Estimated model size:"
-                    f" {(cache_entry.size/GIG):.3f} GB.\n"
+                    f" {(cache_entry.size/GB):.3f} GB.\n"
                     f"{get_pretty_snapshot_diff(snapshot_before, snapshot_after)}"
                 )
 
     def print_cuda_stats(self) -> None:
         """Log CUDA diagnostics."""
-        vram = "%4.2fG" % (torch.cuda.memory_allocated() / GIG)
-        ram = "%4.2fG" % (self.cache_size() / GIG)
+        vram = "%4.2fG" % (torch.cuda.memory_allocated() / GB)
+        ram = "%4.2fG" % (self.cache_size() / GB)
 
         in_ram_models = 0
         in_vram_models = 0
@@ -353,17 +362,20 @@ class ModelCache(ModelCacheBase[AnyModel]):
                 )
 
     def make_room(self, size: int) -> None:
-        """Make enough room in the cache to accommodate a new model of indicated size."""
-        # calculate how much memory this model will require
-        # multiplier = 2 if self.precision==torch.float32 else 1
+        """Make enough room in the cache to accommodate a new model of indicated size.
+
+        Note: This function deletes all of the cache's internal references to a model in order to free it. If there are
+        external references to the model, there's nothing that the cache can do about it, and those models will not be
+        garbage-collected.
+        """
         bytes_needed = size
-        maximum_size = self.max_cache_size * GIG  # stored in GB, convert to bytes
+        maximum_size = self.max_cache_size * GB  # stored in GB, convert to bytes
         current_size = self.cache_size()
 
         if current_size + bytes_needed > maximum_size:
             self.logger.debug(
-                f"Max cache size exceeded: {(current_size/GIG):.2f}/{self.max_cache_size:.2f} GB, need an additional"
-                f" {(bytes_needed/GIG):.2f} GB"
+                f"Max cache size exceeded: {(current_size/GB):.2f}/{self.max_cache_size:.2f} GB, need an additional"
+                f" {(bytes_needed/GB):.2f} GB"
             )
 
         self.logger.debug(f"Before making_room: cached_models={len(self._cached_models)}")
@@ -380,7 +392,7 @@ class ModelCache(ModelCacheBase[AnyModel]):
 
             if not cache_entry.locked:
                 self.logger.debug(
-                    f"Removing {model_key} from RAM cache to free at least {(size/GIG):.2f} GB (-{(cache_entry.size/GIG):.2f} GB)"
+                    f"Removing {model_key} from RAM cache to free at least {(size/GB):.2f} GB (-{(cache_entry.size/GB):.2f} GB)"
                 )
                 current_size -= cache_entry.size
                 models_cleared += 1

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

@@ -0,0 +1,234 @@
+# Copyright (c) 2024, Brandon W. Rising and the InvokeAI Development Team
+"""Class for Flux model loading in InvokeAI."""
+
+from pathlib import Path
+from typing import Optional
+
+import accelerate
+import torch
+from safetensors.torch import load_file
+from transformers import AutoConfig, AutoModelForTextEncoding, CLIPTextModel, CLIPTokenizer, T5EncoderModel, T5Tokenizer
+
+from invokeai.app.services.config.config_default import get_config
+from invokeai.backend.flux.model import Flux
+from invokeai.backend.flux.modules.autoencoder import AutoEncoder
+from invokeai.backend.flux.util import ae_params, params
+from invokeai.backend.model_manager import (
+    AnyModel,
+    AnyModelConfig,
+    BaseModelType,
+    ModelFormat,
+    ModelType,
+    SubModelType,
+)
+from invokeai.backend.model_manager.config import (
+    CheckpointConfigBase,
+    CLIPEmbedDiffusersConfig,
+    MainBnbQuantized4bCheckpointConfig,
+    MainCheckpointConfig,
+    T5EncoderBnbQuantizedLlmInt8bConfig,
+    T5EncoderConfig,
+    VAECheckpointConfig,
+)
+from invokeai.backend.model_manager.load.load_default import ModelLoader
+from invokeai.backend.model_manager.load.model_loader_registry import ModelLoaderRegistry
+from invokeai.backend.util.silence_warnings import SilenceWarnings
+
+try:
+    from invokeai.backend.quantization.bnb_llm_int8 import quantize_model_llm_int8
+    from invokeai.backend.quantization.bnb_nf4 import quantize_model_nf4
+
+    bnb_available = True
+except ImportError:
+    bnb_available = False
+
+app_config = get_config()
+
+
+@ModelLoaderRegistry.register(base=BaseModelType.Flux, type=ModelType.VAE, format=ModelFormat.Checkpoint)
+class FluxVAELoader(ModelLoader):
+    """Class to load VAE models."""
+
+    def _load_model(
+        self,
+        config: AnyModelConfig,
+        submodel_type: Optional[SubModelType] = None,
+    ) -> AnyModel:
+        if not isinstance(config, VAECheckpointConfig):
+            raise ValueError("Only VAECheckpointConfig models are currently supported here.")
+        model_path = Path(config.path)
+
+        with SilenceWarnings():
+            model = AutoEncoder(ae_params[config.config_path])
+            sd = load_file(model_path)
+            model.load_state_dict(sd, assign=True)
+            model.to(dtype=self._torch_dtype)
+
+        return model
+
+
+@ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.CLIPEmbed, format=ModelFormat.Diffusers)
+class ClipCheckpointModel(ModelLoader):
+    """Class to load main models."""
+
+    def _load_model(
+        self,
+        config: AnyModelConfig,
+        submodel_type: Optional[SubModelType] = None,
+    ) -> AnyModel:
+        if not isinstance(config, CLIPEmbedDiffusersConfig):
+            raise ValueError("Only CLIPEmbedDiffusersConfig models are currently supported here.")
+
+        match submodel_type:
+            case SubModelType.Tokenizer:
+                return CLIPTokenizer.from_pretrained(Path(config.path) / "tokenizer")
+            case SubModelType.TextEncoder:
+                return CLIPTextModel.from_pretrained(Path(config.path) / "text_encoder")
+
+        raise ValueError(
+            f"Only Tokenizer and TextEncoder submodels are currently supported. Received: {submodel_type.value if submodel_type else 'None'}"
+        )
+
+
+@ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.T5Encoder, format=ModelFormat.BnbQuantizedLlmInt8b)
+class BnbQuantizedLlmInt8bCheckpointModel(ModelLoader):
+    """Class to load main models."""
+
+    def _load_model(
+        self,
+        config: AnyModelConfig,
+        submodel_type: Optional[SubModelType] = None,
+    ) -> AnyModel:
+        if not isinstance(config, T5EncoderBnbQuantizedLlmInt8bConfig):
+            raise ValueError("Only T5EncoderBnbQuantizedLlmInt8bConfig models are currently supported here.")
+        if not bnb_available:
+            raise ImportError(
+                "The bnb modules are not available. Please install bitsandbytes if available on your platform."
+            )
+        match submodel_type:
+            case SubModelType.Tokenizer2:
+                return T5Tokenizer.from_pretrained(Path(config.path) / "tokenizer_2", max_length=512)
+            case SubModelType.TextEncoder2:
+                te2_model_path = Path(config.path) / "text_encoder_2"
+                model_config = AutoConfig.from_pretrained(te2_model_path)
+                with accelerate.init_empty_weights():
+                    model = AutoModelForTextEncoding.from_config(model_config)
+                    model = quantize_model_llm_int8(model, modules_to_not_convert=set())
+
+                state_dict_path = te2_model_path / "bnb_llm_int8_model.safetensors"
+                state_dict = load_file(state_dict_path)
+                self._load_state_dict_into_t5(model, state_dict)
+
+                return model
+
+        raise ValueError(
+            f"Only Tokenizer and TextEncoder submodels are currently supported. Received: {submodel_type.value if submodel_type else 'None'}"
+        )
+
+    @classmethod
+    def _load_state_dict_into_t5(cls, model: T5EncoderModel, state_dict: dict[str, torch.Tensor]):
+        # There is a shared reference to a single weight tensor in the model.
+        # Both "encoder.embed_tokens.weight" and "shared.weight" refer to the same tensor, so only the latter should
+        # be present in the state_dict.
+        missing_keys, unexpected_keys = model.load_state_dict(state_dict, strict=False, assign=True)
+        assert len(unexpected_keys) == 0
+        assert set(missing_keys) == {"encoder.embed_tokens.weight"}
+        # Assert that the layers we expect to be shared are actually shared.
+        assert model.encoder.embed_tokens.weight is model.shared.weight
+
+
+@ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.T5Encoder, format=ModelFormat.T5Encoder)
+class T5EncoderCheckpointModel(ModelLoader):
+    """Class to load main models."""
+
+    def _load_model(
+        self,
+        config: AnyModelConfig,
+        submodel_type: Optional[SubModelType] = None,
+    ) -> AnyModel:
+        if not isinstance(config, T5EncoderConfig):
+            raise ValueError("Only T5EncoderConfig models are currently supported here.")
+
+        match submodel_type:
+            case SubModelType.Tokenizer2:
+                return T5Tokenizer.from_pretrained(Path(config.path) / "tokenizer_2", max_length=512)
+            case SubModelType.TextEncoder2:
+                return T5EncoderModel.from_pretrained(Path(config.path) / "text_encoder_2")
+
+        raise ValueError(
+            f"Only Tokenizer and TextEncoder submodels are currently supported. Received: {submodel_type.value if submodel_type else 'None'}"
+        )
+
+
+@ModelLoaderRegistry.register(base=BaseModelType.Flux, 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)
+
+        with SilenceWarnings():
+            model = Flux(params[config.config_path])
+            sd = load_file(model_path)
+            model.load_state_dict(sd, assign=True)
+        return model
+
+
+@ModelLoaderRegistry.register(base=BaseModelType.Flux, type=ModelType.Main, format=ModelFormat.BnbQuantizednf4b)
+class FluxBnbQuantizednf4bCheckpointModel(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, MainBnbQuantized4bCheckpointConfig)
+        if not bnb_available:
+            raise ImportError(
+                "The bnb modules are not available. Please install bitsandbytes if available on your platform."
+            )
+        model_path = Path(config.path)
+
+        with SilenceWarnings():
+            with accelerate.init_empty_weights():
+                model = Flux(params[config.config_path])
+                model = quantize_model_nf4(model, modules_to_not_convert=set(), compute_dtype=torch.bfloat16)
+            sd = load_file(model_path)
+            model.load_state_dict(sd, assign=True)
+        return model

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

@@ -78,7 +78,12 @@ class GenericDiffusersLoader(ModelLoader):
 
     # TO DO: Add exception handling
     def _hf_definition_to_type(self, module: str, class_name: str) -> ModelMixin:  # fix with correct type
-        if module in ["diffusers", "transformers"]:
+        if module in [
+            "diffusers",
+            "transformers",
+            "invokeai.backend.quantization.fast_quantized_transformers_model",
+            "invokeai.backend.quantization.fast_quantized_diffusion_model",
+        ]:
             res_type = sys.modules[module]
         else:
             res_type = sys.modules["diffusers"].pipelines

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

@@ -5,8 +5,10 @@ from logging import Logger
 from pathlib import Path
 from typing import Optional
 
+import torch
+from safetensors.torch import load_file
+
 from invokeai.app.services.config import InvokeAIAppConfig
-from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_manager import (
     AnyModel,
     AnyModelConfig,
@@ -18,6 +20,11 @@ from invokeai.backend.model_manager import (
 from invokeai.backend.model_manager.load.load_default import ModelLoader
 from invokeai.backend.model_manager.load.model_cache.model_cache_base import ModelCacheBase
 from invokeai.backend.model_manager.load.model_loader_registry import ModelLoaderRegistry
+from invokeai.backend.peft.conversions.flux_kohya_lora_conversion_utils import (
+    lora_model_from_flux_kohya_state_dict,
+)
+from invokeai.backend.peft.conversions.sd_lora_conversion_utils import lora_model_from_sd_state_dict
+from invokeai.backend.peft.conversions.sdxl_lora_conversion_utils import convert_sdxl_keys_to_diffusers_format
 
 
 @ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.LoRA, format=ModelFormat.Diffusers)
@@ -45,14 +52,28 @@ class LoRALoader(ModelLoader):
             raise ValueError("There are no submodels in a LoRA model.")
         model_path = Path(config.path)
         assert self._model_base is not None
-        model = LoRAModelRaw.from_checkpoint(
-            file_path=model_path,
-            dtype=self._torch_dtype,
-            base_model=self._model_base,
-        )
+
+        # Load the state dict from the model file.
+        if model_path.suffix == ".safetensors":
+            state_dict = load_file(model_path.absolute().as_posix(), device="cpu")
+        else:
+            state_dict = torch.load(model_path, map_location="cpu")
+
+        # Apply state_dict key conversions, if necessary.
+        if self._model_base == BaseModelType.StableDiffusionXL:
+            state_dict = convert_sdxl_keys_to_diffusers_format(state_dict)
+            model = lora_model_from_sd_state_dict(state_dict=state_dict)
+        elif self._model_base == BaseModelType.Flux:
+            model = lora_model_from_flux_kohya_state_dict(state_dict=state_dict)
+        elif self._model_base in [BaseModelType.StableDiffusion1, BaseModelType.StableDiffusion2]:
+            # Currently, we don't apply any conversions for SD1 and SD2 LoRA models.
+            model = lora_model_from_sd_state_dict(state_dict=state_dict)
+        else:
+            raise ValueError(f"Unsupported LoRA base model: {self._model_base}")
+
+        model.to(dtype=self._torch_dtype)
         return model
 
-    # override
     def _get_model_path(self, config: AnyModelConfig) -> Path:
         # cheating a little - we remember this variable for using in the subsequent call to _load_model()
         self._model_base = config.base

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

@@ -36,8 +36,18 @@ VARIANT_TO_IN_CHANNEL_MAP = {
 }
 
 
-@ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.Main, format=ModelFormat.Diffusers)
-@ModelLoaderRegistry.register(base=BaseModelType.Any, type=ModelType.Main, format=ModelFormat.Checkpoint)
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion1, type=ModelType.Main, format=ModelFormat.Diffusers)
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion2, type=ModelType.Main, format=ModelFormat.Diffusers)
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusionXL, type=ModelType.Main, format=ModelFormat.Diffusers)
+@ModelLoaderRegistry.register(
+    base=BaseModelType.StableDiffusionXLRefiner, type=ModelType.Main, format=ModelFormat.Diffusers
+)
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion1, type=ModelType.Main, format=ModelFormat.Checkpoint)
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusion2, type=ModelType.Main, format=ModelFormat.Checkpoint)
+@ModelLoaderRegistry.register(base=BaseModelType.StableDiffusionXL, type=ModelType.Main, format=ModelFormat.Checkpoint)
+@ModelLoaderRegistry.register(
+    base=BaseModelType.StableDiffusionXLRefiner, type=ModelType.Main, format=ModelFormat.Checkpoint
+)
 class StableDiffusionDiffusersModel(GenericDiffusersLoader):
     """Class to load main models."""
 

invokeai/backend/model_manager/load/model_util.py

@@ -9,15 +9,15 @@ from typing import Optional
 import torch
 from diffusers.pipelines.pipeline_utils import DiffusionPipeline
 from diffusers.schedulers.scheduling_utils import SchedulerMixin
-from transformers import CLIPTokenizer
+from transformers import CLIPTokenizer, T5Tokenizer, T5TokenizerFast
 
 from invokeai.backend.image_util.depth_anything.depth_anything_pipeline import DepthAnythingPipeline
 from invokeai.backend.image_util.grounding_dino.grounding_dino_pipeline import GroundingDinoPipeline
 from invokeai.backend.image_util.segment_anything.segment_anything_pipeline import SegmentAnythingPipeline
 from invokeai.backend.ip_adapter.ip_adapter import IPAdapter
-from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_manager.config import AnyModel
 from invokeai.backend.onnx.onnx_runtime import IAIOnnxRuntimeModel
+from invokeai.backend.peft.lora import LoRAModelRaw
 from invokeai.backend.spandrel_image_to_image_model import SpandrelImageToImageModel
 from invokeai.backend.textual_inversion import TextualInversionModelRaw
 
@@ -50,6 +50,17 @@ def calc_model_size_by_data(logger: logging.Logger, model: AnyModel) -> int:
         ),
     ):
         return model.calc_size()
+    elif isinstance(
+        model,
+        (
+            T5TokenizerFast,
+            T5Tokenizer,
+        ),
+    ):
+        # HACK(ryand): len(model) just returns the vocabulary size, so this is blatantly wrong. It should be small
+        # relative to the text encoder that it's used with, so shouldn't matter too much, but we should fix this at some
+        # point.
+        return len(model)
     else:
         # TODO(ryand): Promote this from a log to an exception once we are confident that we are handling all of the
         # supported model types.

invokeai/backend/model_manager/probe.py

@@ -26,6 +26,7 @@ from invokeai.backend.model_manager.config import (
     SchedulerPredictionType,
 )
 from invokeai.backend.model_manager.util.model_util import lora_token_vector_length, read_checkpoint_meta
+from invokeai.backend.peft.conversions.flux_kohya_lora_conversion_utils import is_state_dict_likely_in_flux_kohya_format
 from invokeai.backend.spandrel_image_to_image_model import SpandrelImageToImageModel
 from invokeai.backend.util.silence_warnings import SilenceWarnings
 
@@ -95,6 +96,7 @@ class ModelProbe(object):
     }
 
     CLASS2TYPE = {
+        "FluxPipeline": ModelType.Main,
         "StableDiffusionPipeline": ModelType.Main,
         "StableDiffusionInpaintPipeline": ModelType.Main,
         "StableDiffusionXLPipeline": ModelType.Main,
@@ -106,6 +108,7 @@ class ModelProbe(object):
         "ControlNetModel": ModelType.ControlNet,
         "CLIPVisionModelWithProjection": ModelType.CLIPVision,
         "T2IAdapter": ModelType.T2IAdapter,
+        "CLIPModel": ModelType.CLIPEmbed,
     }
 
     @classmethod
@@ -161,7 +164,7 @@ class ModelProbe(object):
         fields["description"] = (
             fields.get("description") or f"{fields['base'].value} {model_type.value} model {fields['name']}"
         )
-        fields["format"] = fields.get("format") or probe.get_format()
+        fields["format"] = ModelFormat(fields.get("format")) if "format" in fields else probe.get_format()
         fields["hash"] = fields.get("hash") or ModelHash(algorithm=hash_algo).hash(model_path)
 
         fields["default_settings"] = fields.get("default_settings")
@@ -176,10 +179,10 @@ class ModelProbe(object):
             fields["repo_variant"] = fields.get("repo_variant") or probe.get_repo_variant()
 
