InvokeAI 4.2.6rc1__tar.gz → 4.2.7__tar.gz

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/InvokeAI.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: InvokeAI
-Version: 4.2.6rc1
+Version: 4.2.7
 Summary: An implementation of Stable Diffusion which provides various new features and options to aid the image generation process
 Author-email: The InvokeAI Project <lincoln.stein@gmail.com>
 License:                                  Apache License

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/InvokeAI.egg-info/SOURCES.txt

@@ -56,6 +56,7 @@ invokeai/app/invocations/prompt.py
 invokeai/app/invocations/resize_latents.py
 invokeai/app/invocations/scheduler.py
 invokeai/app/invocations/sdxl.py
+invokeai/app/invocations/spandrel_image_to_image.py
 invokeai/app/invocations/strings.py
 invokeai/app/invocations/t2i_adapter.py
 invokeai/app/invocations/tiled_multi_diffusion_denoise_latents.py
@@ -187,6 +188,8 @@ invokeai/app/services/workflow_records/workflow_records_sqlite.py
 invokeai/app/services/workflow_records/default_workflows/ESRGAN Upscaling with Canny ControlNet.json
 invokeai/app/services/workflow_records/default_workflows/Face Detailer with IP-Adapter & Canny (See Note in Details).json
 invokeai/app/services/workflow_records/default_workflows/Multi ControlNet (Canny & Depth).json
+invokeai/app/services/workflow_records/default_workflows/MultiDiffusion SD1.5.json
+invokeai/app/services/workflow_records/default_workflows/MultiDiffusion SDXL.json
 invokeai/app/services/workflow_records/default_workflows/Prompt from File.json
 invokeai/app/services/workflow_records/default_workflows/Text to Image - SD1.5.json
 invokeai/app/services/workflow_records/default_workflows/Text to Image - SDXL.json
@@ -210,6 +213,7 @@ invokeai/backend/__init__.py
 invokeai/backend/lora.py
 invokeai/backend/model_patcher.py
 invokeai/backend/raw_model.py
+invokeai/backend/spandrel_image_to_image_model.py
 invokeai/backend/textual_inversion.py
 invokeai/backend/image_util/__init__.py
 invokeai/backend/image_util/canny.py
@@ -268,6 +272,7 @@ invokeai/backend/model_manager/load/model_loaders/generic_diffusers.py
 invokeai/backend/model_manager/load/model_loaders/ip_adapter.py
 invokeai/backend/model_manager/load/model_loaders/lora.py
 invokeai/backend/model_manager/load/model_loaders/onnx.py
+invokeai/backend/model_manager/load/model_loaders/spandrel_image_to_image.py
 invokeai/backend/model_manager/load/model_loaders/stable_diffusion.py
 invokeai/backend/model_manager/load/model_loaders/textual_inversion.py
 invokeai/backend/model_manager/load/model_loaders/vae.py
@@ -281,7 +286,11 @@ invokeai/backend/model_manager/util/model_util.py
 invokeai/backend/model_manager/util/select_hf_files.py
 invokeai/backend/onnx/onnx_runtime.py
 invokeai/backend/stable_diffusion/__init__.py
+invokeai/backend/stable_diffusion/denoise_context.py
 invokeai/backend/stable_diffusion/diffusers_pipeline.py
+invokeai/backend/stable_diffusion/diffusion_backend.py
+invokeai/backend/stable_diffusion/extension_callback_type.py
+invokeai/backend/stable_diffusion/extensions_manager.py
 invokeai/backend/stable_diffusion/multi_diffusion_pipeline.py
 invokeai/backend/stable_diffusion/seamless.py
 invokeai/backend/stable_diffusion/vae_tiling.py
@@ -292,6 +301,11 @@ invokeai/backend/stable_diffusion/diffusion/regional_ip_data.py
 invokeai/backend/stable_diffusion/diffusion/regional_prompt_data.py
 invokeai/backend/stable_diffusion/diffusion/shared_invokeai_diffusion.py
 invokeai/backend/stable_diffusion/diffusion/unet_attention_patcher.py
+invokeai/backend/stable_diffusion/extensions/base.py
+invokeai/backend/stable_diffusion/extensions/controlnet.py
+invokeai/backend/stable_diffusion/extensions/freeu.py
+invokeai/backend/stable_diffusion/extensions/preview.py
+invokeai/backend/stable_diffusion/extensions/rescale_cfg.py
 invokeai/backend/stable_diffusion/schedulers/__init__.py
 invokeai/backend/stable_diffusion/schedulers/schedulers.py
 invokeai/backend/tiles/__init__.py
@@ -333,10 +347,10 @@ invokeai/frontend/install/import_images.py
 invokeai/frontend/web/__init__.py
 invokeai/frontend/web/dist/index.html
 invokeai/frontend/web/dist/assets/App-DEu4J2pT.css
-invokeai/frontend/web/dist/assets/App-DOTNSDj9.js
-invokeai/frontend/web/dist/assets/ThemeLocaleProvider-Dfzfg7Ib.js
+invokeai/frontend/web/dist/assets/App-Ds9ZtYAt.js
+invokeai/frontend/web/dist/assets/ThemeLocaleProvider-Bn2q0fTt.js
 invokeai/frontend/web/dist/assets/ThemeLocaleProvider-DzjsLZSc.css
-invokeai/frontend/web/dist/assets/index-BmvE-xpm.js
+invokeai/frontend/web/dist/assets/index-DkRlfOU4.js
 invokeai/frontend/web/dist/assets/inter-cyrillic-ext-wght-normal-DIEz8p5i.woff2
 invokeai/frontend/web/dist/assets/inter-cyrillic-wght-normal-BmJJXa8e.woff2
 invokeai/frontend/web/dist/assets/inter-greek-ext-wght-normal-D5AYLNiq.woff2

