From 983cb5ebd2b29a17df44674365d8cf229bb01a92 Mon Sep 17 00:00:00 2001
From: Kent Keirsey <31807370+hipsterusername@users.noreply.github.com>
Date: Thu, 3 Jul 2025 14:17:53 -0400
Subject: [PATCH] ruff ruff

---
 invokeai/backend/flux/denoise.py               | 18 ++++++++++--------
 .../flux/extensions/kontext_extension.py       |  2 +-
 2 files changed, 11 insertions(+), 9 deletions(-)

diff --git a/invokeai/backend/flux/denoise.py b/invokeai/backend/flux/denoise.py
index 7939900879..2ed34d88b9 100644
--- a/invokeai/backend/flux/denoise.py
+++ b/invokeai/backend/flux/denoise.py
@@ -49,10 +49,10 @@ def denoise(
     )
     # guidance_vec is ignored for schnell.
     guidance_vec = torch.full((img.shape[0],), guidance, device=img.device, dtype=img.dtype)
-    
+
     # Store original sequence length for slicing predictions
     original_seq_len = img.shape[1]
-    
+
     for step_index, (t_curr, t_prev) in tqdm(list(enumerate(zip(timesteps[:-1], timesteps[1:], strict=True)))):
         t_vec = torch.full((img.shape[0],), t_curr, dtype=img.dtype, device=img.device)
 
@@ -78,21 +78,23 @@ def denoise(
         # controlnet_residuals datastructure is efficient in that it likely contains multiple references to the same
         # tensors. Calculating the sum materializes each tensor into its own instance.
         merged_controlnet_residuals = sum_controlnet_flux_outputs(controlnet_residuals)
-        
+
         # Prepare input for model - concatenate fresh each step
         img_input = img
         img_input_ids = img_ids
-        
+
         # Add channel-wise conditioning (for ControlNet, FLUX Fill, etc.)
         if img_cond is not None:
             img_input = torch.cat((img_input, img_cond), dim=-1)
-        
+
         # Add sequence-wise conditioning (for Kontext)
         if img_cond_seq is not None:
-            assert img_cond_seq_ids is not None, "You need to provide either both or neither of the sequence conditioning"
+            assert img_cond_seq_ids is not None, (
+                "You need to provide either both or neither of the sequence conditioning"
+            )
             img_input = torch.cat((img_input, img_cond_seq), dim=1)
             img_input_ids = torch.cat((img_input_ids, img_cond_seq_ids), dim=1)
-        
+
         pred = model(
             img=img_input,
             img_ids=img_input_ids,
@@ -108,7 +110,7 @@ def denoise(
             ip_adapter_extensions=pos_ip_adapter_extensions,
             regional_prompting_extension=pos_regional_prompting_extension,
         )
-        
+
         # Slice prediction to only include the main image tokens
         if img_input_ids is not None:
             pred = pred[:, :original_seq_len]
diff --git a/invokeai/backend/flux/extensions/kontext_extension.py b/invokeai/backend/flux/extensions/kontext_extension.py
index 802d49d451..bf2062401d 100644
--- a/invokeai/backend/flux/extensions/kontext_extension.py
+++ b/invokeai/backend/flux/extensions/kontext_extension.py
@@ -1,7 +1,7 @@
 import einops
+import numpy as np
 import torch
 from einops import repeat
-import numpy as np
 from PIL import Image
 
 from invokeai.app.invocations.fields import FluxKontextConditioningField