         # additional fields needed for main and controlnet models
-        if (
-            fields["type"] in [ModelType.Main, ModelType.ControlNet, ModelType.VAE]
-            and fields["format"] is ModelFormat.Checkpoint
-        ):
+        if fields["type"] in [ModelType.Main, ModelType.ControlNet, ModelType.VAE] and fields["format"] in [
+            ModelFormat.Checkpoint,
+            ModelFormat.BnbQuantizednf4b,
+        ]:
             ckpt_config_path = cls._get_checkpoint_config_path(
                 model_path,
                 model_type=fields["type"],
@@ -222,7 +225,8 @@ class ModelProbe(object):
         ckpt = ckpt.get("state_dict", ckpt)
 
         for key in [str(k) for k in ckpt.keys()]:
-            if key.startswith(("cond_stage_model.", "first_stage_model.", "model.diffusion_model.")):
+            if key.startswith(("cond_stage_model.", "first_stage_model.", "model.diffusion_model.", "double_blocks.")):
+                # Keys starting with double_blocks are associated with Flux models
                 return ModelType.Main
             elif key.startswith(("encoder.conv_in", "decoder.conv_in")):
                 return ModelType.VAE
@@ -321,10 +325,27 @@ class ModelProbe(object):
             return possible_conf.absolute()
 
         if model_type is ModelType.Main:
-            config_file = LEGACY_CONFIGS[base_type][variant_type]
-            if isinstance(config_file, dict):  # need another tier for sd-2.x models
-                config_file = config_file[prediction_type]
-            config_file = f"stable-diffusion/{config_file}"
+            if base_type == BaseModelType.Flux:
+                # TODO: Decide between dev/schnell
+                checkpoint = ModelProbe._scan_and_load_checkpoint(model_path)
+                state_dict = checkpoint.get("state_dict") or checkpoint
+                if "guidance_in.out_layer.weight" in state_dict:
+                    # For flux, this is a key in invokeai.backend.flux.util.params
+                    #   Due to model type and format being the descriminator for model configs this
+                    #   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-dev"
+                else:
+                    # For flux, this is a key in invokeai.backend.flux.util.params
+                    #   Due to model type and format being the descriminator for model configs this
+                    #   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"
+            else:
+                config_file = LEGACY_CONFIGS[base_type][variant_type]
+                if isinstance(config_file, dict):  # need another tier for sd-2.x models
+                    config_file = config_file[prediction_type]
+                config_file = f"stable-diffusion/{config_file}"
         elif model_type is ModelType.ControlNet:
             config_file = (
                 "controlnet/cldm_v15.yaml"
@@ -333,7 +354,13 @@ class ModelProbe(object):
             )
         elif model_type is ModelType.VAE:
             config_file = (
-                "stable-diffusion/v1-inference.yaml"
+                # For flux, this is a key in invokeai.backend.flux.util.ae_params
+                #   Due to model type and format being the descriminator for model configs this
+                #   is used rather than attempting to support flux with separate model types and format
+                #   If changed in the future, please fix me
+                "flux"
+                if base_type is BaseModelType.Flux
+                else "stable-diffusion/v1-inference.yaml"
                 if base_type is BaseModelType.StableDiffusion1
                 else "stable-diffusion/sd_xl_base.yaml"
                 if base_type is BaseModelType.StableDiffusionXL
@@ -416,11 +443,15 @@ class CheckpointProbeBase(ProbeBase):
         self.checkpoint = ModelProbe._scan_and_load_checkpoint(model_path)
 
     def get_format(self) -> ModelFormat:
+        state_dict = self.checkpoint.get("state_dict") or self.checkpoint
+        if "double_blocks.0.img_attn.proj.weight.quant_state.bitsandbytes__nf4" in state_dict:
+            return ModelFormat.BnbQuantizednf4b
         return ModelFormat("checkpoint")
 
     def get_variant_type(self) -> ModelVariantType:
         model_type = ModelProbe.get_model_type_from_checkpoint(self.model_path, self.checkpoint)
-        if model_type != ModelType.Main:
+        base_type = self.get_base_type()
+        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]
@@ -440,6 +471,8 @@ class PipelineCheckpointProbe(CheckpointProbeBase):
     def get_base_type(self) -> BaseModelType:
         checkpoint = self.checkpoint
         state_dict = self.checkpoint.get("state_dict") or checkpoint
+        if "double_blocks.0.img_attn.norm.key_norm.scale" in state_dict:
+            return BaseModelType.Flux
         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
@@ -482,6 +515,7 @@ class VaeCheckpointProbe(CheckpointProbeBase):
             (r"xl", BaseModelType.StableDiffusionXL),
             (r"sd2", BaseModelType.StableDiffusion2),
             (r"vae", BaseModelType.StableDiffusion1),
+            (r"FLUX.1-schnell_ae", BaseModelType.Flux),
         ]:
             if re.search(regexp, self.model_path.name, re.IGNORECASE):
                 return basetype
@@ -495,9 +529,11 @@ class LoRACheckpointProbe(CheckpointProbeBase):
         return ModelFormat("lycoris")
 
     def get_base_type(self) -> BaseModelType:
-        checkpoint = self.checkpoint
-        token_vector_length = lora_token_vector_length(checkpoint)
+        if is_state_dict_likely_in_flux_kohya_format(self.checkpoint):
+            return BaseModelType.Flux
 
+        # If we've gotten here, we assume that the model is a Stable Diffusion model.
+        token_vector_length = lora_token_vector_length(self.checkpoint)
         if token_vector_length == 768:
             return BaseModelType.StableDiffusion1
         elif token_vector_length == 1024:
@@ -713,6 +749,11 @@ class TextualInversionFolderProbe(FolderProbeBase):
         return TextualInversionCheckpointProbe(path).get_base_type()
 
 
+class T5EncoderFolderProbe(FolderProbeBase):
+    def get_format(self) -> ModelFormat:
+        return ModelFormat.T5Encoder
+
+
 class ONNXFolderProbe(PipelineFolderProbe):
     def get_base_type(self) -> BaseModelType:
         # Due to the way the installer is set up, the configuration file for safetensors
@@ -805,6 +846,11 @@ class CLIPVisionFolderProbe(FolderProbeBase):
         return BaseModelType.Any
 
 
+class CLIPEmbedFolderProbe(FolderProbeBase):
+    def get_base_type(self) -> BaseModelType:
+        return BaseModelType.Any
+
+
 class SpandrelImageToImageFolderProbe(FolderProbeBase):
     def get_base_type(self) -> BaseModelType:
         raise NotImplementedError()
@@ -835,8 +881,10 @@ ModelProbe.register_probe("diffusers", ModelType.Main, PipelineFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.VAE, VaeFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.LoRA, LoRAFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.TextualInversion, TextualInversionFolderProbe)
+ModelProbe.register_probe("diffusers", ModelType.T5Encoder, T5EncoderFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.ControlNet, ControlNetFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.IPAdapter, IPAdapterFolderProbe)
+ModelProbe.register_probe("diffusers", ModelType.CLIPEmbed, CLIPEmbedFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.CLIPVision, CLIPVisionFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.T2IAdapter, T2IAdapterFolderProbe)
 ModelProbe.register_probe("diffusers", ModelType.SpandrelImageToImage, SpandrelImageToImageFolderProbe)

invokeai/backend/model_manager/starter_models.py

@@ -2,7 +2,7 @@ from typing import Optional
 
 from pydantic import BaseModel
 
-from invokeai.backend.model_manager.config import BaseModelType, ModelType
+from invokeai.backend.model_manager.config import BaseModelType, ModelFormat, ModelType
 
 
 class StarterModelWithoutDependencies(BaseModel):
@@ -11,6 +11,7 @@ class StarterModelWithoutDependencies(BaseModel):
     name: str
     base: BaseModelType
     type: ModelType
+    format: Optional[ModelFormat] = None
     is_installed: bool = False
 
 
@@ -51,10 +52,76 @@ cyberrealistic_negative = StarterModel(
     type=ModelType.TextualInversion,
 )
 
+t5_base_encoder = StarterModel(
+    name="t5_base_encoder",
+    base=BaseModelType.Any,
+    source="InvokeAI/t5-v1_1-xxl::bfloat16",
+    description="T5-XXL text encoder (used in FLUX pipelines). ~8GB",
+    type=ModelType.T5Encoder,
+)
+
+t5_8b_quantized_encoder = StarterModel(
+    name="t5_bnb_int8_quantized_encoder",
+    base=BaseModelType.Any,
+    source="InvokeAI/t5-v1_1-xxl::bnb_llm_int8",
+    description="T5-XXL text encoder with bitsandbytes LLM.int8() quantization (used in FLUX pipelines). ~5GB",
+    type=ModelType.T5Encoder,
+    format=ModelFormat.BnbQuantizedLlmInt8b,
+)
+
+clip_l_encoder = StarterModel(
+    name="clip-vit-large-patch14",
+    base=BaseModelType.Any,
+    source="InvokeAI/clip-vit-large-patch14-text-encoder::bfloat16",
+    description="CLIP-L text encoder (used in FLUX pipelines). ~250MB",
+    type=ModelType.CLIPEmbed,
+)
+
+flux_vae = StarterModel(
+    name="FLUX.1-schnell_ae",
+    base=BaseModelType.Flux,
+    source="black-forest-labs/FLUX.1-schnell::ae.safetensors",
+    description="FLUX VAE compatible with both schnell and dev variants.",
+    type=ModelType.VAE,
+)
+
+
 # List of starter models, displayed on the frontend.
 # The order/sort of this list is not changed by the frontend - set it how you want it here.
 STARTER_MODELS: list[StarterModel] = [
     # region: Main
+    StarterModel(
+        name="FLUX Schnell (Quantized)",
+        base=BaseModelType.Flux,
+        source="InvokeAI/flux_schnell::transformer/bnb_nf4/flux1-schnell-bnb_nf4.safetensors",
+        description="FLUX schnell transformer quantized to bitsandbytes NF4 format. Total size with dependencies: ~12GB",
+        type=ModelType.Main,
+        dependencies=[t5_8b_quantized_encoder, flux_vae, clip_l_encoder],
+    ),
+    StarterModel(
+        name="FLUX Dev (Quantized)",
+        base=BaseModelType.Flux,
+        source="InvokeAI/flux_dev::transformer/bnb_nf4/flux1-dev-bnb_nf4.safetensors",
+        description="FLUX dev transformer quantized to bitsandbytes NF4 format. Total size with dependencies: ~12GB",
+        type=ModelType.Main,
+        dependencies=[t5_8b_quantized_encoder, flux_vae, clip_l_encoder],
+    ),
+    StarterModel(
+        name="FLUX Schnell",
+        base=BaseModelType.Flux,
+        source="InvokeAI/flux_schnell::transformer/base/flux1-schnell.safetensors",
+        description="FLUX schnell transformer in bfloat16. Total size with dependencies: ~33GB",
+        type=ModelType.Main,
+        dependencies=[t5_base_encoder, flux_vae, clip_l_encoder],
+    ),
+    StarterModel(
+        name="FLUX Dev",
+        base=BaseModelType.Flux,
+        source="InvokeAI/flux_dev::transformer/base/flux1-dev.safetensors",
+        description="FLUX dev transformer in bfloat16. Total size with dependencies: ~33GB",
+        type=ModelType.Main,
+        dependencies=[t5_base_encoder, flux_vae, clip_l_encoder],
+    ),
     StarterModel(
         name="CyberRealistic v4.1",
         base=BaseModelType.StableDiffusion1,
@@ -125,6 +192,7 @@ STARTER_MODELS: list[StarterModel] = [
     # endregion
     # region VAE
     sdxl_fp16_vae_fix,
+    flux_vae,
     # endregion
     # region LoRA
     StarterModel(
@@ -450,6 +518,11 @@ STARTER_MODELS: list[StarterModel] = [
         type=ModelType.SpandrelImageToImage,
     ),
     # endregion
+    # region TextEncoders
+    t5_base_encoder,
+    t5_8b_quantized_encoder,
+    clip_l_encoder,
+    # endregion
 ]
 
 assert len(STARTER_MODELS) == len({m.source for m in STARTER_MODELS}), "Duplicate starter models"

invokeai/backend/model_manager/util/select_hf_files.py

@@ -54,6 +54,7 @@ def filter_files(
                 "lora_weights.safetensors",
                 "weights.pb",
                 "onnx_data",
+                "spiece.model",  # Added for `black-forest-labs/FLUX.1-schnell`.
             )
         ):
             paths.append(file)
@@ -62,13 +63,13 @@ def filter_files(
         # downloading random checkpoints that might also be in the repo. However there is no guarantee
         # that a checkpoint doesn't contain "model" in its name, and no guarantee that future diffusers models
         # will adhere to this naming convention, so this is an area to be careful of.
-        elif re.search(r"model(\.[^.]+)?\.(safetensors|bin|onnx|xml|pth|pt|ckpt|msgpack)$", file.name):
+        elif re.search(r"model.*\.(safetensors|bin|onnx|xml|pth|pt|ckpt|msgpack)$", file.name):
             paths.append(file)
 
     # limit search to subfolder if requested
     if subfolder:
         subfolder = root / subfolder
-        paths = [x for x in paths if x.parent == Path(subfolder)]
+        paths = [x for x in paths if Path(subfolder) in x.parents]
 
     # _filter_by_variant uniquifies the paths and returns a set
     return sorted(_filter_by_variant(paths, variant))
@@ -97,7 +98,9 @@ def _filter_by_variant(files: List[Path], variant: ModelRepoVariant) -> Set[Path
             if variant == ModelRepoVariant.Flax:
                 result.add(path)
 
-        elif path.suffix in [".json", ".txt"]:
+        # Note: '.model' was added to support:
+        # https://huggingface.co/black-forest-labs/FLUX.1-schnell/blob/768d12a373ed5cc9ef9a9dea7504dc09fcc14842/tokenizer_2/spiece.model
+        elif path.suffix in [".json", ".txt", ".model"]:
             result.add(path)
 
         elif variant in [
@@ -140,6 +143,23 @@ def _filter_by_variant(files: List[Path], variant: ModelRepoVariant) -> Set[Path
             continue
 
     for candidate_list in subfolder_weights.values():
+        # Check if at least one of the files has the explicit fp16 variant.
+        at_least_one_fp16 = False
+        for candidate in candidate_list:
+            if len(candidate.path.suffixes) == 2 and candidate.path.suffixes[0] == ".fp16":
+                at_least_one_fp16 = True
+                break
+
+        if not at_least_one_fp16:
+            # If none of the candidates in this candidate_list have the explicit fp16 variant label, then this
+            # candidate_list probably doesn't adhere to the variant naming convention that we expected. In this case,
+            # we'll simply keep all the candidates. An example of a model that hits this case is
+            # `black-forest-labs/FLUX.1-schnell` (as of commit 012d2fd).
+            for candidate in candidate_list:
+                result.add(candidate.path)
+
+        # The candidate_list seems to have the expected variant naming convention. We'll select the highest scoring
+        # candidate.
         highest_score_candidate = max(candidate_list, key=lambda candidate: candidate.score)
         if highest_score_candidate:
             result.add(highest_score_candidate.path)

invokeai/backend/model_patcher.py

@@ -13,10 +13,10 @@ from diffusers import OnnxRuntimeModel, UNet2DConditionModel
 from transformers import CLIPTextModel, CLIPTextModelWithProjection, CLIPTokenizer
 
 from invokeai.app.shared.models import FreeUConfig
-from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_manager import AnyModel
 from invokeai.backend.model_manager.load.optimizations import skip_torch_weight_init
 from invokeai.backend.onnx.onnx_runtime import IAIOnnxRuntimeModel
+from invokeai.backend.peft.lora import LoRAModelRaw
 from invokeai.backend.stable_diffusion.extensions.lora import LoRAExt
 from invokeai.backend.textual_inversion import TextualInversionManager, TextualInversionModelRaw
 from invokeai.backend.util.original_weights_storage import OriginalWeightsStorage

invokeai/backend/peft/conversions/flux_kohya_lora_conversion_utils.py

@@ -0,0 +1,84 @@
+import re
+from typing import Any, Dict, TypeVar
+
+import torch
+
+from invokeai.backend.peft.layers.any_lora_layer import AnyLoRALayer
+from invokeai.backend.peft.layers.utils import peft_layer_from_state_dict
+from invokeai.backend.peft.lora import LoRAModelRaw
+
+# A regex pattern that matches all of the keys in the Kohya FLUX LoRA format.
+# Example keys:
+#   lora_unet_double_blocks_0_img_attn_proj.alpha
+#   lora_unet_double_blocks_0_img_attn_proj.lora_down.weight
+#   lora_unet_double_blocks_0_img_attn_proj.lora_up.weight
+FLUX_KOHYA_KEY_REGEX = (
+    r"lora_unet_(\w+_blocks)_(\d+)_(img_attn|img_mlp|img_mod|txt_attn|txt_mlp|txt_mod|linear1|linear2|modulation)_?(.*)"
+)
+
+
+def is_state_dict_likely_in_flux_kohya_format(state_dict: Dict[str, Any]) -> bool:
+    """Checks if the provided state dict is likely in the Kohya FLUX LoRA format.
+
+    This is intended to be a high-precision detector, but it is not guaranteed to have perfect precision. (A
+    perfect-precision detector would require checking all keys against a whitelist and verifying tensor shapes.)
+    """
+    for k in state_dict.keys():
+        if not re.match(FLUX_KOHYA_KEY_REGEX, k):
+            return False
+    return True
+
+
+def lora_model_from_flux_kohya_state_dict(state_dict: Dict[str, torch.Tensor]) -> LoRAModelRaw:
+    # Group keys by layer.
+    grouped_state_dict: dict[str, dict[str, torch.Tensor]] = {}
+    for key, value in state_dict.items():
+        layer_name, param_name = key.split(".", 1)
+        if layer_name not in grouped_state_dict:
+            grouped_state_dict[layer_name] = {}
+        grouped_state_dict[layer_name][param_name] = value
+
+    # Convert the state dict to the InvokeAI format.
+    grouped_state_dict = convert_flux_kohya_state_dict_to_invoke_format(grouped_state_dict)
+
+    # Create LoRA layers.
+    layers: dict[str, AnyLoRALayer] = {}
+    for layer_key, layer_state_dict in grouped_state_dict.items():
+        layer = peft_layer_from_state_dict(layer_key, layer_state_dict)
+        layers[layer_key] = layer
+
+    # Create and return the LoRAModelRaw.
+    return LoRAModelRaw(layers=layers)
+
+
+T = TypeVar("T")
+
+
+def convert_flux_kohya_state_dict_to_invoke_format(state_dict: Dict[str, T]) -> Dict[str, T]:
+    """Converts a state dict from the Kohya FLUX LoRA format to LoRA weight format used internally by InvokeAI.
+
+    Example key conversions:
+    "lora_unet_double_blocks_0_img_attn_proj" -> "double_blocks.0.img_attn.proj"
+    "lora_unet_double_blocks_0_img_attn_proj" -> "double_blocks.0.img_attn.proj"
+    "lora_unet_double_blocks_0_img_attn_proj" -> "double_blocks.0.img_attn.proj"
+    "lora_unet_double_blocks_0_img_attn_qkv" -> "double_blocks.0.img_attn.qkv"
+    "lora_unet_double_blocks_0_img_attn_qkv" -> "double_blocks.0.img.attn.qkv"
+    "lora_unet_double_blocks_0_img_attn_qkv" -> "double_blocks.0.img.attn.qkv"
+    """
+
+    def replace_func(match: re.Match[str]) -> str:
+        s = f"{match.group(1)}.{match.group(2)}.{match.group(3)}"
+        if match.group(4):
+            s += f".{match.group(4)}"
+        return s
+
+    converted_dict: dict[str, T] = {}
+    for k, v in state_dict.items():
+        match = re.match(FLUX_KOHYA_KEY_REGEX, k)
+        if match:
+            new_key = re.sub(FLUX_KOHYA_KEY_REGEX, replace_func, k)
+            converted_dict[new_key] = v
+        else:
+            raise ValueError(f"Key '{k}' does not match the expected pattern for FLUX LoRA weights.")
+
+    return converted_dict