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/InvokeAI.egg-info/requires.txt

@@ -11,13 +11,14 @@ onnxruntime==1.16.3
 opencv-python==4.9.0.80
 pytorch-lightning==2.1.3
 safetensors==0.4.3
+spandrel==0.3.4
 timm==0.6.13
 torch==2.2.2
 torchmetrics==0.11.4
 torchsde==0.2.6
 torchvision==0.17.2
 transformers==4.41.1
-fastapi-events==0.11.0
+fastapi-events==0.11.1
 fastapi==0.111.0
 huggingface-hub==0.23.1
 pydantic-settings==2.2.1

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: InvokeAI
-Version: 4.2.6rc1
+Version: 4.2.7
 Summary: An implementation of Stable Diffusion which provides various new features and options to aid the image generation process
 Author-email: The InvokeAI Project <lincoln.stein@gmail.com>
 License:                                  Apache License

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/invokeai/app/api/routers/model_manager.py

@@ -6,7 +6,7 @@ import pathlib
 import traceback
 from copy import deepcopy
 from tempfile import TemporaryDirectory
-from typing import Any, Dict, List, Optional, Type
+from typing import List, Optional, Type
 
 from fastapi import Body, Path, Query, Response, UploadFile
 from fastapi.responses import FileResponse, HTMLResponse
@@ -430,13 +430,11 @@ async def delete_model_image(
 async def install_model(
     source: str = Query(description="Model source to install, can be a local path, repo_id, or remote URL"),
     inplace: Optional[bool] = Query(description="Whether or not to install a local model in place", default=False),
-    # TODO(MM2): Can we type this?
-    config: Optional[Dict[str, Any]] = Body(
-        description="Dict of fields that override auto-probed values in the model config record, such as name, description and prediction_type ",
-        default=None,
+    access_token: Optional[str] = Query(description="access token for the remote resource", default=None),
+    config: ModelRecordChanges = Body(
+        description="Object containing fields that override auto-probed values in the model config record, such as name, description and prediction_type ",
         example={"name": "string", "description": "string"},
     ),
-    access_token: Optional[str] = None,
 ) -> ModelInstallJob:
     """Install a model using a string identifier.
 
@@ -451,8 +449,9 @@ async def install_model(
        - model/name:fp16:path/to/model.safetensors
        - model/name::path/to/model.safetensors
 
-    `config` is an optional dict containing model configuration values that will override
-    the ones that are probed automatically.
+    `config` is a ModelRecordChanges object. Fields in this object will override
+    the ones that are probed automatically. Pass an empty object to accept
+    all the defaults.
 
