fix

src/diffusers/__init__.py

@@ -297,8 +297,6 @@ else:
             "ComponentSpec",
             "ModularPipeline",
             "ModularPipelineBlocks",
-            "InputParam",
-            "OutputParam",
         ]
     )
     _import_structure["optimization"] = [
@@ -1062,7 +1060,7 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
             ZImageTransformer2DModel,
             attention_backend,
         )
-        from .modular_pipelines import ComponentsManager, ComponentSpec, ModularPipeline, ModularPipelineBlocks, InputParam, OutputParam
+        from .modular_pipelines import ComponentsManager, ComponentSpec, ModularPipeline, ModularPipelineBlocks
         from .optimization import (
             get_constant_schedule,
             get_constant_schedule_with_warmup,

src/diffusers/modular_pipelines/flux/__init__.py

@@ -31,7 +31,9 @@ else:
         "FluxAutoBeforeDenoiseStep",
         "FluxAutoBlocks",
         "FluxAutoDecodeStep",
+        "FluxAutoDenoiseStep",
         "FluxKontextAutoBlocks",
+        "FluxKontextAutoDenoiseStep",
         "FluxKontextBeforeDenoiseStep",
     ]
     _import_structure["modular_pipeline"] = ["FluxKontextModularPipeline", "FluxModularPipeline"]
@@ -53,7 +55,9 @@ if TYPE_CHECKING or DIFFUSERS_SLOW_IMPORT:
             FluxAutoBeforeDenoiseStep,
             FluxAutoBlocks,
             FluxAutoDecodeStep,
+            FluxAutoDenoiseStep,
             FluxKontextAutoBlocks,
+            FluxKontextAutoDenoiseStep,
             FluxKontextBeforeDenoiseStep,
         )
         from .modular_pipeline import FluxKontextModularPipeline, FluxModularPipeline

src/diffusers/modular_pipelines/flux/modular_blocks.py

@@ -201,6 +201,37 @@ class FluxKontextAutoBeforeDenoiseStep(AutoPipelineBlocks):
         )
 
 
+# denoise: text2image
+class FluxAutoDenoiseStep(AutoPipelineBlocks):
+    block_classes = [FluxDenoiseStep]
+    block_names = ["denoise"]
+    block_trigger_inputs = [None]
+
+    @property
+    def description(self) -> str:
+        return (
+            "Denoise step that iteratively denoise the latents. "
+            "This is a auto pipeline block that works for text2image and img2img tasks."
+            " - `FluxDenoiseStep` (denoise) for text2image and img2img tasks."
+        )
+
+
+# denoise: Flux Kontext
+
+
+class FluxKontextAutoDenoiseStep(AutoPipelineBlocks):
+    block_classes = [FluxKontextDenoiseStep]
+    block_names = ["denoise"]
+    block_trigger_inputs = [None]
+
+    @property
+    def description(self) -> str:
+        return (
+            "Denoise step that iteratively denoise the latents for Flux Kontext. "
+            "This is a auto pipeline block that works for text2image and img2img tasks."
+            " - `FluxDenoiseStep` (denoise) for text2image and img2img tasks."
+        )
+
 
 # decode: all task (text2img, img2img)
 class FluxAutoDecodeStep(AutoPipelineBlocks):
@@ -291,7 +322,7 @@ class FluxKontextAutoInputStep(AutoPipelineBlocks):
 
 class FluxCoreDenoiseStep(SequentialPipelineBlocks):
     model_name = "flux"
-    block_classes = [FluxAutoInputStep, FluxAutoBeforeDenoiseStep, FluxDenoiseStep]
+    block_classes = [FluxAutoInputStep, FluxAutoBeforeDenoiseStep, FluxAutoDenoiseStep]
     block_names = ["input", "before_denoise", "denoise"]
 
     @property
@@ -300,7 +331,7 @@ class FluxCoreDenoiseStep(SequentialPipelineBlocks):
             "Core step that performs the denoising process. \n"
             + " - `FluxAutoInputStep` (input) standardizes the inputs for the denoising step.\n"
             + " - `FluxAutoBeforeDenoiseStep` (before_denoise) prepares the inputs for the denoising step.\n"
-            + " - `FluxDenoiseStep` (denoise) iteratively denoises the latents.\n"
+            + " - `FluxAutoDenoiseStep` (denoise) iteratively denoises the latents.\n"
             + "This step supports text-to-image and image-to-image tasks for Flux:\n"
             + " - for image-to-image generation, you need to provide `image_latents`\n"
             + " - for text-to-image generation, all you need to provide is prompt embeddings."
@@ -309,7 +340,7 @@ class FluxCoreDenoiseStep(SequentialPipelineBlocks):
 
 class FluxKontextCoreDenoiseStep(SequentialPipelineBlocks):
     model_name = "flux-kontext"
-    block_classes = [FluxKontextAutoInputStep, FluxKontextAutoBeforeDenoiseStep, FluxKontextDenoiseStep]
+    block_classes = [FluxKontextAutoInputStep, FluxKontextAutoBeforeDenoiseStep, FluxKontextAutoDenoiseStep]
     block_names = ["input", "before_denoise", "denoise"]
 
     @property
@@ -318,7 +349,7 @@ class FluxKontextCoreDenoiseStep(SequentialPipelineBlocks):
             "Core step that performs the denoising process. \n"
             + " - `FluxKontextAutoInputStep` (input) standardizes the inputs for the denoising step.\n"
             + " - `FluxKontextAutoBeforeDenoiseStep` (before_denoise) prepares the inputs for the denoising step.\n"
-            + " - `FluxKontextDenoiseStep` (denoise) iteratively denoises the latents.\n"
+            + " - `FluxKontextAutoDenoiseStep` (denoise) iteratively denoises the latents.\n"
             + "This step supports text-to-image and image-to-image tasks for Flux:\n"
             + " - for image-to-image generation, you need to provide `image_latents`\n"
             + " - for text-to-image generation, all you need to provide is prompt embeddings."

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_diffusion_forcing.py

@@ -18,7 +18,6 @@ import re
 from copy import deepcopy
 from typing import Any, Callable, Dict, List, Optional, Union
 
-import ftfy
 import torch
 from transformers import AutoTokenizer, UMT5EncoderModel
 

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_diffusion_forcing_i2v.py

@@ -18,7 +18,6 @@ import re
 from copy import deepcopy
 from typing import Any, Callable, Dict, List, Optional, Tuple, Union
 
-import ftfy
 import PIL
 import torch
 from transformers import AutoTokenizer, UMT5EncoderModel

src/diffusers/pipelines/skyreels_v2/pipeline_skyreels_v2_diffusion_forcing_v2v.py

@@ -19,7 +19,6 @@ import re
 from copy import deepcopy
 from typing import Any, Callable, Dict, List, Optional, Union
 
-import ftfy
 import torch
 from PIL import Image
 from transformers import AutoTokenizer, UMT5EncoderModel