invokeai/backend/peft/conversions/sd_lora_conversion_utils.py

@@ -0,0 +1,30 @@
+from typing import Dict
+
+import torch
+
+from invokeai.backend.peft.layers.any_lora_layer import AnyLoRALayer
+from invokeai.backend.peft.layers.utils import peft_layer_from_state_dict
+from invokeai.backend.peft.lora import LoRAModelRaw
+
+
+def lora_model_from_sd_state_dict(state_dict: Dict[str, torch.Tensor]) -> LoRAModelRaw:
+    grouped_state_dict: dict[str, dict[str, torch.Tensor]] = _group_state(state_dict)
+
+    layers: dict[str, AnyLoRALayer] = {}
+    for layer_key, values in grouped_state_dict.items():
+        layer = peft_layer_from_state_dict(layer_key, values)
+        layers[layer_key] = layer
+
+    return LoRAModelRaw(layers=layers)
+
+
+def _group_state(state_dict: Dict[str, torch.Tensor]) -> Dict[str, Dict[str, torch.Tensor]]:
+    state_dict_groupped: Dict[str, Dict[str, torch.Tensor]] = {}
+
+    for key, value in state_dict.items():
+        stem, leaf = key.split(".", 1)
+        if stem not in state_dict_groupped:
+            state_dict_groupped[stem] = {}
+        state_dict_groupped[stem][leaf] = value
+
+    return state_dict_groupped

invokeai/backend/peft/conversions/sdxl_lora_conversion_utils.py

@@ -0,0 +1,154 @@
+import bisect
+from typing import Dict, List, Tuple, TypeVar
+
+T = TypeVar("T")
+
+
+def convert_sdxl_keys_to_diffusers_format(state_dict: Dict[str, T]) -> dict[str, T]:
+    """Convert the keys of an SDXL LoRA state_dict to diffusers format.
+
+    The input state_dict can be in either Stability AI format or diffusers format. If the state_dict is already in
+    diffusers format, then this function will have no effect.
+
+    This function is adapted from:
+    https://github.com/bmaltais/kohya_ss/blob/2accb1305979ba62f5077a23aabac23b4c37e935/networks/lora_diffusers.py#L385-L409
+
+    Args:
+        state_dict (Dict[str, Tensor]): The SDXL LoRA state_dict.
+
+    Raises:
+        ValueError: If state_dict contains an unrecognized key, or not all keys could be converted.
+
+    Returns:
+        Dict[str, Tensor]: The diffusers-format state_dict.
+    """
+    converted_count = 0  # The number of Stability AI keys converted to diffusers format.
+    not_converted_count = 0  # The number of keys that were not converted.
+
+    # Get a sorted list of Stability AI UNet keys so that we can efficiently search for keys with matching prefixes.
+    # For example, we want to efficiently find `input_blocks_4_1` in the list when searching for
+    # `input_blocks_4_1_proj_in`.
+    stability_unet_keys = list(SDXL_UNET_STABILITY_TO_DIFFUSERS_MAP)
+    stability_unet_keys.sort()
+
+    new_state_dict: dict[str, T] = {}
+    for full_key, value in state_dict.items():
+        if full_key.startswith("lora_unet_"):
+            search_key = full_key.replace("lora_unet_", "")
+            # Use bisect to find the key in stability_unet_keys that *may* match the search_key's prefix.
+            position = bisect.bisect_right(stability_unet_keys, search_key)
+            map_key = stability_unet_keys[position - 1]
+            # Now, check if the map_key *actually* matches the search_key.
+            if search_key.startswith(map_key):
+                new_key = full_key.replace(map_key, SDXL_UNET_STABILITY_TO_DIFFUSERS_MAP[map_key])
+                new_state_dict[new_key] = value
+                converted_count += 1
+            else:
+                new_state_dict[full_key] = value
+                not_converted_count += 1
+        elif full_key.startswith("lora_te1_") or full_key.startswith("lora_te2_"):
+            # The CLIP text encoders have the same keys in both Stability AI and diffusers formats.
+            new_state_dict[full_key] = value
+            continue
+        else:
+            raise ValueError(f"Unrecognized SDXL LoRA key prefix: '{full_key}'.")
+
+    if converted_count > 0 and not_converted_count > 0:
+        raise ValueError(
+            f"The SDXL LoRA could only be partially converted to diffusers format. converted={converted_count},"
+            f" not_converted={not_converted_count}"
+        )
+
+    return new_state_dict
+
+
+# code from
+# https://github.com/bmaltais/kohya_ss/blob/2accb1305979ba62f5077a23aabac23b4c37e935/networks/lora_diffusers.py#L15C1-L97C32
+def _make_sdxl_unet_conversion_map() -> List[Tuple[str, str]]:
+    """Create a dict mapping state_dict keys from Stability AI SDXL format to diffusers SDXL format."""
+    unet_conversion_map_layer: list[tuple[str, str]] = []
+
+    for i in range(3):  # num_blocks is 3 in sdxl
+        # loop over downblocks/upblocks
+        for j in range(2):
+            # loop over resnets/attentions for downblocks
+            hf_down_res_prefix = f"down_blocks.{i}.resnets.{j}."
+            sd_down_res_prefix = f"input_blocks.{3*i + j + 1}.0."
+            unet_conversion_map_layer.append((sd_down_res_prefix, hf_down_res_prefix))
+
+            if i < 3:
+                # no attention layers in down_blocks.3
+                hf_down_atn_prefix = f"down_blocks.{i}.attentions.{j}."
+                sd_down_atn_prefix = f"input_blocks.{3*i + j + 1}.1."
+                unet_conversion_map_layer.append((sd_down_atn_prefix, hf_down_atn_prefix))
+
+        for j in range(3):
+            # loop over resnets/attentions for upblocks
+            hf_up_res_prefix = f"up_blocks.{i}.resnets.{j}."
+            sd_up_res_prefix = f"output_blocks.{3*i + j}.0."
+            unet_conversion_map_layer.append((sd_up_res_prefix, hf_up_res_prefix))
+
+            # if i > 0: commentout for sdxl
+            # no attention layers in up_blocks.0
+            hf_up_atn_prefix = f"up_blocks.{i}.attentions.{j}."
+            sd_up_atn_prefix = f"output_blocks.{3*i + j}.1."
+            unet_conversion_map_layer.append((sd_up_atn_prefix, hf_up_atn_prefix))
+
+        if i < 3:
+            # no downsample in down_blocks.3
+            hf_downsample_prefix = f"down_blocks.{i}.downsamplers.0.conv."
+            sd_downsample_prefix = f"input_blocks.{3*(i+1)}.0.op."
+            unet_conversion_map_layer.append((sd_downsample_prefix, hf_downsample_prefix))
+
+            # no upsample in up_blocks.3
+            hf_upsample_prefix = f"up_blocks.{i}.upsamplers.0."
+            sd_upsample_prefix = f"output_blocks.{3*i + 2}.{2}."  # change for sdxl
+            unet_conversion_map_layer.append((sd_upsample_prefix, hf_upsample_prefix))
+
+    hf_mid_atn_prefix = "mid_block.attentions.0."
+    sd_mid_atn_prefix = "middle_block.1."
+    unet_conversion_map_layer.append((sd_mid_atn_prefix, hf_mid_atn_prefix))
+
+    for j in range(2):
+        hf_mid_res_prefix = f"mid_block.resnets.{j}."
+        sd_mid_res_prefix = f"middle_block.{2*j}."
+        unet_conversion_map_layer.append((sd_mid_res_prefix, hf_mid_res_prefix))
+
+    unet_conversion_map_resnet = [
+        # (stable-diffusion, HF Diffusers)
+        ("in_layers.0.", "norm1."),
+        ("in_layers.2.", "conv1."),
+        ("out_layers.0.", "norm2."),
+        ("out_layers.3.", "conv2."),
+        ("emb_layers.1.", "time_emb_proj."),
+        ("skip_connection.", "conv_shortcut."),
+    ]
+
+    unet_conversion_map: list[tuple[str, str]] = []
+    for sd, hf in unet_conversion_map_layer:
+        if "resnets" in hf:
+            for sd_res, hf_res in unet_conversion_map_resnet:
+                unet_conversion_map.append((sd + sd_res, hf + hf_res))
+        else:
+            unet_conversion_map.append((sd, hf))
+
+    for j in range(2):
+        hf_time_embed_prefix = f"time_embedding.linear_{j+1}."
+        sd_time_embed_prefix = f"time_embed.{j*2}."
+        unet_conversion_map.append((sd_time_embed_prefix, hf_time_embed_prefix))
+
+    for j in range(2):
+        hf_label_embed_prefix = f"add_embedding.linear_{j+1}."
+        sd_label_embed_prefix = f"label_emb.0.{j*2}."
+        unet_conversion_map.append((sd_label_embed_prefix, hf_label_embed_prefix))
+
+    unet_conversion_map.append(("input_blocks.0.0.", "conv_in."))
+    unet_conversion_map.append(("out.0.", "conv_norm_out."))
+    unet_conversion_map.append(("out.2.", "conv_out."))
+
+    return unet_conversion_map
+
+
+SDXL_UNET_STABILITY_TO_DIFFUSERS_MAP = {
+    sd.rstrip(".").replace(".", "_"): hf.rstrip(".").replace(".", "_") for sd, hf in _make_sdxl_unet_conversion_map()
+}

invokeai/backend/peft/layers/any_lora_layer.py

@@ -0,0 +1,10 @@
+from typing import Union
+
+from invokeai.backend.peft.layers.full_layer import FullLayer
+from invokeai.backend.peft.layers.ia3_layer import IA3Layer
+from invokeai.backend.peft.layers.loha_layer import LoHALayer
+from invokeai.backend.peft.layers.lokr_layer import LoKRLayer
+from invokeai.backend.peft.layers.lora_layer import LoRALayer
+from invokeai.backend.peft.layers.norm_layer import NormLayer
+
+AnyLoRALayer = Union[LoRALayer, LoHALayer, LoKRLayer, FullLayer, IA3Layer, NormLayer]

invokeai/backend/peft/layers/full_layer.py

@@ -0,0 +1,37 @@
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.lora_layer_base import LoRALayerBase
+
+
+class FullLayer(LoRALayerBase):
+    # bias handled in LoRALayerBase(calc_size, to)
+    # weight: torch.Tensor
+    # bias: Optional[torch.Tensor]
+
+    def __init__(
+        self,
+        layer_key: str,
+        values: Dict[str, torch.Tensor],
+    ):
+        super().__init__(layer_key, values)
+
+        self.weight = values["diff"]
+        self.bias = values.get("diff_b", None)
+
+        self.rank = None  # unscaled
+        self.check_keys(values, {"diff", "diff_b"})
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        return self.weight
+
+    def calc_size(self) -> int:
+        model_size = super().calc_size()
+        model_size += self.weight.nelement() * self.weight.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        super().to(device=device, dtype=dtype)
+
+        self.weight = self.weight.to(device=device, dtype=dtype)

invokeai/backend/peft/layers/ia3_layer.py

@@ -0,0 +1,42 @@
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.lora_layer_base import LoRALayerBase
+
+
+class IA3Layer(LoRALayerBase):
+    # weight: torch.Tensor
+    # on_input: torch.Tensor
+
+    def __init__(
+        self,
+        layer_key: str,
+        values: Dict[str, torch.Tensor],
+    ):
+        super().__init__(layer_key, values)
+
+        self.weight = values["weight"]
+        self.on_input = values["on_input"]
+
+        self.rank = None  # unscaled
+        self.check_keys(values, {"weight", "on_input"})
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        weight = self.weight
+        if not self.on_input:
+            weight = weight.reshape(-1, 1)
+        assert orig_weight is not None
+        return orig_weight * weight
+
+    def calc_size(self) -> int:
+        model_size = super().calc_size()
+        model_size += self.weight.nelement() * self.weight.element_size()
+        model_size += self.on_input.nelement() * self.on_input.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None):
+        super().to(device=device, dtype=dtype)
+
+        self.weight = self.weight.to(device=device, dtype=dtype)
+        self.on_input = self.on_input.to(device=device, dtype=dtype)

invokeai/backend/peft/layers/loha_layer.py

@@ -0,0 +1,68 @@
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.lora_layer_base import LoRALayerBase
+
+
+class LoHALayer(LoRALayerBase):
+    # w1_a: torch.Tensor
+    # w1_b: torch.Tensor
+    # w2_a: torch.Tensor
+    # w2_b: torch.Tensor
+    # t1: Optional[torch.Tensor] = None
+    # t2: Optional[torch.Tensor] = None
+
+    def __init__(self, layer_key: str, values: Dict[str, torch.Tensor]):
+        super().__init__(layer_key, values)
+
+        self.w1_a = values["hada_w1_a"]
+        self.w1_b = values["hada_w1_b"]
+        self.w2_a = values["hada_w2_a"]
+        self.w2_b = values["hada_w2_b"]
+        self.t1 = values.get("hada_t1", None)
+        self.t2 = values.get("hada_t2", None)
+
+        self.rank = self.w1_b.shape[0]
+        self.check_keys(
+            values,
+            {
+                "hada_w1_a",
+                "hada_w1_b",
+                "hada_w2_a",
+                "hada_w2_b",
+                "hada_t1",
+                "hada_t2",
+            },
+        )
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        if self.t1 is None:
+            weight: torch.Tensor = (self.w1_a @ self.w1_b) * (self.w2_a @ self.w2_b)
+
+        else:
+            rebuild1 = torch.einsum("i j k l, j r, i p -> p r k l", self.t1, self.w1_b, self.w1_a)
+            rebuild2 = torch.einsum("i j k l, j r, i p -> p r k l", self.t2, self.w2_b, self.w2_a)
+            weight = rebuild1 * rebuild2
+
+        return weight
+
+    def calc_size(self) -> int:
+        model_size = super().calc_size()
+        for val in [self.w1_a, self.w1_b, self.w2_a, self.w2_b, self.t1, self.t2]:
+            if val is not None:
+                model_size += val.nelement() * val.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        super().to(device=device, dtype=dtype)
+
+        self.w1_a = self.w1_a.to(device=device, dtype=dtype)
+        self.w1_b = self.w1_b.to(device=device, dtype=dtype)
+        if self.t1 is not None:
+            self.t1 = self.t1.to(device=device, dtype=dtype)
+
+        self.w2_a = self.w2_a.to(device=device, dtype=dtype)
+        self.w2_b = self.w2_b.to(device=device, dtype=dtype)
+        if self.t2 is not None:
+            self.t2 = self.t2.to(device=device, dtype=dtype)

invokeai/backend/peft/layers/lokr_layer.py

@@ -0,0 +1,114 @@
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.lora_layer_base import LoRALayerBase
+
+
+class LoKRLayer(LoRALayerBase):
+    # w1: Optional[torch.Tensor] = None
+    # w1_a: Optional[torch.Tensor] = None
+    # w1_b: Optional[torch.Tensor] = None
+    # w2: Optional[torch.Tensor] = None
+    # w2_a: Optional[torch.Tensor] = None
+    # w2_b: Optional[torch.Tensor] = None
+    # t2: Optional[torch.Tensor] = None
+
+    def __init__(
+        self,
+        layer_key: str,
+        values: Dict[str, torch.Tensor],
+    ):
+        super().__init__(layer_key, values)
+
+        self.w1 = values.get("lokr_w1", None)
+        if self.w1 is None:
+            self.w1_a = values["lokr_w1_a"]
+            self.w1_b = values["lokr_w1_b"]
+        else:
+            self.w1_b = None
+            self.w1_a = None
+
+        self.w2 = values.get("lokr_w2", None)
+        if self.w2 is None:
+            self.w2_a = values["lokr_w2_a"]
+            self.w2_b = values["lokr_w2_b"]
+        else:
+            self.w2_a = None
+            self.w2_b = None
+
+        self.t2 = values.get("lokr_t2", None)
+
+        if self.w1_b is not None:
+            self.rank = self.w1_b.shape[0]
+        elif self.w2_b is not None:
+            self.rank = self.w2_b.shape[0]
+        else:
+            self.rank = None  # unscaled
+
+        self.check_keys(
+            values,
+            {
+                "lokr_w1",
+                "lokr_w1_a",
+                "lokr_w1_b",
+                "lokr_w2",
+                "lokr_w2_a",
+                "lokr_w2_b",
+                "lokr_t2",
+            },
+        )
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        w1: Optional[torch.Tensor] = self.w1
+        if w1 is None:
+            assert self.w1_a is not None
+            assert self.w1_b is not None
+            w1 = self.w1_a @ self.w1_b
+
+        w2 = self.w2
+        if w2 is None:
+            if self.t2 is None:
+                assert self.w2_a is not None
+                assert self.w2_b is not None
+                w2 = self.w2_a @ self.w2_b
+            else:
+                w2 = torch.einsum("i j k l, i p, j r -> p r k l", self.t2, self.w2_a, self.w2_b)
+
+        if len(w2.shape) == 4:
+            w1 = w1.unsqueeze(2).unsqueeze(2)
+        w2 = w2.contiguous()
+        assert w1 is not None
+        assert w2 is not None
+        weight = torch.kron(w1, w2)
+
+        return weight
+
+    def calc_size(self) -> int:
+        model_size = super().calc_size()
+        for val in [self.w1, self.w1_a, self.w1_b, self.w2, self.w2_a, self.w2_b, self.t2]:
+            if val is not None:
+                model_size += val.nelement() * val.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        super().to(device=device, dtype=dtype)
+
+        if self.w1 is not None:
+            self.w1 = self.w1.to(device=device, dtype=dtype)
+        else:
+            assert self.w1_a is not None
+            assert self.w1_b is not None
+            self.w1_a = self.w1_a.to(device=device, dtype=dtype)
+            self.w1_b = self.w1_b.to(device=device, dtype=dtype)
+
+        if self.w2 is not None:
+            self.w2 = self.w2.to(device=device, dtype=dtype)
+        else:
+            assert self.w2_a is not None
+            assert self.w2_b is not None
+            self.w2_a = self.w2_a.to(device=device, dtype=dtype)
+            self.w2_b = self.w2_b.to(device=device, dtype=dtype)
+
+        if self.t2 is not None:
+            self.t2 = self.t2.to(device=device, dtype=dtype)