     `access_token` is an optional access token for use with Urls that require
     authentication.
@@ -737,7 +736,7 @@ async def convert_model(
         # write the converted file to the convert path
         raw_model = converted_model.model
         assert hasattr(raw_model, "save_pretrained")
-        raw_model.save_pretrained(convert_path)
+        raw_model.save_pretrained(convert_path)  # type: ignore
         assert convert_path.exists()
 
         # temporarily rename the original safetensors file so that there is no naming conflict
@@ -750,12 +749,12 @@ async def convert_model(
         try:
             new_key = installer.install_path(
                 convert_path,
-                config={
-                    "name": original_name,
-                    "description": model_config.description,
-                    "hash": model_config.hash,
-                    "source": model_config.source,
-                },
+                config=ModelRecordChanges(
+                    name=original_name,
+                    description=model_config.description,
+                    hash=model_config.hash,
+                    source=model_config.source,
+                ),
             )
         except Exception as e:
             logger.error(str(e))

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/invokeai/app/invocations/denoise_latents.py

@@ -1,5 +1,6 @@
 # Copyright (c) 2023 Kyle Schouviller (https://github.com/kyle0654)
 import inspect
+import os
 from contextlib import ExitStack
 from typing import Any, Dict, Iterator, List, Optional, Tuple, Union
 
@@ -39,6 +40,7 @@ from invokeai.backend.lora import LoRAModelRaw
 from invokeai.backend.model_manager import BaseModelType
 from invokeai.backend.model_patcher import ModelPatcher
 from invokeai.backend.stable_diffusion import PipelineIntermediateState, set_seamless
+from invokeai.backend.stable_diffusion.denoise_context import DenoiseContext, DenoiseInputs
 from invokeai.backend.stable_diffusion.diffusers_pipeline import (
     ControlNetData,
     StableDiffusionGeneratorPipeline,
@@ -53,6 +55,14 @@ from invokeai.backend.stable_diffusion.diffusion.conditioning_data import (
     TextConditioningData,
     TextConditioningRegions,
 )
+from invokeai.backend.stable_diffusion.diffusion.custom_atttention import CustomAttnProcessor2_0
+from invokeai.backend.stable_diffusion.diffusion_backend import StableDiffusionBackend
+from invokeai.backend.stable_diffusion.extension_callback_type import ExtensionCallbackType
+from invokeai.backend.stable_diffusion.extensions.controlnet import ControlNetExt
+from invokeai.backend.stable_diffusion.extensions.freeu import FreeUExt
+from invokeai.backend.stable_diffusion.extensions.preview import PreviewExt
+from invokeai.backend.stable_diffusion.extensions.rescale_cfg import RescaleCFGExt
+from invokeai.backend.stable_diffusion.extensions_manager import ExtensionsManager
 from invokeai.backend.stable_diffusion.schedulers import SCHEDULER_MAP
 from invokeai.backend.stable_diffusion.schedulers.schedulers import SCHEDULER_NAME_VALUES
 from invokeai.backend.util.devices import TorchDevice
@@ -314,9 +324,10 @@ class DenoiseLatentsInvocation(BaseInvocation):
         context: InvocationContext,
         positive_conditioning_field: Union[ConditioningField, list[ConditioningField]],
         negative_conditioning_field: Union[ConditioningField, list[ConditioningField]],
-        unet: UNet2DConditionModel,
         latent_height: int,
         latent_width: int,
+        device: torch.device,
+        dtype: torch.dtype,
         cfg_scale: float | list[float],
         steps: int,
         cfg_rescale_multiplier: float,
@@ -330,10 +341,10 @@ class DenoiseLatentsInvocation(BaseInvocation):
             uncond_list = [uncond_list]
 
         cond_text_embeddings, cond_text_embedding_masks = DenoiseLatentsInvocation._get_text_embeddings_and_masks(
-            cond_list, context, unet.device, unet.dtype
+            cond_list, context, device, dtype
         )
         uncond_text_embeddings, uncond_text_embedding_masks = DenoiseLatentsInvocation._get_text_embeddings_and_masks(
-            uncond_list, context, unet.device, unet.dtype
+            uncond_list, context, device, dtype
         )
 