invokeai/backend/peft/layers/lora_layer.py

@@ -0,0 +1,59 @@
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.lora_layer_base import LoRALayerBase
+
+
+# TODO: find and debug lora/locon with bias
+class LoRALayer(LoRALayerBase):
+    # up: torch.Tensor
+    # mid: Optional[torch.Tensor]
+    # down: torch.Tensor
+
+    def __init__(
+        self,
+        layer_key: str,
+        values: Dict[str, torch.Tensor],
+    ):
+        super().__init__(layer_key, values)
+
+        self.up = values["lora_up.weight"]
+        self.down = values["lora_down.weight"]
+        self.mid = values.get("lora_mid.weight", None)
+
+        self.rank = self.down.shape[0]
+        self.check_keys(
+            values,
+            {
+                "lora_up.weight",
+                "lora_down.weight",
+                "lora_mid.weight",
+            },
+        )
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        if self.mid is not None:
+            up = self.up.reshape(self.up.shape[0], self.up.shape[1])
+            down = self.down.reshape(self.down.shape[0], self.down.shape[1])
+            weight = torch.einsum("m n w h, i m, n j -> i j w h", self.mid, up, down)
+        else:
+            weight = self.up.reshape(self.up.shape[0], -1) @ self.down.reshape(self.down.shape[0], -1)
+
+        return weight
+
+    def calc_size(self) -> int:
+        model_size = super().calc_size()
+        for val in [self.up, self.mid, self.down]:
+            if val is not None:
+                model_size += val.nelement() * val.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        super().to(device=device, dtype=dtype)
+
+        self.up = self.up.to(device=device, dtype=dtype)
+        self.down = self.down.to(device=device, dtype=dtype)
+
+        if self.mid is not None:
+            self.mid = self.mid.to(device=device, dtype=dtype)

invokeai/backend/peft/layers/lora_layer_base.py

@@ -0,0 +1,74 @@
+from typing import Dict, Optional, Set
+
+import torch
+
+import invokeai.backend.util.logging as logger
+
+
+class LoRALayerBase:
+    # rank: Optional[int]
+    # alpha: Optional[float]
+    # bias: Optional[torch.Tensor]
+    # layer_key: str
+
+    # @property
+    # def scale(self):
+    #    return self.alpha / self.rank if (self.alpha and self.rank) else 1.0
+
+    def __init__(
+        self,
+        layer_key: str,
+        values: Dict[str, torch.Tensor],
+    ):
+        if "alpha" in values:
+            self.alpha = values["alpha"].item()
+        else:
+            self.alpha = None
+
+        if "bias_indices" in values and "bias_values" in values and "bias_size" in values:
+            self.bias: Optional[torch.Tensor] = torch.sparse_coo_tensor(
+                values["bias_indices"],
+                values["bias_values"],
+                tuple(values["bias_size"]),
+            )
+
+        else:
+            self.bias = None
+
+        self.rank = None  # set in layer implementation
+        self.layer_key = layer_key
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        raise NotImplementedError()
+
+    def get_bias(self, orig_bias: torch.Tensor) -> Optional[torch.Tensor]:
+        return self.bias
+
+    def get_parameters(self, orig_module: torch.nn.Module) -> Dict[str, torch.Tensor]:
+        params = {"weight": self.get_weight(orig_module.weight)}
+        bias = self.get_bias(orig_module.bias)
+        if bias is not None:
+            params["bias"] = bias
+        return params
+
+    def calc_size(self) -> int:
+        model_size = 0
+        for val in [self.bias]:
+            if val is not None:
+                model_size += val.nelement() * val.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        if self.bias is not None:
+            self.bias = self.bias.to(device=device, dtype=dtype)
+
+    def check_keys(self, values: Dict[str, torch.Tensor], known_keys: Set[str]):
+        """Log a warning if values contains unhandled keys."""
+        # {"alpha", "bias_indices", "bias_values", "bias_size"} are hard-coded, because they are handled by
+        # `LoRALayerBase`. Sub-classes should provide the known_keys that they handled.
+        all_known_keys = known_keys | {"alpha", "bias_indices", "bias_values", "bias_size"}
+        unknown_keys = set(values.keys()) - all_known_keys
+        if unknown_keys:
+            logger.warning(
+                f"Unexpected keys found in LoRA/LyCORIS layer, model might work incorrectly! Keys: {unknown_keys}"
+            )

invokeai/backend/peft/layers/norm_layer.py

@@ -0,0 +1,37 @@
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.lora_layer_base import LoRALayerBase
+
+
+class NormLayer(LoRALayerBase):
+    # bias handled in LoRALayerBase(calc_size, to)
+    # weight: torch.Tensor
+    # bias: Optional[torch.Tensor]
+
+    def __init__(
+        self,
+        layer_key: str,
+        values: Dict[str, torch.Tensor],
+    ):
+        super().__init__(layer_key, values)
+
+        self.weight = values["w_norm"]
+        self.bias = values.get("b_norm", None)
+
+        self.rank = None  # unscaled
+        self.check_keys(values, {"w_norm", "b_norm"})
+
+    def get_weight(self, orig_weight: torch.Tensor) -> torch.Tensor:
+        return self.weight
+
+    def calc_size(self) -> int:
+        model_size = super().calc_size()
+        model_size += self.weight.nelement() * self.weight.element_size()
+        return model_size
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        super().to(device=device, dtype=dtype)
+
+        self.weight = self.weight.to(device=device, dtype=dtype)

invokeai/backend/peft/layers/utils.py

@@ -0,0 +1,33 @@
+from typing import Dict
+
+import torch
+
+from invokeai.backend.peft.layers.any_lora_layer import AnyLoRALayer
+from invokeai.backend.peft.layers.full_layer import FullLayer
+from invokeai.backend.peft.layers.ia3_layer import IA3Layer
+from invokeai.backend.peft.layers.loha_layer import LoHALayer
+from invokeai.backend.peft.layers.lokr_layer import LoKRLayer
+from invokeai.backend.peft.layers.lora_layer import LoRALayer
+from invokeai.backend.peft.layers.norm_layer import NormLayer
+
+
+def peft_layer_from_state_dict(layer_key: str, state_dict: Dict[str, torch.Tensor]) -> AnyLoRALayer:
+    # Detect layers according to LyCORIS detection logic(`weight_list_det`)
+    # https://github.com/KohakuBlueleaf/LyCORIS/tree/8ad8000efb79e2b879054da8c9356e6143591bad/lycoris/modules
+
+    if "lora_up.weight" in state_dict:
+        # LoRA a.k.a LoCon
+        return LoRALayer(layer_key, state_dict)
+    elif "hada_w1_a" in state_dict:
+        return LoHALayer(layer_key, state_dict)
+    elif "lokr_w1" in state_dict or "lokr_w1_a" in state_dict:
+        return LoKRLayer(layer_key, state_dict)
+    elif "diff" in state_dict:
+        # Full a.k.a Diff
+        return FullLayer(layer_key, state_dict)
+    elif "on_input" in state_dict:
+        return IA3Layer(layer_key, state_dict)
+    elif "w_norm" in state_dict:
+        return NormLayer(layer_key, state_dict)
+    else:
+        raise ValueError(f"Unsupported lora format: {state_dict.keys()}")

invokeai/backend/peft/lora.py

@@ -0,0 +1,22 @@
+# Copyright (c) 2024 The InvokeAI Development team
+from typing import Dict, Optional
+
+import torch
+
+from invokeai.backend.peft.layers.any_lora_layer import AnyLoRALayer
+from invokeai.backend.raw_model import RawModel
+
+
+class LoRAModelRaw(RawModel):  # (torch.nn.Module):
+    def __init__(self, layers: Dict[str, AnyLoRALayer]):
+        self.layers = layers
+
+    def to(self, device: Optional[torch.device] = None, dtype: Optional[torch.dtype] = None) -> None:
+        for _key, layer in self.layers.items():
+            layer.to(device=device, dtype=dtype)
+
+    def calc_size(self) -> int:
+        model_size = 0
+        for _, layer in self.layers.items():
+            model_size += layer.calc_size()
+        return model_size

invokeai/backend/peft/peft_patcher.py

@@ -0,0 +1,102 @@
+from contextlib import contextmanager
+from typing import Dict, Iterator, Optional, Tuple
+
+import torch
+
+from invokeai.backend.peft.lora import LoRAModelRaw
+from invokeai.backend.util.devices import TorchDevice
+from invokeai.backend.util.original_weights_storage import OriginalWeightsStorage
+
+
+class PeftPatcher:
+    @classmethod
+    @torch.no_grad()
+    @contextmanager
+    def apply_peft_patches(
+        cls,
+        model: torch.nn.Module,
+        patches: Iterator[Tuple[LoRAModelRaw, float]],
+        prefix: str,
+        cached_weights: Optional[Dict[str, torch.Tensor]] = None,
+    ):
+        """Apply one or more PEFT patches to a model.
+
+        :param model: The model to patch.
+        :param loras: An iterator that returns tuples of PEFT patches and associated weights. An iterator is used so
+            that the PEFT patches do not need to be loaded into memory all at once.
+        :param prefix: The keys in the patches will be filtered to only include weights with this prefix.
+        :cached_weights: Read-only copy of the model's state dict in CPU, for efficient unpatching purposes.
+        """
+        original_weights = OriginalWeightsStorage(cached_weights)
+        try:
+            for patch, patch_weight in patches:
+                cls._apply_peft_patch(
+                    model=model,
+                    prefix=prefix,
+                    patch=patch,
+                    patch_weight=patch_weight,
+                    original_weights=original_weights,
+                )
+
+            yield
+        finally:
+            for param_key, weight in original_weights.get_changed_weights():
+                model.get_parameter(param_key).copy_(weight)
+
+    @classmethod
+    @torch.no_grad()
+    def _apply_peft_patch(
+        cls,
+        model: torch.nn.Module,
+        prefix: str,
+        patch: LoRAModelRaw,
+        patch_weight: float,
+        original_weights: OriginalWeightsStorage,
+    ):
+        """
+        Apply one a LoRA to a model.
+        :param model: The model to patch.
+        :param patch: LoRA model to patch in.
+        :param patch_weight: LoRA patch weight.
+        :param prefix: A string prefix that precedes keys used in the LoRAs weight layers.
+        :param original_weights: Storage with original weights, filled by weights which lora patches, used for unpatching.
+        """
+
+        if patch_weight == 0:
+            return
+
+        for layer_key, layer in patch.layers.items():
+            if not layer_key.startswith(prefix):
+                continue
+
+            module = model.get_submodule(layer_key)
+
+            # All of the LoRA weight calculations will be done on the same device as the module weight.
+            # (Performance will be best if this is a CUDA device.)
+            device = module.weight.device
+            dtype = module.weight.dtype
+
+            layer_scale = layer.alpha / layer.rank if (layer.alpha and layer.rank) else 1.0
+
+            # We intentionally move to the target device first, then cast. Experimentally, this was found to
+            # be significantly faster for 16-bit CPU tensors being moved to a CUDA device than doing the
+            # same thing in a single call to '.to(...)'.
+            layer.to(device=device)
+            layer.to(dtype=torch.float32)
+
+            # TODO(ryand): Using torch.autocast(...) over explicit casting may offer a speed benefit on CUDA
+            # devices here. Experimentally, it was found to be very slow on CPU. More investigation needed.
+            for param_name, lora_param_weight in layer.get_parameters(module).items():
+                param_key = layer_key + "." + param_name
+                module_param = module.get_parameter(param_name)
+
+                # Save original weight
+                original_weights.save(param_key, module_param)
+
+                if module_param.shape != lora_param_weight.shape:
+                    lora_param_weight = lora_param_weight.reshape(module_param.shape)
+
+                lora_param_weight *= patch_weight * layer_scale
+                module_param += lora_param_weight.to(dtype=dtype)
+
+            layer.to(device=TorchDevice.CPU_DEVICE)

invokeai/backend/quantization/bnb_llm_int8.py

@@ -0,0 +1,135 @@
+import bitsandbytes as bnb
+import torch
+
+# This file contains utils for working with models that use bitsandbytes LLM.int8() quantization.
+# The utils in this file are partially inspired by:
+# https://github.com/Lightning-AI/pytorch-lightning/blob/1551a16b94f5234a4a78801098f64d0732ef5cb5/src/lightning/fabric/plugins/precision/bitsandbytes.py
+
+
+# NOTE(ryand): All of the custom state_dict manipulation logic in this file is pretty hacky. This could be made much
+# cleaner by re-implementing bnb.nn.Linear8bitLt with proper use of buffers and less magic. But, for now, we try to
+# stick close to the bitsandbytes classes to make interoperability easier with other models that might use bitsandbytes.
+
+
+class InvokeInt8Params(bnb.nn.Int8Params):
+    """We override cuda() to avoid re-quantizing the weights in the following cases:
+    - We loaded quantized weights from a state_dict on the cpu, and then moved the model to the gpu.
+    - We are moving the model back-and-forth between the cpu and gpu.
+    """
+
+    def cuda(self, device):
+        if self.has_fp16_weights:
+            return super().cuda(device)
+        elif self.CB is not None and self.SCB is not None:
+            self.data = self.data.cuda()
+            self.CB = self.data
+            self.SCB = self.SCB.cuda()
+        else:
+            # we store the 8-bit rows-major weight
+            # we convert this weight to the turning/ampere weight during the first inference pass
+            B = self.data.contiguous().half().cuda(device)
+            CB, CBt, SCB, SCBt, coo_tensorB = bnb.functional.double_quant(B)
+            del CBt
+            del SCBt
+            self.data = CB
+            self.CB = CB
+            self.SCB = SCB
+
+        return self
+
+
+class InvokeLinear8bitLt(bnb.nn.Linear8bitLt):
+    def _load_from_state_dict(
+        self,
+        state_dict: dict[str, torch.Tensor],
+        prefix: str,
+        local_metadata,
+        strict,
+        missing_keys,
+        unexpected_keys,
+        error_msgs,
+    ):
+        weight = state_dict.pop(prefix + "weight")
+        bias = state_dict.pop(prefix + "bias", None)
+
+        # See `bnb.nn.Linear8bitLt._save_to_state_dict()` for the serialization logic of SCB and weight_format.
+        scb = state_dict.pop(prefix + "SCB", None)
+
+        # Currently, we only support weight_format=0.
+        weight_format = state_dict.pop(prefix + "weight_format", None)
+        assert weight_format == 0
+
+        # TODO(ryand): Technically, we should be using `strict`, `missing_keys`, `unexpected_keys`, and `error_msgs`
+        # rather than raising an exception to correctly implement this API.
+        assert len(state_dict) == 0
+
+        if scb is not None:
+            # We are loading a pre-quantized state dict.
+            self.weight = InvokeInt8Params(
+                data=weight,
+                requires_grad=self.weight.requires_grad,
+                has_fp16_weights=False,
+                # Note: After quantization, CB is the same as weight.
+                CB=weight,
+                SCB=scb,
+            )
+            self.bias = bias if bias is None else torch.nn.Parameter(bias)
+        else:
+            # We are loading a non-quantized state dict.
+
+            # We could simply call the `super()._load_from_state_dict()` method here, but then we wouldn't be able to
+            # load from a state_dict into a model on the "meta" device. Attempting to load into a model on the "meta"
+            # device requires setting `assign=True`, doing this with the default `super()._load_from_state_dict()`
+            # implementation causes `Params4Bit` to be replaced by a `torch.nn.Parameter`. By initializing a new
+            # `Params4bit` object, we work around this issue. It's a bit hacky, but it gets the job done.
+            self.weight = InvokeInt8Params(
+                data=weight,
+                requires_grad=self.weight.requires_grad,
+                has_fp16_weights=False,
+                CB=None,
+                SCB=None,
+            )
+            self.bias = bias if bias is None else torch.nn.Parameter(bias)
+
+        # Reset the state. The persisted fields are based on the initialization behaviour in
+        # `bnb.nn.Linear8bitLt.__init__()`.
+        new_state = bnb.MatmulLtState()
+        new_state.threshold = self.state.threshold
+        new_state.has_fp16_weights = False
+        new_state.use_pool = self.state.use_pool
+        self.state = new_state
+
+
+def _convert_linear_layers_to_llm_8bit(
+    module: torch.nn.Module, ignore_modules: set[str], outlier_threshold: float, prefix: str = ""
+) -> None:
+    """Convert all linear layers in the module to bnb.nn.Linear8bitLt layers."""
+    for name, child in module.named_children():
+        fullname = f"{prefix}.{name}" if prefix else name
+        if isinstance(child, torch.nn.Linear) and not any(fullname.startswith(s) for s in ignore_modules):
+            has_bias = child.bias is not None
+            replacement = InvokeLinear8bitLt(
+                child.in_features,
+                child.out_features,
+                bias=has_bias,
+                has_fp16_weights=False,
+                threshold=outlier_threshold,
+            )
+            replacement.weight.data = child.weight.data
+            if has_bias:
+                replacement.bias.data = child.bias.data
+            replacement.requires_grad_(False)
+            module.__setattr__(name, replacement)
+        else:
+            _convert_linear_layers_to_llm_8bit(
+                child, ignore_modules, outlier_threshold=outlier_threshold, prefix=fullname
+            )
+
+
+def quantize_model_llm_int8(model: torch.nn.Module, modules_to_not_convert: set[str], outlier_threshold: float = 6.0):
+    """Apply bitsandbytes LLM.8bit() quantization to the model."""
+    _convert_linear_layers_to_llm_8bit(
+        module=model, ignore_modules=modules_to_not_convert, outlier_threshold=outlier_threshold
+    )
+
+    return model

invokeai/backend/quantization/bnb_nf4.py

@@ -0,0 +1,156 @@
+import bitsandbytes as bnb
+import torch
+
+# This file contains utils for working with models that use bitsandbytes NF4 quantization.
+# The utils in this file are partially inspired by:
+# https://github.com/Lightning-AI/pytorch-lightning/blob/1551a16b94f5234a4a78801098f64d0732ef5cb5/src/lightning/fabric/plugins/precision/bitsandbytes.py
+
+# NOTE(ryand): All of the custom state_dict manipulation logic in this file is pretty hacky. This could be made much
+# cleaner by re-implementing bnb.nn.LinearNF4 with proper use of buffers and less magic. But, for now, we try to stick
+# close to the bitsandbytes classes to make interoperability easier with other models that might use bitsandbytes.
+
+
+class InvokeLinearNF4(bnb.nn.LinearNF4):
+    """A class that extends `bnb.nn.LinearNF4` to add the following functionality:
+    - Ability to load Linear NF4 layers from a pre-quantized state_dict.
+    - Ability to load Linear NF4 layers from a state_dict when the model is on the "meta" device.
+    """
+
+    def _load_from_state_dict(
+        self,
+        state_dict: dict[str, torch.Tensor],
+        prefix: str,
+        local_metadata,
+        strict,
+        missing_keys,
+        unexpected_keys,
+        error_msgs,
+    ):
+        """This method is based on the logic in the bitsandbytes serialization unit tests for `Linear4bit`:
+        https://github.com/bitsandbytes-foundation/bitsandbytes/blob/6d714a5cce3db5bd7f577bc447becc7a92d5ccc7/tests/test_linear4bit.py#L52-L71
+        """
+        weight = state_dict.pop(prefix + "weight")
+        bias = state_dict.pop(prefix + "bias", None)
+        # We expect the remaining keys to be quant_state keys.
+        quant_state_sd = state_dict
+
+        # During serialization, the quant_state is stored as subkeys of "weight." (See
+        # `bnb.nn.LinearNF4._save_to_state_dict()`). We validate that they at least have the correct prefix.
+        # TODO(ryand): Technically, we should be using `strict`, `missing_keys`, `unexpected_keys`, and `error_msgs`
+        # rather than raising an exception to correctly implement this API.
+        assert all(k.startswith(prefix + "weight.") for k in quant_state_sd.keys())
+
+        if len(quant_state_sd) > 0:
+            # We are loading a pre-quantized state dict.
+            self.weight = bnb.nn.Params4bit.from_prequantized(
+                data=weight, quantized_stats=quant_state_sd, device=weight.device
+            )
+            self.bias = bias if bias is None else torch.nn.Parameter(bias, requires_grad=False)
+        else:
+            # We are loading a non-quantized state dict.
+
+            # We could simply call the `super()._load_from_state_dict()` method here, but then we wouldn't be able to
+            # load from a state_dict into a model on the "meta" device. Attempting to load into a model on the "meta"
+            # device requires setting `assign=True`, doing this with the default `super()._load_from_state_dict()`
+            # implementation causes `Params4Bit` to be replaced by a `torch.nn.Parameter`. By initializing a new
+            # `Params4bit` object, we work around this issue. It's a bit hacky, but it gets the job done.
+            self.weight = bnb.nn.Params4bit(
+                data=weight,
+                requires_grad=self.weight.requires_grad,
+                compress_statistics=self.weight.compress_statistics,
+                quant_type=self.weight.quant_type,
+                quant_storage=self.weight.quant_storage,
+                module=self,
+            )
+            self.bias = bias if bias is None else torch.nn.Parameter(bias)
+
+
+def _replace_param(
+    param: torch.nn.Parameter | bnb.nn.Params4bit,
+    data: torch.Tensor,
+) -> torch.nn.Parameter:
+    """A helper function to replace the data of a model parameter with new data in a way that allows replacing params on
+    the "meta" device.
+
+    Supports both `torch.nn.Parameter` and `bnb.nn.Params4bit` parameters.
+    """
+    if param.device.type == "meta":
+        # Doing `param.data = data` raises a RuntimeError if param.data was on the "meta" device, so we need to
+        # re-create the param instead of overwriting the data.
+        if isinstance(param, bnb.nn.Params4bit):
+            return bnb.nn.Params4bit(
+                data,
+                requires_grad=data.requires_grad,
+                quant_state=param.quant_state,
+                compress_statistics=param.compress_statistics,
+                quant_type=param.quant_type,
+            )
+        return torch.nn.Parameter(data, requires_grad=data.requires_grad)
+
+    param.data = data
+    return param
+
+
+def _convert_linear_layers_to_nf4(
+    module: torch.nn.Module,
+    ignore_modules: set[str],
+    compute_dtype: torch.dtype,
+    compress_statistics: bool = False,
+    prefix: str = "",
+) -> None:
+    """Convert all linear layers in the model to NF4 quantized linear layers.
+
+    Args:
+        module: All linear layers in this module will be converted.
+        ignore_modules: A set of module prefixes to ignore when converting linear layers.
+        compute_dtype: The dtype to use for computation in the quantized linear layers.
+        compress_statistics: Whether to enable nested quantization (aka double quantization) where the quantization
+           constants from the first quantization are quantized again.
+        prefix: The prefix of the current module in the model. Used to call this function recursively.
+    """
+    for name, child in module.named_children():
+        fullname = f"{prefix}.{name}" if prefix else name
+        if isinstance(child, torch.nn.Linear) and not any(fullname.startswith(s) for s in ignore_modules):
+            has_bias = child.bias is not None
+            replacement = InvokeLinearNF4(
+                child.in_features,
+                child.out_features,
+                bias=has_bias,
+                compute_dtype=compute_dtype,
+                compress_statistics=compress_statistics,
+            )
+            if has_bias:
+                replacement.bias = _replace_param(replacement.bias, child.bias.data)
+            replacement.weight = _replace_param(replacement.weight, child.weight.data)
+            replacement.requires_grad_(False)
+            module.__setattr__(name, replacement)
+        else:
+            _convert_linear_layers_to_nf4(child, ignore_modules, compute_dtype=compute_dtype, prefix=fullname)
+
+
+def quantize_model_nf4(model: torch.nn.Module, modules_to_not_convert: set[str], compute_dtype: torch.dtype):
+    """Apply bitsandbytes nf4 quantization to the model.
+
+    You likely want to call this function inside a `accelerate.init_empty_weights()` context.
+
+    Example usage:
+    ```
+    # Initialize the model from a config on the meta device.
+    with accelerate.init_empty_weights():
+        model = ModelClass.from_config(...)
+
+    # Add NF4 quantization linear layers to the model - still on the meta device.
+    with accelerate.init_empty_weights():
+        model = quantize_model_nf4(model, modules_to_not_convert=set(), compute_dtype=torch.float16)
+
+    # Load a state_dict into the model. (Could be either a prequantized or non-quantized state_dict.)
+    model.load_state_dict(state_dict, strict=True, assign=True)
+
+    # Move the model to the "cuda" device. If the model was non-quantized, this is where the weight quantization takes
+    # place.
+    model.to("cuda")
+    ```
+    """
+    _convert_linear_layers_to_nf4(module=model, ignore_modules=modules_to_not_convert, compute_dtype=compute_dtype)
+
+    return model

invokeai/backend/quantization/scripts/load_flux_model_bnb_llm_int8.py

@@ -0,0 +1,79 @@
+from pathlib import Path
+
+import accelerate
+from safetensors.torch import load_file, save_file
+
+from invokeai.backend.flux.model import Flux
+from invokeai.backend.flux.util import params
+from invokeai.backend.quantization.bnb_llm_int8 import quantize_model_llm_int8
+from invokeai.backend.quantization.scripts.load_flux_model_bnb_nf4 import log_time
+
+
+def main():
+    """A script for quantizing a FLUX transformer model using the bitsandbytes LLM.int8() quantization method.
+
+    This script is primarily intended for reference. The script params (e.g. the model_path, modules_to_not_convert,
+    etc.) are hardcoded and would need to be modified for other use cases.
+    """
+    # Load the FLUX transformer model onto the meta device.
+    model_path = Path(
+        "/data/invokeai/models/.download_cache/https__huggingface.co_black-forest-labs_flux.1-schnell_resolve_main_flux1-schnell.safetensors/flux1-schnell.safetensors"
+    )
+
+    with log_time("Intialize FLUX transformer on meta device"):
+        # TODO(ryand): Determine if this is a schnell model or a dev model and load the appropriate config.
+        p = params["flux-schnell"]
+
+        # Initialize the model on the "meta" device.
+        with accelerate.init_empty_weights():
+            model = Flux(p)
+
+    # TODO(ryand): We may want to add some modules to not quantize here (e.g. the proj_out layer). See the accelerate
+    # `get_keys_to_not_convert(...)` function for a heuristic to determine which modules to not quantize.
+    modules_to_not_convert: set[str] = set()
+
+    model_int8_path = model_path.parent / "bnb_llm_int8.safetensors"
+    if model_int8_path.exists():
+        # The quantized model already exists, load it and return it.
+        print(f"A pre-quantized model already exists at '{model_int8_path}'. Attempting to load it...")
+
+        # Replace the linear layers with LLM.int8() quantized linear layers (still on the meta device).
+        with log_time("Replace linear layers with LLM.int8() layers"), accelerate.init_empty_weights():
+            model = quantize_model_llm_int8(model, modules_to_not_convert=modules_to_not_convert)
+
+        with log_time("Load state dict into model"):
+            sd = load_file(model_int8_path)
+            model.load_state_dict(sd, strict=True, assign=True)
+
+        with log_time("Move model to cuda"):
+            model = model.to("cuda")
+
+        print(f"Successfully loaded pre-quantized model from '{model_int8_path}'.")
+
+    else:
+        # The quantized model does not exist, quantize the model and save it.
+        print(f"No pre-quantized model found at '{model_int8_path}'. Quantizing the model...")
+
+        with log_time("Replace linear layers with LLM.int8() layers"), accelerate.init_empty_weights():
+            model = quantize_model_llm_int8(model, modules_to_not_convert=modules_to_not_convert)
+
+        with log_time("Load state dict into model"):
+            state_dict = load_file(model_path)
+            # TODO(ryand): Cast the state_dict to the appropriate dtype?
+            model.load_state_dict(state_dict, strict=True, assign=True)
+
+        with log_time("Move model to cuda and quantize"):
+            model = model.to("cuda")
+
+        with log_time("Save quantized model"):
+            model_int8_path.parent.mkdir(parents=True, exist_ok=True)
+            save_file(model.state_dict(), model_int8_path)
+
+        print(f"Successfully quantized and saved model to '{model_int8_path}'.")
+
+    assert isinstance(model, Flux)
+    return model
+
+
+if __name__ == "__main__":
+    main()

invokeai/backend/quantization/scripts/load_flux_model_bnb_nf4.py

@@ -0,0 +1,96 @@
+import time
+from contextlib import contextmanager
+from pathlib import Path
+
+import accelerate
+import torch
+from safetensors.torch import load_file, save_file
+
+from invokeai.backend.flux.model import Flux
+from invokeai.backend.flux.util import params
+from invokeai.backend.quantization.bnb_nf4 import quantize_model_nf4
+
+
+@contextmanager
+def log_time(name: str):
+    """Helper context manager to log the time taken by a block of code."""
+    start = time.time()
+    try:
+        yield None
+    finally:
+        end = time.time()
+        print(f"'{name}' took {end - start:.4f} secs")
+
+
+def main():
+    """A script for quantizing a FLUX transformer model using the bitsandbytes NF4 quantization method.
+
+    This script is primarily intended for reference. The script params (e.g. the model_path, modules_to_not_convert,
+    etc.) are hardcoded and would need to be modified for other use cases.
+    """
+    model_path = Path(
+        "/data/invokeai/models/.download_cache/https__huggingface.co_black-forest-labs_flux.1-schnell_resolve_main_flux1-schnell.safetensors/flux1-schnell.safetensors"
+    )
+
+    # inference_dtype = torch.bfloat16
+    with log_time("Intialize FLUX transformer on meta device"):
+        # TODO(ryand): Determine if this is a schnell model or a dev model and load the appropriate config.
+        p = params["flux-schnell"]
+
+        # Initialize the model on the "meta" device.
+        with accelerate.init_empty_weights():
+            model = Flux(p)
+
+    # TODO(ryand): We may want to add some modules to not quantize here (e.g. the proj_out layer). See the accelerate
+    # `get_keys_to_not_convert(...)` function for a heuristic to determine which modules to not quantize.
+    modules_to_not_convert: set[str] = set()
+
+    model_nf4_path = model_path.parent / "bnb_nf4.safetensors"
+    if model_nf4_path.exists():
+        # The quantized model already exists, load it and return it.
+        print(f"A pre-quantized model already exists at '{model_nf4_path}'. Attempting to load it...")
+
+        # Replace the linear layers with NF4 quantized linear layers (still on the meta device).
+        with log_time("Replace linear layers with NF4 layers"), accelerate.init_empty_weights():
+            model = quantize_model_nf4(
+                model, modules_to_not_convert=modules_to_not_convert, compute_dtype=torch.bfloat16
+            )
+
+        with log_time("Load state dict into model"):
+            state_dict = load_file(model_nf4_path)
+            model.load_state_dict(state_dict, strict=True, assign=True)
+
+        with log_time("Move model to cuda"):
+            model = model.to("cuda")
+
+        print(f"Successfully loaded pre-quantized model from '{model_nf4_path}'.")
+
+    else:
+        # The quantized model does not exist, quantize the model and save it.
+        print(f"No pre-quantized model found at '{model_nf4_path}'. Quantizing the model...")
+
+        with log_time("Replace linear layers with NF4 layers"), accelerate.init_empty_weights():
+            model = quantize_model_nf4(
+                model, modules_to_not_convert=modules_to_not_convert, compute_dtype=torch.bfloat16
+            )
+
+        with log_time("Load state dict into model"):
+            state_dict = load_file(model_path)
+            # TODO(ryand): Cast the state_dict to the appropriate dtype?
+            model.load_state_dict(state_dict, strict=True, assign=True)
+
+        with log_time("Move model to cuda and quantize"):
+            model = model.to("cuda")
+
+        with log_time("Save quantized model"):
+            model_nf4_path.parent.mkdir(parents=True, exist_ok=True)
+            save_file(model.state_dict(), model_nf4_path)
+
+        print(f"Successfully quantized and saved model to '{model_nf4_path}'.")
+
+    assert isinstance(model, Flux)
+    return model
+
+
+if __name__ == "__main__":
+    main()

invokeai/backend/quantization/scripts/quantize_t5_xxl_bnb_llm_int8.py

@@ -0,0 +1,92 @@
+from pathlib import Path
+
+import accelerate
+from safetensors.torch import load_file, save_file
+from transformers import AutoConfig, AutoModelForTextEncoding, T5EncoderModel
+
+from invokeai.backend.quantization.bnb_llm_int8 import quantize_model_llm_int8
+from invokeai.backend.quantization.scripts.load_flux_model_bnb_nf4 import log_time
+
+
+def load_state_dict_into_t5(model: T5EncoderModel, state_dict: dict):
+    # There is a shared reference to a single weight tensor in the model.
+    # Both "encoder.embed_tokens.weight" and "shared.weight" refer to the same tensor, so only the latter should
+    # be present in the state_dict.
+    missing_keys, unexpected_keys = model.load_state_dict(state_dict, strict=False, assign=True)
+    assert len(unexpected_keys) == 0
+    assert set(missing_keys) == {"encoder.embed_tokens.weight"}
+    # Assert that the layers we expect to be shared are actually shared.
+    assert model.encoder.embed_tokens.weight is model.shared.weight
+
+
+def main():
+    """A script for quantizing a T5 text encoder model using the bitsandbytes LLM.int8() quantization method.
+
+    This script is primarily intended for reference. The script params (e.g. the model_path, modules_to_not_convert,
+    etc.) are hardcoded and would need to be modified for other use cases.
+    """
+    model_path = Path("/data/misc/text_encoder_2")
+
+    with log_time("Intialize T5 on meta device"):
+        model_config = AutoConfig.from_pretrained(model_path)
+        with accelerate.init_empty_weights():
+            model = AutoModelForTextEncoding.from_config(model_config)
+
+    # TODO(ryand): We may want to add some modules to not quantize here (e.g. the proj_out layer). See the accelerate
+    # `get_keys_to_not_convert(...)` function for a heuristic to determine which modules to not quantize.
+    modules_to_not_convert: set[str] = set()
+
+    model_int8_path = model_path / "bnb_llm_int8.safetensors"
+    if model_int8_path.exists():
+        # The quantized model already exists, load it and return it.
+        print(f"A pre-quantized model already exists at '{model_int8_path}'. Attempting to load it...")
+
+        # Replace the linear layers with LLM.int8() quantized linear layers (still on the meta device).
+        with log_time("Replace linear layers with LLM.int8() layers"), accelerate.init_empty_weights():
+            model = quantize_model_llm_int8(model, modules_to_not_convert=modules_to_not_convert)
+
+        with log_time("Load state dict into model"):
+            sd = load_file(model_int8_path)
+            load_state_dict_into_t5(model, sd)
+
+        with log_time("Move model to cuda"):
+            model = model.to("cuda")
+
+        print(f"Successfully loaded pre-quantized model from '{model_int8_path}'.")
+
+    else:
+        # The quantized model does not exist, quantize the model and save it.
+        print(f"No pre-quantized model found at '{model_int8_path}'. Quantizing the model...")
+
+        with log_time("Replace linear layers with LLM.int8() layers"), accelerate.init_empty_weights():
+            model = quantize_model_llm_int8(model, modules_to_not_convert=modules_to_not_convert)
+
+        with log_time("Load state dict into model"):
+            # Load sharded state dict.
+            files = list(model_path.glob("*.safetensors"))
+            state_dict = {}
+            for file in files:
+                sd = load_file(file)
+                state_dict.update(sd)
+            load_state_dict_into_t5(model, state_dict)
+
+        with log_time("Move model to cuda and quantize"):
+            model = model.to("cuda")
+
+        with log_time("Save quantized model"):
+            model_int8_path.parent.mkdir(parents=True, exist_ok=True)
+            state_dict = model.state_dict()
+            state_dict.pop("encoder.embed_tokens.weight")
+            save_file(state_dict, model_int8_path)
+            # This handling of shared weights could also be achieved with save_model(...), but then we'd lose control
+            # over which keys are kept. And, the corresponding load_model(...) function does not support assign=True.
+            # save_model(model, model_int8_path)
+
+        print(f"Successfully quantized and saved model to '{model_int8_path}'.")
+
+    assert isinstance(model, T5EncoderModel)
+    return model
+
+
+if __name__ == "__main__":
+    main()

invokeai/backend/stable_diffusion/diffusion/conditioning_data.py

@@ -25,11 +25,6 @@ class BasicConditioningInfo:
         return self
 
 
-@dataclass
-class ConditioningFieldData:
-    conditionings: List[BasicConditioningInfo]
-
-
 @dataclass
 class SDXLConditioningInfo(BasicConditioningInfo):
     """SDXL text conditioning information produced by Compel."""
@@ -43,6 +38,22 @@ class SDXLConditioningInfo(BasicConditioningInfo):
         return super().to(device=device, dtype=dtype)
 
 
+@dataclass
+class FLUXConditioningInfo:
+    clip_embeds: torch.Tensor
+    t5_embeds: torch.Tensor
+
+    def to(self, device: torch.device | None = None, dtype: torch.dtype | None = None):
+        self.clip_embeds = self.clip_embeds.to(device=device, dtype=dtype)
+        self.t5_embeds = self.t5_embeds.to(device=device, dtype=dtype)
+        return self
+
+
+@dataclass
+class ConditioningFieldData:
+    conditionings: List[BasicConditioningInfo] | List[SDXLConditioningInfo] | List[FLUXConditioningInfo]
+
+
 @dataclass
 class IPAdapterConditioningInfo:
     cond_image_prompt_embeds: torch.Tensor

invokeai/backend/stable_diffusion/extensions/lora.py

@@ -12,7 +12,7 @@ from invokeai.backend.util.devices import TorchDevice
 if TYPE_CHECKING:
     from invokeai.app.invocations.model import ModelIdentifierField
     from invokeai.app.services.shared.invocation_context import InvocationContext
-    from invokeai.backend.lora import LoRAModelRaw
+    from invokeai.backend.peft.lora import LoRAModelRaw
     from invokeai.backend.util.original_weights_storage import OriginalWeightsStorage
 
 

invokeai/backend/util/__init__.py

@@ -3,10 +3,9 @@ Initialization file for invokeai.backend.util
 """
 
 from invokeai.backend.util.logging import InvokeAILogger
-from invokeai.backend.util.util import GIG, Chdir, directory_size
+from invokeai.backend.util.util import Chdir, directory_size
 