         cond_text_embedding, cond_regions = DenoiseLatentsInvocation._concat_regional_text_embeddings(
@@ -341,14 +352,14 @@ class DenoiseLatentsInvocation(BaseInvocation):
             masks=cond_text_embedding_masks,
             latent_height=latent_height,
             latent_width=latent_width,
-            dtype=unet.dtype,
+            dtype=dtype,
         )
         uncond_text_embedding, uncond_regions = DenoiseLatentsInvocation._concat_regional_text_embeddings(
             text_conditionings=uncond_text_embeddings,
             masks=uncond_text_embedding_masks,
             latent_height=latent_height,
             latent_width=latent_width,
-            dtype=unet.dtype,
+            dtype=dtype,
         )
 
         if isinstance(cfg_scale, list):
@@ -455,6 +466,38 @@ class DenoiseLatentsInvocation(BaseInvocation):
 
         return controlnet_data
 
+    @staticmethod
+    def parse_controlnet_field(
+        exit_stack: ExitStack,
+        context: InvocationContext,
+        control_input: ControlField | list[ControlField] | None,
+        ext_manager: ExtensionsManager,
+    ) -> None:
+        # Normalize control_input to a list.
+        control_list: list[ControlField]
+        if isinstance(control_input, ControlField):
+            control_list = [control_input]
+        elif isinstance(control_input, list):
+            control_list = control_input
+        elif control_input is None:
+            control_list = []
+        else:
+            raise ValueError(f"Unexpected control_input type: {type(control_input)}")
+
+        for control_info in control_list:
+            model = exit_stack.enter_context(context.models.load(control_info.control_model))
+            ext_manager.add_extension(
+                ControlNetExt(
+                    model=model,
+                    image=context.images.get_pil(control_info.image.image_name),
+                    weight=control_info.control_weight,
+                    begin_step_percent=control_info.begin_step_percent,
+                    end_step_percent=control_info.end_step_percent,
+                    control_mode=control_info.control_mode,
+                    resize_mode=control_info.resize_mode,
+                )
+            )
+
     def prep_ip_adapter_image_prompts(
         self,
         context: InvocationContext,
@@ -707,9 +750,124 @@ class DenoiseLatentsInvocation(BaseInvocation):
 
         return seed, noise, latents
 
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        if os.environ.get("USE_MODULAR_DENOISE", False):
+            return self._new_invoke(context)
+        else:
+            return self._old_invoke(context)
+
     @torch.no_grad()
     @SilenceWarnings()  # This quenches the NSFW nag from diffusers.
-    def invoke(self, context: InvocationContext) -> LatentsOutput:
+    def _new_invoke(self, context: InvocationContext) -> LatentsOutput:
+        ext_manager = ExtensionsManager(is_canceled=context.util.is_canceled)
+
+        device = TorchDevice.choose_torch_device()
+        dtype = TorchDevice.choose_torch_dtype()
+
+        seed, noise, latents = self.prepare_noise_and_latents(context, self.noise, self.latents)
+        latents = latents.to(device=device, dtype=dtype)
+        if noise is not None:
+            noise = noise.to(device=device, dtype=dtype)
+
+        _, _, latent_height, latent_width = latents.shape
+
+        conditioning_data = self.get_conditioning_data(
+            context=context,
+            positive_conditioning_field=self.positive_conditioning,
+            negative_conditioning_field=self.negative_conditioning,
+            cfg_scale=self.cfg_scale,
+            steps=self.steps,
+            latent_height=latent_height,
+            latent_width=latent_width,
+            device=device,
+            dtype=dtype,
+            # TODO: old backend, remove
+            cfg_rescale_multiplier=self.cfg_rescale_multiplier,
+        )
+
+        scheduler = get_scheduler(
+            context=context,
+            scheduler_info=self.unet.scheduler,
+            scheduler_name=self.scheduler,
+            seed=seed,
+        )
+
+        timesteps, init_timestep, scheduler_step_kwargs = self.init_scheduler(
+            scheduler,
+            seed=seed,
+            device=device,
+            steps=self.steps,
+            denoising_start=self.denoising_start,
+            denoising_end=self.denoising_end,
+        )
+
+        denoise_ctx = DenoiseContext(
+            inputs=DenoiseInputs(
+                orig_latents=latents,
+                timesteps=timesteps,
+                init_timestep=init_timestep,
+                noise=noise,
+                seed=seed,
+                scheduler_step_kwargs=scheduler_step_kwargs,
+                conditioning_data=conditioning_data,
+                attention_processor_cls=CustomAttnProcessor2_0,
+            ),
+            unet=None,
+            scheduler=scheduler,
+        )
+
+        # get the unet's config so that we can pass the base to sd_step_callback()
+        unet_config = context.models.get_config(self.unet.unet.key)
+
+        ### preview
+        def step_callback(state: PipelineIntermediateState) -> None:
+            context.util.sd_step_callback(state, unet_config.base)
+
+        ext_manager.add_extension(PreviewExt(step_callback))
+
+        ### cfg rescale
+        if self.cfg_rescale_multiplier > 0:
+            ext_manager.add_extension(RescaleCFGExt(self.cfg_rescale_multiplier))
+
+        ### freeu
+        if self.unet.freeu_config:
+            ext_manager.add_extension(FreeUExt(self.unet.freeu_config))
+
+        # context for loading additional models
+        with ExitStack() as exit_stack:
+            # later should be smth like:
+            # for extension_field in self.extensions:
+            #    ext = extension_field.to_extension(exit_stack, context, ext_manager)
+            #    ext_manager.add_extension(ext)
+            self.parse_controlnet_field(exit_stack, context, self.control, ext_manager)
+
+            # ext: t2i/ip adapter
+            ext_manager.run_callback(ExtensionCallbackType.SETUP, denoise_ctx)
+
+            unet_info = context.models.load(self.unet.unet)
+            assert isinstance(unet_info.model, UNet2DConditionModel)
+            with (
+                unet_info.model_on_device() as (cached_weights, unet),
+                ModelPatcher.patch_unet_attention_processor(unet, denoise_ctx.inputs.attention_processor_cls),
+                # ext: controlnet
+                ext_manager.patch_extensions(denoise_ctx),
+                # ext: freeu, seamless, ip adapter, lora
+                ext_manager.patch_unet(unet, cached_weights),
+            ):
+                sd_backend = StableDiffusionBackend(unet, scheduler)
+                denoise_ctx.unet = unet
+                result_latents = sd_backend.latents_from_embeddings(denoise_ctx, ext_manager)
+
+        # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        result_latents = result_latents.detach().to("cpu")
+        TorchDevice.empty_cache()
+
+        name = context.tensors.save(tensor=result_latents)
+        return LatentsOutput.build(latents_name=name, latents=result_latents, seed=None)
+
+    @torch.no_grad()
+    @SilenceWarnings()  # This quenches the NSFW nag from diffusers.
+    def _old_invoke(self, context: InvocationContext) -> LatentsOutput:
         seed, noise, latents = self.prepare_noise_and_latents(context, self.noise, self.latents)
 
         mask, masked_latents, gradient_mask = self.prep_inpaint_mask(context, latents)
@@ -788,7 +946,8 @@ class DenoiseLatentsInvocation(BaseInvocation):
                 context=context,
                 positive_conditioning_field=self.positive_conditioning,
                 negative_conditioning_field=self.negative_conditioning,
-                unet=unet,
+                device=unet.device,
+                dtype=unet.dtype,
                 latent_height=latent_height,
                 latent_width=latent_width,
                 cfg_scale=self.cfg_scale,

{InvokeAI-4.2.6rc1 → InvokeAI-4.2.7}/invokeai/app/invocations/fields.py

@@ -48,6 +48,7 @@ class UIType(str, Enum, metaclass=MetaEnum):
     ControlNetModel = "ControlNetModelField"
     IPAdapterModel = "IPAdapterModelField"
     T2IAdapterModel = "T2IAdapterModelField"
+    SpandrelImageToImageModel = "SpandrelImageToImageModelField"
     # endregion
 
     # region Misc Field Types
@@ -134,6 +135,7 @@ class FieldDescriptions:
     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"
+    spandrel_image_to_image_model = "Image-to-Image model"
     lora_weight = "The weight at which the LoRA is applied to each model"
     compel_prompt = "Prompt to be parsed by Compel to create a conditioning tensor"
     raw_prompt = "Raw prompt text (no parsing)"