 __all__ = [
-    "GIG",
     "directory_size",
     "Chdir",
     "InvokeAILogger",

invokeai/backend/util/util.py

@@ -7,9 +7,6 @@ from pathlib import Path
 
 from PIL import Image
 
-# actual size of a gig
-GIG = 1073741824
-
 
 def slugify(value: str, allow_unicode: bool = False) -> str:
     """

invokeai/frontend/web/.eslintrc.js

@@ -12,10 +12,6 @@ module.exports = {
     'i18next/no-literal-string': 'error',
     // https://eslint.org/docs/latest/rules/no-console
     'no-console': 'error',
-    // https://eslint.org/docs/latest/rules/no-promise-executor-return
-    'no-promise-executor-return': 'error',
-    // https://eslint.org/docs/latest/rules/require-await
-    'require-await': 'error',
   },
   overrides: [
     /**

invokeai/frontend/web/.storybook/ReduxInit.tsx

@@ -1,5 +1,5 @@
 import { PropsWithChildren, memo, useEffect } from 'react';
-import { modelChanged } from '../src/features/controlLayers/store/canvasV2Slice';
+import { modelChanged } from '../src/features/parameters/store/generationSlice';
 import { useAppDispatch } from '../src/app/store/storeHooks';
 import { useGlobalModifiersInit } from '@invoke-ai/ui-library';
 /**
@@ -10,9 +10,7 @@ export const ReduxInit = memo((props: PropsWithChildren) => {
   const dispatch = useAppDispatch();
   useGlobalModifiersInit();
   useEffect(() => {
-    dispatch(
-      modelChanged({ model: { key: 'test_model', hash: 'some_hash', name: 'some name', base: 'sd-1', type: 'main' } })
-    );
+    dispatch(modelChanged({ key: 'test_model', hash: 'some_hash', name: 'some name', base: 'sd-1', type: 'main' }));
   }, []);
 
   return props.children;

invokeai/frontend/web/knip.ts

@@ -9,8 +9,6 @@ const config: KnipConfig = {
     'src/services/api/schema.ts',
     'src/features/nodes/types/v1/**',
     'src/features/nodes/types/v2/**',
-    // TODO(psyche): maybe we can clean up these utils after canvas v2 release
-    'src/features/controlLayers/konva/util.ts',
   ],
   ignoreBinaries: ['only-allow'],
   paths: {

invokeai/frontend/web/package.json

@@ -24,7 +24,7 @@
     "build": "pnpm run lint && vite build",
     "typegen": "node scripts/typegen.js",
     "preview": "vite preview",
-    "lint:knip": "knip --tags=-knipignore",
+    "lint:knip": "knip",
     "lint:dpdm": "dpdm --no-warning --no-tree --transform --exit-code circular:1 src/main.tsx",
     "lint:eslint": "eslint --max-warnings=0 .",
     "lint:prettier": "prettier --check .",
@@ -52,17 +52,17 @@
     }
   },
   "dependencies": {
+    "@chakra-ui/react-use-size": "^2.1.0",
     "@dagrejs/dagre": "^1.1.3",
     "@dagrejs/graphlib": "^2.2.3",
     "@dnd-kit/core": "^6.1.0",
     "@dnd-kit/sortable": "^8.0.0",
     "@dnd-kit/utilities": "^3.2.2",
     "@fontsource-variable/inter": "^5.0.20",
-    "@invoke-ai/ui-library": "^0.0.31",
+    "@invoke-ai/ui-library": "^0.0.29",
     "@nanostores/react": "^0.7.3",
     "@reduxjs/toolkit": "2.2.3",
     "@roarr/browser-log-writer": "^1.3.0",
-    "async-mutex": "^0.5.0",
     "chakra-react-select": "^4.9.1",
     "compare-versions": "^6.1.1",
     "dateformat": "^5.0.3",
@@ -74,8 +74,6 @@
     "jsondiffpatch": "^0.6.0",
     "konva": "^9.3.14",
     "lodash-es": "^4.17.21",
-    "lru-cache": "^11.0.0",
-    "nanoid": "^5.0.7",
     "nanostores": "^0.11.2",
     "new-github-issue-url": "^1.0.0",
     "overlayscrollbars": "^2.10.0",
@@ -90,6 +88,7 @@
     "react-hotkeys-hook": "4.5.0",
     "react-i18next": "^14.1.3",
     "react-icons": "^5.2.1",
+    "react-konva": "^18.2.10",
     "react-redux": "9.1.2",
     "react-resizable-panels": "^2.0.23",
     "react-select": "5.8.0",
@@ -103,9 +102,9 @@
     "roarr": "^7.21.1",
     "serialize-error": "^11.0.3",
     "socket.io-client": "^4.7.5",
-    "stable-hash": "^0.0.4",
     "use-debounce": "^10.0.2",
     "use-device-pixel-ratio": "^1.1.2",
+    "use-image": "^1.1.1",
     "uuid": "^10.0.0",
     "zod": "^3.23.8",
     "zod-validation-error": "^3.3.1"

invokeai/frontend/web/pnpm-lock.yaml

@@ -5,6 +5,9 @@ settings:
   excludeLinksFromLockfile: false
 
 dependencies:
+  '@chakra-ui/react-use-size':
+    specifier: ^2.1.0
+    version: 2.1.0(react@18.3.1)
   '@dagrejs/dagre':
     specifier: ^1.1.3
     version: 1.1.3
@@ -24,8 +27,8 @@ dependencies:
     specifier: ^5.0.20
     version: 5.0.20
   '@invoke-ai/ui-library':
-    specifier: ^0.0.31
-    version: 0.0.31(@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@5.0.20)(@types/react@18.3.3)(i18next@23.12.2)(react-dom@18.3.1)(react@18.3.1)
+    specifier: ^0.0.29
+    version: 0.0.29(@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@5.0.20)(@types/react@18.3.3)(i18next@23.12.2)(react-dom@18.3.1)(react@18.3.1)
   '@nanostores/react':
     specifier: ^0.7.3
     version: 0.7.3(nanostores@0.11.2)(react@18.3.1)
@@ -35,9 +38,6 @@ dependencies:
   '@roarr/browser-log-writer':
     specifier: ^1.3.0
     version: 1.3.0
-  async-mutex:
-    specifier: ^0.5.0
-    version: 0.5.0
   chakra-react-select:
     specifier: ^4.9.1
     version: 4.9.1(@chakra-ui/form-control@2.2.0)(@chakra-ui/icon@3.2.0)(@chakra-ui/layout@2.3.1)(@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)(@emotion/react@11.13.0)(@types/react@18.3.3)(react-dom@18.3.1)(react@18.3.1)
@@ -71,12 +71,6 @@ dependencies:
   lodash-es:
     specifier: ^4.17.21
     version: 4.17.21
-  lru-cache:
-    specifier: ^11.0.0
-    version: 11.0.0
-  nanoid:
-    specifier: ^5.0.7
-    version: 5.0.7
   nanostores:
     specifier: ^0.11.2
     version: 0.11.2
@@ -119,6 +113,9 @@ dependencies:
   react-icons:
     specifier: ^5.2.1
     version: 5.2.1(react@18.3.1)
+  react-konva:
+    specifier: ^18.2.10
+    version: 18.2.10(konva@9.3.14)(react-dom@18.3.1)(react@18.3.1)
   react-redux:
     specifier: 9.1.2
     version: 9.1.2(@types/react@18.3.3)(react@18.3.1)(redux@5.0.1)
@@ -158,15 +155,15 @@ dependencies:
   socket.io-client:
     specifier: ^4.7.5
     version: 4.7.5
-  stable-hash:
-    specifier: ^0.0.4
-    version: 0.0.4
   use-debounce:
     specifier: ^10.0.2
     version: 10.0.2(react@18.3.1)
   use-device-pixel-ratio:
     specifier: ^1.1.2
     version: 1.1.2(react@18.3.1)
+  use-image:
+    specifier: ^1.1.1
+    version: 1.1.1(react-dom@18.3.1)(react@18.3.1)
   uuid:
     specifier: ^10.0.0
     version: 10.0.0
@@ -3564,8 +3561,8 @@ packages:
       prettier: 3.3.3
     dev: true
 
-  /@invoke-ai/ui-library@0.0.31(@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@5.0.20)(@types/react@18.3.3)(i18next@23.12.2)(react-dom@18.3.1)(react@18.3.1):
-    resolution: {integrity: sha512-7LtOUN/bcGHc8jCRd2m22DvP2eeogqwM/shdXQpLH5RY2FzWJNXlWdVT4hIPGDu7znnk3xvXlZvo6tiGSjbnCQ==}
+  /@invoke-ai/ui-library@0.0.29(@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@5.0.20)(@types/react@18.3.3)(i18next@23.12.2)(react-dom@18.3.1)(react@18.3.1):
+    resolution: {integrity: sha512-7SYOaiEEKk9iHk0hg2R2yVxiuV3I1x6bDEv0R3Y2tCH/Aq5XDG2tR+d7SQAPqf5+za3S+qfNFjbjl7GvEMwqmA==}
     peerDependencies:
       '@fontsource-variable/inter': ^5.0.16
       react: ^18.2.0
@@ -5205,6 +5202,12 @@ packages:
       '@types/react': 18.3.3
     dev: true
 
+  /@types/react-reconciler@0.28.8:
+    resolution: {integrity: sha512-SN9c4kxXZonFhbX4hJrZy37yw9e7EIxcpHCxQv5JUS18wDE5ovkQKlqQEkufdJCCMfuI9BnjUJvhYeJ9x5Ra7g==}
+    dependencies:
+      '@types/react': 18.3.3
+    dev: false
+
   /@types/react-transition-group@4.4.10:
     resolution: {integrity: sha512-hT/+s0VQs2ojCX823m60m5f0sL5idt9SO6Tj6Dg+rdphGPIeJbJ6CxvBYkgkGKrYeDjvIpKTR38UzmtHJOGW3Q==}
     dependencies:
@@ -5900,12 +5903,6 @@ packages:
       tslib: 2.6.3
     dev: true
 
-  /async-mutex@0.5.0:
-    resolution: {integrity: sha512-1A94B18jkJ3DYq284ohPxoXbfTA5HsQ7/Mf4DEhcyLx3Bz27Rh59iScbB6EPiP+B+joue6YCxcMXSbFC1tZKwA==}
-    dependencies:
-      tslib: 2.6.3
-    dev: false
-
   /attr-accept@2.2.2:
     resolution: {integrity: sha512-7prDjvt9HmqiZ0cl5CRjtS84sEyhsHP2coDkaZKRKVfCDo9s7iw7ChVmar78Gu9pC4SoR/28wFu/G5JJhTnqEg==}
     engines: {node: '>=4'}
@@ -8387,6 +8384,15 @@ packages:
       set-function-name: 2.0.2
     dev: true
 
+  /its-fine@1.2.5(react@18.3.1):
+    resolution: {integrity: sha512-fXtDA0X0t0eBYAGLVM5YsgJGsJ5jEmqZEPrGbzdf5awjv0xE7nqv3TVnvtUF060Tkes15DbDAKW/I48vsb6SyA==}
+    peerDependencies:
+      react: '>=18.0'
+    dependencies:
+      '@types/react-reconciler': 0.28.8
+      react: 18.3.1
+    dev: false
+
   /jackspeak@2.3.6:
     resolution: {integrity: sha512-N3yCS/NegsOBokc8GAdM8UcmfsKiSS8cipheD/nivzr700H+nsMOxJjQnvwOcRYVuFkdH0wGUvW2WbXGmrZGbQ==}
     engines: {node: '>=14'}
@@ -8704,11 +8710,6 @@ packages:
     engines: {node: 14 || >=16.14}
     dev: true
 
-  /lru-cache@11.0.0:
-    resolution: {integrity: sha512-Qv32eSV1RSCfhY3fpPE2GNZ8jgM9X7rdAfemLWqTUxwiyIC4jJ6Sy0fZ8H+oLWevO6i4/bizg7c8d8i6bxrzbA==}
-    engines: {node: 20 || >=22}
-    dev: false
-
   /lru-cache@5.1.1:
     resolution: {integrity: sha512-KpNARQA3Iwv+jTA0utUVVbrh+Jlrr1Fv0e56GGzAFOXN7dk/FviaDW8LHmK52DlcH4WP2n6gI8vN1aesBFgo9w==}
     dependencies:
@@ -8996,12 +8997,6 @@ packages:
     hasBin: true
     dev: true
 
-  /nanoid@5.0.7:
-    resolution: {integrity: sha512-oLxFY2gd2IqnjcYyOXD8XGCftpGtZP2AbHbOkthDkvRywH5ayNtPVy9YlOPcHckXzbLTCHpkb7FB+yuxKV13pQ==}
-    engines: {node: ^18 || >=20}
-    hasBin: true
-    dev: false
-
   /nanostores@0.11.2:
     resolution: {integrity: sha512-6bucNxMJA5rNV554WQl+MWGng0QVMzlRgpKTHHfIbVLrhQ+yRXBychV9ECGVuuUfCMQPjfIG9bj8oJFZ9hYP/Q==}
     engines: {node: ^18.0.0 || >=20.0.0}
@@ -9806,6 +9801,33 @@ packages:
     resolution: {integrity: sha512-/LLMVyas0ljjAtoYiPqYiL8VWXzUUdThrmU5+n20DZv+a+ClRoevUzw5JxU+Ieh5/c87ytoTBV9G1FiKfNJdmg==}
     dev: true
 
+  /react-konva@18.2.10(konva@9.3.14)(react-dom@18.3.1)(react@18.3.1):
+    resolution: {integrity: sha512-ohcX1BJINL43m4ynjZ24MxFI1syjBdrXhqVxYVDw2rKgr3yuS0x/6m1Y2Z4sl4T/gKhfreBx8KHisd0XC6OT1g==}
+    peerDependencies:
+      konva: ^8.0.1 || ^7.2.5 || ^9.0.0
+      react: '>=18.0.0'
+      react-dom: '>=18.0.0'
+    dependencies:
+      '@types/react-reconciler': 0.28.8
+      its-fine: 1.2.5(react@18.3.1)
+      konva: 9.3.14
+      react: 18.3.1
+      react-dom: 18.3.1(react@18.3.1)
+      react-reconciler: 0.29.2(react@18.3.1)
+      scheduler: 0.23.2
+    dev: false
+
+  /react-reconciler@0.29.2(react@18.3.1):
+    resolution: {integrity: sha512-zZQqIiYgDCTP/f1N/mAR10nJGrPD2ZR+jDSEsKWJHYC7Cm2wodlwbR3upZRdC3cjIjSlTLNVyO7Iu0Yy7t2AYg==}
+    engines: {node: '>=0.10.0'}
+    peerDependencies:
+      react: ^18.3.1
+    dependencies:
+      loose-envify: 1.4.0
+      react: 18.3.1
+      scheduler: 0.23.2
+    dev: false
+
   /react-redux@9.1.2(@types/react@18.3.3)(react@18.3.1)(redux@5.0.1):
     resolution: {integrity: sha512-0OA4dhM1W48l3uzmv6B7TXPCGmokUU4p1M44DGN2/D9a1FjVPukVjER1PcPX97jIg6aUeLq1XJo1IpfbgULn0w==}
     peerDependencies:
@@ -10603,10 +10625,6 @@ packages:
     resolution: {integrity: sha512-D9cPgkvLlV3t3IzL0D0YLvGA9Ahk4PcvVwUbN0dSGr1aP0Nrt4AEnTUbuGvquEC0mA64Gqt1fzirlRs5ibXx8g==}
     dev: true
 
-  /stable-hash@0.0.4:
-    resolution: {integrity: sha512-LjdcbuBeLcdETCrPn9i8AYAZ1eCtu4ECAWtP7UleOiZ9LzVxRzzUZEoZ8zB24nhkQnDWyET0I+3sWokSDS3E7g==}
-    dev: false
-
   /stack-generator@2.0.10:
     resolution: {integrity: sha512-mwnua/hkqM6pF4k8SnmZ2zfETsRUpWXREfA/goT8SLCV4iOFa4bzOX2nDipWAZFPTjLvQB82f5yaodMVhK0yJQ==}
     dependencies:
@@ -11312,6 +11330,16 @@ packages:
       react: 18.3.1
     dev: false
 
+  /use-image@1.1.1(react-dom@18.3.1)(react@18.3.1):
+    resolution: {integrity: sha512-n4YO2k8AJG/BcDtxmBx8Aa+47kxY5m335dJiCQA5tTeVU4XdhrhqR6wT0WISRXwdMEOv5CSjqekDZkEMiiWaYQ==}
+    peerDependencies:
+      react: '>=16.8.0'
+      react-dom: '>=16.8.0'
+    dependencies:
+      react: 18.3.1
+      react-dom: 18.3.1(react@18.3.1)
+    dev: false
+
   /use-isomorphic-layout-effect@1.1.2(@types/react@18.3.3)(react@18.3.1):
     resolution: {integrity: sha512-49L8yCO3iGT/ZF9QttjwLF/ZD9Iwto5LnH5LmEdk/6cFmXddqi2ulF0edxTwjj+7mqvpVVGQWvbXZdn32wRSHA==}
     peerDependencies:

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

@@ -80,7 +80,6 @@
         "aboutDesc": "Using Invoke for work? Check out:",
         "aboutHeading": "Own Your Creative Power",
         "accept": "Accept",
-        "apply": "Apply",
         "add": "Add",
         "advanced": "Advanced",
         "ai": "ai",
@@ -116,7 +115,6 @@
         "githubLabel": "Github",
         "goTo": "Go to",
         "hotkeysLabel": "Hotkeys",
-        "loadingImage": "Loading Image",
         "imageFailedToLoad": "Unable to Load Image",
         "img2img": "Image To Image",
         "inpaint": "inpaint",
@@ -327,10 +325,6 @@
         "canceled": "Canceled",
         "completedIn": "Completed in",
         "batch": "Batch",
-        "origin": "Origin",
-        "originCanvas": "Canvas",
-        "originWorkflows": "Workflows",
-        "originOther": "Other",
         "batchFieldValues": "Batch Field Values",
         "item": "Item",
         "session": "Session",
@@ -702,6 +696,8 @@
         "availableModels": "Available Models",
         "baseModel": "Base Model",
         "cancel": "Cancel",
+        "clipEmbed": "CLIP Embed",
+        "clipVision": "CLIP Vision",
         "config": "Config",
         "convert": "Convert",
         "convertingModelBegin": "Converting Model. Please wait.",
@@ -789,13 +785,16 @@
         "settings": "Settings",
         "simpleModelPlaceholder": "URL or path to a local file or diffusers folder",
         "source": "Source",
+        "spandrelImageToImage": "Image to Image (Spandrel)",
         "starterModels": "Starter Models",
+        "starterModelsInModelManager": "Starter Models can be found in Model Manager",
         "syncModels": "Sync Models",
         "textualInversions": "Textual Inversions",
         "triggerPhrases": "Trigger Phrases",
         "loraTriggerPhrases": "LoRA Trigger Phrases",
         "mainModelTriggerPhrases": "Main Model Trigger Phrases",
         "typePhraseHere": "Type phrase here",
+        "t5Encoder": "T5 Encoder",
         "upcastAttention": "Upcast Attention",
         "uploadImage": "Upload Image",
         "urlOrLocalPath": "URL or Local Path",
@@ -1101,6 +1100,7 @@
         "confirmOnDelete": "Confirm On Delete",
         "developer": "Developer",
         "displayInProgress": "Display Progress Images",
+        "enableImageDebugging": "Enable Image Debugging",
         "enableInformationalPopovers": "Enable Informational Popovers",
         "informationalPopoversDisabled": "Informational Popovers Disabled",
         "informationalPopoversDisabledDesc": "Informational popovers have been disabled. Enable them in Settings.",
@@ -1567,7 +1567,7 @@
         "copyToClipboard": "Copy to Clipboard",
         "cursorPosition": "Cursor Position",
         "darkenOutsideSelection": "Darken Outside Selection",
-        "discardAll": "Discard All & Cancel Pending Generations",
+        "discardAll": "Discard All",
         "discardCurrent": "Discard Current",
         "downloadAsImage": "Download As Image",
         "enableMask": "Enable Mask",
@@ -1645,32 +1645,16 @@
         "storeNotInitialized": "Store is not initialized"
     },
     "controlLayers": {
-        "generateMode": "Generate",
-        "generateModeDesc": "Create individual images. Generated images are added directly to the gallery.",
-        "composeMode": "Compose",
-        "composeModeDesc": "Compose your work iterative. Generated images are added back to the canvas.",
-        "autoSave": "Auto-save to Gallery",
-        "resetCanvas": "Reset Canvas",
-        "resetAll": "Reset All",
         "deleteAll": "Delete All",
-        "clearCaches": "Clear Caches",
-        "recalculateRects": "Recalculate Rects",
-        "clipToBbox": "Clip Strokes to Bbox",
         "addLayer": "Add Layer",
-        "duplicate": "Duplicate",
         "moveToFront": "Move to Front",
         "moveToBack": "Move to Back",
         "moveForward": "Move Forward",
         "moveBackward": "Move Backward",
         "brushSize": "Brush Size",
-        "width": "Width",
-        "zoom": "Zoom",
-        "resetView": "Reset View",
         "controlLayers": "Control Layers",
         "globalMaskOpacity": "Global Mask Opacity",
         "autoNegative": "Auto Negative",
-        "enableAutoNegative": "Enable Auto Negative",
-        "disableAutoNegative": "Disable Auto Negative",
         "deletePrompt": "Delete Prompt",
         "resetRegion": "Reset Region",
         "debugLayers": "Debug Layers",
@@ -1679,83 +1663,21 @@
         "addPositivePrompt": "Add $t(common.positivePrompt)",
         "addNegativePrompt": "Add $t(common.negativePrompt)",
         "addIPAdapter": "Add $t(common.ipAdapter)",
+        "regionalGuidance": "Regional Guidance",
         "regionalGuidanceLayer": "$t(controlLayers.regionalGuidance) $t(unifiedCanvas.layer)",
-        "raster": "Raster",
-        "rasterLayer_one": "Raster Layer",
-        "controlLayer_one": "Control Layer",
-        "inpaintMask_one": "Inpaint Mask",
-        "regionalGuidance_one": "Regional Guidance",
-        "ipAdapter_one": "IP Adapter",
-        "rasterLayer_other": "Raster Layers",
-        "controlLayer_other": "Control Layers",
-        "inpaintMask_other": "Inpaint Masks",
-        "regionalGuidance_other": "Regional Guidance",
-        "ipAdapter_other": "IP Adapters",
         "opacity": "Opacity",
-        "regionalGuidance_withCount_hidden": "Regional Guidance ({{count}} hidden)",
-        "controlAdapters_withCount_hidden": "Control Adapters ({{count}} hidden)",
-        "controlLayers_withCount_hidden": "Control Layers ({{count}} hidden)",
-        "rasterLayers_withCount_hidden": "Raster Layers ({{count}} hidden)",
-        "ipAdapters_withCount_hidden": "IP Adapters ({{count}} hidden)",
-        "inpaintMasks_withCount_hidden": "Inpaint Masks ({{count}} hidden)",
-        "regionalGuidance_withCount_visible": "Regional Guidance ({{count}})",
-        "controlAdapters_withCount_visible": "Control Adapters ({{count}})",
-        "controlLayers_withCount_visible": "Control Layers ({{count}})",
-        "rasterLayers_withCount_visible": "Raster Layers ({{count}})",
-        "ipAdapters_withCount_visible": "IP Adapters ({{count}})",
-        "inpaintMasks_withCount_visible": "Inpaint Masks ({{count}})",
         "globalControlAdapter": "Global $t(controlnet.controlAdapter_one)",
         "globalControlAdapterLayer": "Global $t(controlnet.controlAdapter_one) $t(unifiedCanvas.layer)",
         "globalIPAdapter": "Global $t(common.ipAdapter)",
         "globalIPAdapterLayer": "Global $t(common.ipAdapter) $t(unifiedCanvas.layer)",
         "globalInitialImage": "Global Initial Image",
         "globalInitialImageLayer": "$t(controlLayers.globalInitialImage) $t(unifiedCanvas.layer)",
-        "layer": "Layer",
         "opacityFilter": "Opacity Filter",
         "clearProcessor": "Clear Processor",
         "resetProcessor": "Reset Processor to Defaults",
         "noLayersAdded": "No Layers Added",
         "layers_one": "Layer",
-        "layers_other": "Layers",
-        "objects_zero": "empty",
-        "objects_one": "{{count}} object",
-        "objects_other": "{{count}} objects",
-        "convertToControlLayer": "Convert to Control Layer",
-        "convertToRasterLayer": "Convert to Raster Layer",
-        "transparency": "Transparency",
-        "enableTransparencyEffect": "Enable Transparency Effect",
-        "disableTransparencyEffect": "Disable Transparency Effect",
-        "hidingType": "Hiding {{type}}",
-        "showingType": "Showing {{type}}",
-        "dynamicGrid": "Dynamic Grid",
-        "logDebugInfo": "Log Debug Info",
-        "fill": {
-            "fillStyle": "Fill Style",
-            "solid": "Solid",
-            "grid": "Grid",
-            "crosshatch": "Crosshatch",
-            "vertical": "Vertical",
-            "horizontal": "Horizontal",
-            "diagonal": "Diagonal"
-        },
-        "tool": {
-            "brush": "Brush",
-            "eraser": "Eraser",
-            "rectangle": "Rectangle",
-            "bbox": "Bbox",
-            "move": "Move",
-            "view": "View",
-            "transform": "Transform",
-            "colorPicker": "Color Picker"
-        },
-        "filter": {
-            "filter": "Filter",
-            "filters": "Filters",
-            "filterType": "Filter Type",
-            "preview": "Preview",
-            "apply": "Apply",
-            "cancel": "Cancel"
-        }
+        "layers_other": "Layers"
     },
     "upscaling": {
         "upscale": "Upscale",
@@ -1843,30 +1765,5 @@
             "upscaling": "Upscaling",
             "upscalingTab": "$t(ui.tabs.upscaling) $t(common.tab)"
         }
-    },
-    "system": {
-        "enableLogging": "Enable Logging",
-        "logLevel": {
-            "logLevel": "Log Level",
-            "trace": "Trace",
-            "debug": "Debug",
-            "info": "Info",
-            "warn": "Warn",
-            "error": "Error",
-            "fatal": "Fatal"
-        },
-        "logNamespaces": {
-            "logNamespaces": "Log Namespaces",
-            "gallery": "Gallery",
-            "models": "Models",
-            "config": "Config",
-            "canvas": "Canvas",
-            "generation": "Generation",
-            "workflows": "Workflows",
-            "system": "System",
-            "events": "Events",
-            "queue": "Queue",
-            "metadata": "Metadata"
-        }
     }
 }

invokeai/frontend/web/scripts/typegen.js

@@ -38,7 +38,7 @@ async function generateTypes(schema) {
   process.stdout.write(`\nOK!\r\n`);
 }
 
-function main() {
+async function main() {
   const encoding = 'utf-8';
 
   if (process.stdin.isTTY) {

invokeai/frontend/web/src/app/components/App.tsx

@@ -6,7 +6,6 @@ import { appStarted } from 'app/store/middleware/listenerMiddleware/listeners/ap
 import { useAppDispatch, useAppSelector } from 'app/store/storeHooks';
 import type { PartialAppConfig } from 'app/types/invokeai';
 import ImageUploadOverlay from 'common/components/ImageUploadOverlay';
-import { useScopeFocusWatcher } from 'common/hooks/interactionScopes';
 import { useClearStorage } from 'common/hooks/useClearStorage';
 import { useFullscreenDropzone } from 'common/hooks/useFullscreenDropzone';
 import { useGlobalHotkeys } from 'common/hooks/useGlobalHotkeys';
@@ -14,13 +13,13 @@ import ChangeBoardModal from 'features/changeBoardModal/components/ChangeBoardMo
 import DeleteImageModal from 'features/deleteImageModal/components/DeleteImageModal';
 import { DynamicPromptsModal } from 'features/dynamicPrompts/components/DynamicPromptsPreviewModal';
 import { useStarterModelsToast } from 'features/modelManagerV2/hooks/useStarterModelsToast';
-import { ClearQueueConfirmationsAlertDialog } from 'features/queue/components/ClearQueueConfirmationAlertDialog';
 import { StylePresetModal } from 'features/stylePresets/components/StylePresetForm/StylePresetModal';
+import { activeStylePresetIdChanged } from 'features/stylePresets/store/stylePresetSlice';
 import { configChanged } from 'features/system/store/configSlice';
 import { languageSelector } from 'features/system/store/systemSelectors';
-import { AppContent } from 'features/ui/components/AppContent';
+import InvokeTabs from 'features/ui/components/InvokeTabs';
+import type { InvokeTabName } from 'features/ui/store/tabMap';
 import { setActiveTab } from 'features/ui/store/uiSlice';
-import type { TabName } from 'features/ui/store/uiTypes';
 import { useGetAndLoadLibraryWorkflow } from 'features/workflowLibrary/hooks/useGetAndLoadLibraryWorkflow';
 import { AnimatePresence } from 'framer-motion';
 import i18n from 'i18n';
@@ -41,10 +40,17 @@ interface Props {
     action: 'sendToImg2Img' | 'sendToCanvas' | 'useAllParameters';
   };
   selectedWorkflowId?: string;
-  destination?: TabName | undefined;
+  selectedStylePresetId?: string;
+  destination?: InvokeTabName | undefined;
 }
 
-const App = ({ config = DEFAULT_CONFIG, selectedImage, selectedWorkflowId, destination }: Props) => {
+const App = ({
+  config = DEFAULT_CONFIG,
+  selectedImage,
+  selectedWorkflowId,
+  selectedStylePresetId,
+  destination,
+}: Props) => {
   const language = useAppSelector(languageSelector);
   const logger = useLogger('system');
   const dispatch = useAppDispatch();
@@ -83,6 +89,12 @@ const App = ({ config = DEFAULT_CONFIG, selectedImage, selectedWorkflowId, desti
     }
   }, [selectedWorkflowId, getAndLoadWorkflow]);
 
+  useEffect(() => {
+    if (selectedStylePresetId) {
+      dispatch(activeStylePresetIdChanged(selectedStylePresetId));
+    }
+  }, [dispatch, selectedStylePresetId]);
+
   useEffect(() => {
     if (destination) {
       dispatch(setActiveTab(destination));
@@ -95,7 +107,6 @@ const App = ({ config = DEFAULT_CONFIG, selectedImage, selectedWorkflowId, desti
 
   useStarterModelsToast();
   useSyncQueueStatus();
-  useScopeFocusWatcher();
 
   return (
     <ErrorBoundary onReset={handleReset} FallbackComponent={AppErrorBoundaryFallback}>
@@ -108,7 +119,7 @@ const App = ({ config = DEFAULT_CONFIG, selectedImage, selectedWorkflowId, desti
         {...dropzone.getRootProps()}
       >
         <input {...dropzone.getInputProps()} />
-        <AppContent />
+        <InvokeTabs />
         <AnimatePresence>
           {dropzone.isDragActive && isHandlingUpload && (
             <ImageUploadOverlay dropzone={dropzone} setIsHandlingUpload={setIsHandlingUpload} />
@@ -119,7 +130,6 @@ const App = ({ config = DEFAULT_CONFIG, selectedImage, selectedWorkflowId, desti
       <ChangeBoardModal />
       <DynamicPromptsModal />
       <StylePresetModal />
-      <ClearQueueConfirmationsAlertDialog />
       <PreselectedImage selectedImage={selectedImage} />
     </ErrorBoundary>
   );

invokeai/frontend/web/src/app/components/InvokeAIUI.tsx

@@ -19,7 +19,7 @@ import type { PartialAppConfig } from 'app/types/invokeai';
 import Loading from 'common/components/Loading/Loading';
 import AppDndContext from 'features/dnd/components/AppDndContext';
 import type { WorkflowCategory } from 'features/nodes/types/workflow';
-import type { TabName } from 'features/ui/store/uiTypes';
+import type { InvokeTabName } from 'features/ui/store/tabMap';
 import type { PropsWithChildren, ReactNode } from 'react';
 import React, { lazy, memo, useEffect, useMemo } from 'react';
 import { Provider } from 'react-redux';
@@ -45,7 +45,8 @@ interface Props extends PropsWithChildren {
     action: 'sendToImg2Img' | 'sendToCanvas' | 'useAllParameters';
   };
   selectedWorkflowId?: string;
-  destination?: TabName;
+  selectedStylePresetId?: string;
+  destination?: InvokeTabName;
   customStarUi?: CustomStarUi;
   socketOptions?: Partial<ManagerOptions & SocketOptions>;
   isDebugging?: boolean;
@@ -66,6 +67,7 @@ const InvokeAIUI = ({
   queueId,
   selectedImage,
   selectedWorkflowId,
+  selectedStylePresetId,
   destination,
   customStarUi,
   socketOptions,
@@ -227,6 +229,7 @@ const InvokeAIUI = ({
                 config={config}
                 selectedImage={selectedImage}
                 selectedWorkflowId={selectedWorkflowId}
+                selectedStylePresetId={selectedStylePresetId}
                 destination={destination}
               />
             </AppDndContext>

invokeai/frontend/web/src/app/hooks/useSocketIO.ts

@@ -2,7 +2,7 @@ import { useStore } from '@nanostores/react';
 import { $authToken } from 'app/store/nanostores/authToken';
 import { $baseUrl } from 'app/store/nanostores/baseUrl';
 import { $isDebugging } from 'app/store/nanostores/isDebugging';
-import { useAppStore } from 'app/store/nanostores/store';
+import { useAppDispatch } from 'app/store/storeHooks';
 import type { MapStore } from 'nanostores';
 import { atom, map } from 'nanostores';
 import { useEffect, useMemo } from 'react';
@@ -18,19 +18,14 @@ declare global {
   }
 }
 
-export type AppSocket = Socket<ServerToClientEvents, ClientToServerEvents>;
-
-export const $socket = atom<AppSocket | null>(null);
 export const $socketOptions = map<Partial<ManagerOptions & SocketOptions>>({});
-
 const $isSocketInitialized = atom<boolean>(false);
-export const $isConnected = atom<boolean>(false);
 
 /**
  * Initializes the socket.io connection and sets up event listeners.
  */
 export const useSocketIO = () => {
-  const { dispatch, getState } = useAppStore();
+  const dispatch = useAppDispatch();
   const baseUrl = useStore($baseUrl);
   const authToken = useStore($authToken);
   const addlSocketOptions = useStore($socketOptions);
@@ -66,9 +61,8 @@ export const useSocketIO = () => {
       return;
     }
 
-    const socket: AppSocket = io(socketUrl, socketOptions);
-    $socket.set(socket);
-    setEventListeners({ socket, dispatch, getState, setIsConnected: $isConnected.set });
+    const socket: Socket<ServerToClientEvents, ClientToServerEvents> = io(socketUrl, socketOptions);
+    setEventListeners({ dispatch, socket });
     socket.connect();
 
     if ($isDebugging.get() || import.meta.env.MODE === 'development') {
@@ -90,5 +84,5 @@ export const useSocketIO = () => {
       socket.disconnect();
       $isSocketInitialized.set(false);
     };
-  }, [dispatch, getState, socketOptions, socketUrl]);
+  }, [dispatch, socketOptions, socketUrl]);
 };

invokeai/frontend/web/src/app/logging/logger.ts

@@ -15,21 +15,21 @@ export const BASE_CONTEXT = {};
 
 export const $logger = atom<Logger>(Roarr.child(BASE_CONTEXT));
 
-export const zLogNamespace = z.enum([
-  'canvas',
-  'config',
-  'events',
-  'gallery',
-  'generation',
-  'metadata',
-  'models',
-  'system',
-  'queue',
-  'workflows',
-]);
-export type LogNamespace = z.infer<typeof zLogNamespace>;
+export type LoggerNamespace =
+  | 'images'
+  | 'models'
+  | 'config'
+  | 'canvas'
+  | 'generation'
+  | 'nodes'
+  | 'system'
+  | 'socketio'
+  | 'session'
+  | 'queue'
+  | 'dnd'
+  | 'controlLayers';
 
-export const logger = (namespace: LogNamespace) => $logger.get().child({ namespace });
+export const logger = (namespace: LoggerNamespace) => $logger.get().child({ namespace });
 
 export const zLogLevel = z.enum(['trace', 'debug', 'info', 'warn', 'error', 'fatal']);
 export type LogLevel = z.infer<typeof zLogLevel>;

invokeai/frontend/web/src/app/logging/useLogger.ts

@@ -3,34 +3,27 @@ import { useAppSelector } from 'app/store/storeHooks';
 import { useEffect, useMemo } from 'react';
 import { ROARR, Roarr } from 'roarr';
 
-import type { LogNamespace } from './logger';
+import type { LoggerNamespace } from './logger';
 import { $logger, BASE_CONTEXT, LOG_LEVEL_MAP, logger } from './logger';
 
-export const useLogger = (namespace: LogNamespace) => {
-  const logLevel = useAppSelector((s) => s.system.logLevel);
-  const logNamespaces = useAppSelector((s) => s.system.logNamespaces);
-  const logIsEnabled = useAppSelector((s) => s.system.logIsEnabled);
+export const useLogger = (namespace: LoggerNamespace) => {
+  const consoleLogLevel = useAppSelector((s) => s.system.consoleLogLevel);
+  const shouldLogToConsole = useAppSelector((s) => s.system.shouldLogToConsole);
 
   // The provided Roarr browser log writer uses localStorage to config logging to console
   useEffect(() => {
-    if (logIsEnabled) {
+    if (shouldLogToConsole) {
       // Enable console log output
       localStorage.setItem('ROARR_LOG', 'true');
 
       // Use a filter to show only logs of the given level
-      let filter = `context.logLevel:>=${LOG_LEVEL_MAP[logLevel]}`;
-      if (logNamespaces.length > 0) {
-        filter += ` AND (${logNamespaces.map((ns) => `context.namespace:${ns}`).join(' OR ')})`;
-      } else {
-        filter += ' AND context.namespace:undefined';
-      }
-      localStorage.setItem('ROARR_FILTER', filter);
+      localStorage.setItem('ROARR_FILTER', `context.logLevel:>=${LOG_LEVEL_MAP[consoleLogLevel]}`);
     } else {
       // Disable console log output
       localStorage.setItem('ROARR_LOG', 'false');
     }
     ROARR.write = createLogWriter();
-  }, [logLevel, logIsEnabled, logNamespaces]);
+  }, [consoleLogLevel, shouldLogToConsole]);
 
   // Update the module-scoped logger context as needed
   useEffect(() => {

invokeai/frontend/web/src/app/store/actions.ts

@@ -1,7 +1,7 @@
 import { createAction } from '@reduxjs/toolkit';
-import type { TabName } from 'features/ui/store/uiTypes';
+import type { InvokeTabName } from 'features/ui/store/tabMap';
 
 export const enqueueRequested = createAction<{
-  tabName: TabName;
+  tabName: InvokeTabName;
   prepend: boolean;
 }>('app/enqueueRequested');

invokeai/frontend/web/src/app/store/createMemoizedSelector.ts

@@ -1,6 +1,5 @@
 import { createDraftSafeSelectorCreator, createSelectorCreator, lruMemoize } from '@reduxjs/toolkit';
 import type { GetSelectorsOptions } from '@reduxjs/toolkit/dist/entities/state_selectors';
-import type { RootState } from 'app/store/store';
 import { isEqual } from 'lodash-es';
 
 /**
@@ -20,5 +19,3 @@ export const getSelectorsOptions: GetSelectorsOptions = {
     argsMemoize: lruMemoize,
   }),
 };
-
-export const createMemoizedAppSelector = createMemoizedSelector.withTypes<RootState>();

invokeai/frontend/web/src/app/store/enhancers/reduxRemember/errors.ts

@@ -1,4 +1,5 @@
 import { logger } from 'app/logging/logger';
+import { parseify } from 'common/util/serialize';
 import { PersistError, RehydrateError } from 'redux-remember';
 import { serializeError } from 'serialize-error';
 
@@ -40,6 +41,6 @@ export const errorHandler = (err: PersistError | RehydrateError) => {
   } else if (err instanceof RehydrateError) {
     log.error({ error: serializeError(err) }, 'Problem rehydrating state');
   } else {
-    log.error({ error: serializeError(err) }, 'Problem in persistence layer');
+    log.error({ error: parseify(err) }, 'Problem in persistence layer');
   }
 };

invokeai/frontend/web/src/app/store/middleware/devtools/actionSanitizer.ts

@@ -1,7 +1,9 @@
 import type { UnknownAction } from '@reduxjs/toolkit';
+import { deepClone } from 'common/util/deepClone';
 import { isAnyGraphBuilt } from 'features/nodes/store/actions';
 import { appInfoApi } from 'services/api/endpoints/appInfo';
 import type { Graph } from 'services/api/types';
+import { socketGeneratorProgress } from 'services/events/actions';
 
 export const actionSanitizer = <A extends UnknownAction>(action: A): A => {
   if (isAnyGraphBuilt(action)) {
@@ -22,5 +24,13 @@ export const actionSanitizer = <A extends UnknownAction>(action: A): A => {
     };
   }
 
+  if (socketGeneratorProgress.match(action)) {
+    const sanitized = deepClone(action);
+    if (sanitized.payload.data.progress_image) {
+      sanitized.payload.data.progress_image.dataURL = '<Progress image omitted>';
+    }
+    return sanitized;
+  }
+
   return action;
 };

invokeai/frontend/web/src/app/store/middleware/listenerMiddleware/index.ts

@@ -1,7 +1,7 @@
 import type { TypedStartListening } from '@reduxjs/toolkit';
 import { createListenerMiddleware } from '@reduxjs/toolkit';
 import { addAdHocPostProcessingRequestedListener } from 'app/store/middleware/listenerMiddleware/listeners/addAdHocPostProcessingRequestedListener';
-import { addStagingListeners } from 'app/store/middleware/listenerMiddleware/listeners/addCommitStagingAreaImageListener';
+import { addCommitStagingAreaImageListener } from 'app/store/middleware/listenerMiddleware/listeners/addCommitStagingAreaImageListener';
 import { addAnyEnqueuedListener } from 'app/store/middleware/listenerMiddleware/listeners/anyEnqueued';
 import { addAppConfigReceivedListener } from 'app/store/middleware/listenerMiddleware/listeners/appConfigReceived';
 import { addAppStartedListener } from 'app/store/middleware/listenerMiddleware/listeners/appStarted';
@@ -9,6 +9,17 @@ import { addBatchEnqueuedListener } from 'app/store/middleware/listenerMiddlewar
 import { addDeleteBoardAndImagesFulfilledListener } from 'app/store/middleware/listenerMiddleware/listeners/boardAndImagesDeleted';
 import { addBoardIdSelectedListener } from 'app/store/middleware/listenerMiddleware/listeners/boardIdSelected';
 import { addBulkDownloadListeners } from 'app/store/middleware/listenerMiddleware/listeners/bulkDownload';
+import { addCanvasCopiedToClipboardListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasCopiedToClipboard';
+import { addCanvasDownloadedAsImageListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasDownloadedAsImage';
+import { addCanvasImageToControlNetListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasImageToControlNet';
+import { addCanvasMaskSavedToGalleryListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasMaskSavedToGallery';
+import { addCanvasMaskToControlNetListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasMaskToControlNet';
+import { addCanvasMergedListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasMerged';
+import { addCanvasSavedToGalleryListener } from 'app/store/middleware/listenerMiddleware/listeners/canvasSavedToGallery';
+import { addControlAdapterPreprocessor } from 'app/store/middleware/listenerMiddleware/listeners/controlAdapterPreprocessor';
+import { addControlNetAutoProcessListener } from 'app/store/middleware/listenerMiddleware/listeners/controlNetAutoProcess';
+import { addControlNetImageProcessedListener } from 'app/store/middleware/listenerMiddleware/listeners/controlNetImageProcessed';
+import { addEnqueueRequestedCanvasListener } from 'app/store/middleware/listenerMiddleware/listeners/enqueueRequestedCanvas';
 import { addEnqueueRequestedLinear } from 'app/store/middleware/listenerMiddleware/listeners/enqueueRequestedLinear';
 import { addEnqueueRequestedNodes } from 'app/store/middleware/listenerMiddleware/listeners/enqueueRequestedNodes';
 import { addGalleryImageClickedListener } from 'app/store/middleware/listenerMiddleware/listeners/galleryImageClicked';
@@ -26,7 +37,16 @@ import { addModelSelectedListener } from 'app/store/middleware/listenerMiddlewar
 import { addModelsLoadedListener } from 'app/store/middleware/listenerMiddleware/listeners/modelsLoaded';
 import { addDynamicPromptsListener } from 'app/store/middleware/listenerMiddleware/listeners/promptChanged';
 import { addSetDefaultSettingsListener } from 'app/store/middleware/listenerMiddleware/listeners/setDefaultSettings';
-import { addSocketConnectedEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketConnected';
+import { addSocketConnectedEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketConnected';
+import { addSocketDisconnectedEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketDisconnected';
+import { addGeneratorProgressEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketGeneratorProgress';
+import { addInvocationCompleteEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketInvocationComplete';
+import { addInvocationErrorEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketInvocationError';
+import { addInvocationStartedEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketInvocationStarted';
+import { addModelInstallEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketModelInstall';
+import { addModelLoadEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketModelLoad';
+import { addSocketQueueItemStatusChangedEventListener } from 'app/store/middleware/listenerMiddleware/listeners/socketio/socketQueueItemStatusChanged';
+import { addStagingAreaImageSavedListener } from 'app/store/middleware/listenerMiddleware/listeners/stagingAreaImageSaved';
 import { addUpdateAllNodesRequestedListener } from 'app/store/middleware/listenerMiddleware/listeners/updateAllNodesRequested';
 import { addWorkflowLoadRequestedListener } from 'app/store/middleware/listenerMiddleware/listeners/workflowLoadRequested';
 import type { AppDispatch, RootState } from 'app/store/store';
@@ -63,6 +83,7 @@ addGalleryImageClickedListener(startAppListening);
 addGalleryOffsetChangedListener(startAppListening);
 
 // User Invoked
+addEnqueueRequestedCanvasListener(startAppListening);
 addEnqueueRequestedNodes(startAppListening);
 addEnqueueRequestedLinear(startAppListening);
 addEnqueueRequestedUpscale(startAppListening);
@@ -70,23 +91,32 @@ addAnyEnqueuedListener(startAppListening);
 addBatchEnqueuedListener(startAppListening);
 
 // Canvas actions
-// addCanvasSavedToGalleryListener(startAppListening);
-// addCanvasMaskSavedToGalleryListener(startAppListening);
-// addCanvasImageToControlNetListener(startAppListening);
-// addCanvasMaskToControlNetListener(startAppListening);
-// addCanvasDownloadedAsImageListener(startAppListening);
-// addCanvasCopiedToClipboardListener(startAppListening);
-// addCanvasMergedListener(startAppListening);
-// addStagingAreaImageSavedListener(startAppListening);
-// addCommitStagingAreaImageListener(startAppListening);
-addStagingListeners(startAppListening);
+addCanvasSavedToGalleryListener(startAppListening);
+addCanvasMaskSavedToGalleryListener(startAppListening);
+addCanvasImageToControlNetListener(startAppListening);
+addCanvasMaskToControlNetListener(startAppListening);
+addCanvasDownloadedAsImageListener(startAppListening);
+addCanvasCopiedToClipboardListener(startAppListening);
+addCanvasMergedListener(startAppListening);
+addStagingAreaImageSavedListener(startAppListening);
+addCommitStagingAreaImageListener(startAppListening);
 
 // Socket.IO
+addGeneratorProgressEventListener(startAppListening);
+addInvocationCompleteEventListener(startAppListening);
+addInvocationErrorEventListener(startAppListening);
+addInvocationStartedEventListener(startAppListening);
 addSocketConnectedEventListener(startAppListening);
-
-// Gallery bulk download
+addSocketDisconnectedEventListener(startAppListening);
+addModelLoadEventListener(startAppListening);
+addModelInstallEventListener(startAppListening);
+addSocketQueueItemStatusChangedEventListener(startAppListening);
 addBulkDownloadListeners(startAppListening);
 
+// ControlNet
+addControlNetImageProcessedListener(startAppListening);
+addControlNetAutoProcessListener(startAppListening);
+
 // Boards
 addImageAddedToBoardFulfilledListener(startAppListening);
 addImageRemovedFromBoardFulfilledListener(startAppListening);
@@ -118,4 +148,4 @@ addAdHocPostProcessingRequestedListener(startAppListening);
 addDynamicPromptsListener(startAppListening);
 
 addSetDefaultSettingsListener(startAppListening);
-// addControlAdapterPreprocessor(startAppListening);
+addControlAdapterPreprocessor(startAppListening);

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

@@ -1,21 +1,21 @@
 import { createAction } from '@reduxjs/toolkit';
 import { logger } from 'app/logging/logger';
 import type { AppStartListening } from 'app/store/middleware/listenerMiddleware';
-import type { SerializableObject } from 'common/types';
+import { parseify } from 'common/util/serialize';
 import { buildAdHocPostProcessingGraph } from 'features/nodes/util/graph/buildAdHocPostProcessingGraph';
 import { toast } from 'features/toast/toast';
 import { t } from 'i18next';
 import { queueApi } from 'services/api/endpoints/queue';
 import type { BatchConfig, ImageDTO } from 'services/api/types';
 
-const log = logger('queue');
-
 export const adHocPostProcessingRequested = createAction<{ imageDTO: ImageDTO }>(`upscaling/postProcessingRequested`);
 
 export const addAdHocPostProcessingRequestedListener = (startAppListening: AppStartListening) => {
   startAppListening({
     actionCreator: adHocPostProcessingRequested,
     effect: async (action, { dispatch, getState }) => {
+      const log = logger('session');
+
       const { imageDTO } = action.payload;
       const state = getState();
 
@@ -39,9 +39,9 @@ export const addAdHocPostProcessingRequestedListener = (startAppListening: AppSt
 
         const enqueueResult = await req.unwrap();
         req.reset();
-        log.debug({ enqueueResult } as SerializableObject, t('queue.graphQueued'));
+        log.debug({ enqueueResult: parseify(enqueueResult) }, t('queue.graphQueued'));
       } catch (error) {
-        log.error({ enqueueBatchArg } as SerializableObject, t('queue.graphFailedToQueue'));
+        log.error({ enqueueBatchArg: parseify(enqueueBatchArg) }, t('queue.graphFailedToQueue'));
 
         if (error instanceof Object && 'status' in error && error.status === 403) {
           return;

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

@@ -23,7 +23,7 @@ export const addArchivedOrDeletedBoardListener = (startAppListening: AppStartLis
    */
   startAppListening({
     matcher: matchAnyBoardDeleted,
-    effect: (action, { dispatch, getState }) => {
+    effect: async (action, { dispatch, getState }) => {
       const state = getState();
       const deletedBoardId = action.meta.arg.originalArgs;
       const { autoAddBoardId, selectedBoardId } = state.gallery;
@@ -44,7 +44,7 @@ export const addArchivedOrDeletedBoardListener = (startAppListening: AppStartLis
   // If we archived a board, it may end up hidden. If it's selected or the auto-add board, we should reset those.
   startAppListening({
     matcher: boardsApi.endpoints.updateBoard.matchFulfilled,
-    effect: (action, { dispatch, getState }) => {
+    effect: async (action, { dispatch, getState }) => {
       const state = getState();
       const { shouldShowArchivedBoards } = state.gallery;
 
@@ -61,7 +61,7 @@ export const addArchivedOrDeletedBoardListener = (startAppListening: AppStartLis
   // When we hide archived boards, if the selected or the auto-add board is archived, we should reset those.
   startAppListening({
     actionCreator: shouldShowArchivedBoardsChanged,
-    effect: (action, { dispatch, getState }) => {
+    effect: async (action, { dispatch, getState }) => {
       const shouldShowArchivedBoards = action.payload;
 
       // We only need to take action if we have just hidden archived boards.
@@ -100,7 +100,7 @@ export const addArchivedOrDeletedBoardListener = (startAppListening: AppStartLis
    */
   startAppListening({
     matcher: boardsApi.endpoints.listAllBoards.matchFulfilled,
-    effect: (action, { dispatch, getState }) => {
+    effect: async (action, { dispatch, getState }) => {
       const boards = action.payload;
       const state = getState();
       const { selectedBoardId, autoAddBoardId } = state.gallery;

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

@@ -1,36 +1,33 @@
+import { isAnyOf } from '@reduxjs/toolkit';
 import { logger } from 'app/logging/logger';
 import type { AppStartListening } from 'app/store/middleware/listenerMiddleware';
 import {
-  rasterLayerAdded,
-  sessionStagingAreaImageAccepted,
-  sessionStagingAreaReset,
-} from 'features/controlLayers/store/canvasV2Slice';
-import type { CanvasRasterLayerState } from 'features/controlLayers/store/types';
-import { imageDTOToImageObject } from 'features/controlLayers/store/types';
+  canvasBatchIdsReset,
+  commitStagingAreaImage,
+  discardStagedImages,
+  resetCanvas,
+  setInitialCanvasImage,
+} from 'features/canvas/store/canvasSlice';
 import { toast } from 'features/toast/toast';
 import { t } from 'i18next';
 import { queueApi } from 'services/api/endpoints/queue';
-import { $lastCanvasProgressEvent } from 'services/events/setEventListeners';
-import { assert } from 'tsafe';
 
-const log = logger('canvas');
+const matcher = isAnyOf(commitStagingAreaImage, discardStagedImages, resetCanvas, setInitialCanvasImage);
 
-export const addStagingListeners = (startAppListening: AppStartListening) => {
+export const addCommitStagingAreaImageListener = (startAppListening: AppStartListening) => {
   startAppListening({
-    actionCreator: sessionStagingAreaReset,
-    effect: async (_, { dispatch }) => {
+    matcher,
+    effect: async (_, { dispatch, getState }) => {
+      const log = logger('canvas');
+      const state = getState();
+      const { batchIds } = state.canvas;
+
       try {
         const req = dispatch(
-          queueApi.endpoints.cancelByBatchOrigin.initiate(
-            { origin: 'canvas' },
-            { fixedCacheKey: 'cancelByBatchOrigin' }
-          )
+          queueApi.endpoints.cancelByBatchIds.initiate({ batch_ids: batchIds }, { fixedCacheKey: 'cancelByBatchIds' })
         );
         const { canceled } = await req.unwrap();
         req.reset();
-
-        $lastCanvasProgressEvent.set(null);
-
         if (canceled > 0) {
           log.debug(`Canceled ${canceled} canvas batches`);
           toast({
@@ -39,6 +36,7 @@ export const addStagingListeners = (startAppListening: AppStartListening) => {
             status: 'success',
           });
         }
+        dispatch(canvasBatchIdsReset());
       } catch {
         log.error('Failed to cancel canvas batches');
         toast({
@@ -49,26 +47,4 @@ export const addStagingListeners = (startAppListening: AppStartListening) => {
       }
     },
   });
-
-  startAppListening({
-    actionCreator: sessionStagingAreaImageAccepted,
-    effect: (action, api) => {
-      const { index } = action.payload;
-      const state = api.getState();
-      const stagingAreaImage = state.canvasV2.session.stagedImages[index];
-
-      assert(stagingAreaImage, 'No staged image found to accept');
-      const { x, y } = state.canvasV2.bbox.rect;
-
-      const { imageDTO, offsetX, offsetY } = stagingAreaImage;
-      const imageObject = imageDTOToImageObject(imageDTO);
-      const overrides: Partial<CanvasRasterLayerState> = {
-        position: { x: x + offsetX, y: y + offsetY },
-        objects: [imageObject],
-      };
-
-      api.dispatch(rasterLayerAdded({ overrides, isSelected: true }));
-      api.dispatch(sessionStagingAreaReset());
-    },
-  });
 };

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

@@ -4,7 +4,7 @@ import { queueApi, selectQueueStatus } from 'services/api/endpoints/queue';
 export const addAnyEnqueuedListener = (startAppListening: AppStartListening) => {
   startAppListening({
     matcher: queueApi.endpoints.enqueueBatch.matchFulfilled,
-    effect: (_, { dispatch, getState }) => {
+    effect: async (_, { dispatch, getState }) => {
       const { data } = selectQueueStatus(getState());
 
       if (!data || data.processor.is_started) {

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

@@ -1,14 +1,14 @@
 import type { AppStartListening } from 'app/store/middleware/listenerMiddleware';
-import { setInfillMethod } from 'features/controlLayers/store/canvasV2Slice';
+import { setInfillMethod } from 'features/parameters/store/generationSlice';
 import { shouldUseNSFWCheckerChanged, shouldUseWatermarkerChanged } from 'features/system/store/systemSlice';
 import { appInfoApi } from 'services/api/endpoints/appInfo';
 
 export const addAppConfigReceivedListener = (startAppListening: AppStartListening) => {
   startAppListening({
     matcher: appInfoApi.endpoints.getAppConfig.matchFulfilled,
-    effect: (action, { getState, dispatch }) => {
+    effect: async (action, { getState, dispatch }) => {
       const { infill_methods = [], nsfw_methods = [], watermarking_methods = [] } = action.payload;
-      const infillMethod = getState().canvasV2.compositing.infillMethod;
+      const infillMethod = getState().generation.infillMethod;
 
       if (!infill_methods.includes(infillMethod)) {
         // if there is no infill method, set it to the first one