Revert "much stricter ruff linting"

This reverts commit 310dbf1574.

CHANGELOG.md

@@ -1,12 +1,12 @@
 # Change Log for SD.Next
 
-## Update for 2026-05-11
+## Update for 2026-05-10
 
-### Highlights for 2026-05-11
+### Highlights for 2026-05-10
 
 *What's New?*
 - Image editing models now can work with multiple image inputs!
-- Five new base models: *HiDream-O1 Image*, *JoyAI Image Edit*, *Step1X-Edit*, *VIBE Image Edit* and *UltraFlux*
+- New models: *HiDream-O1 Image*, *JoyAI Image Edit*, *Step1X-Edit*, *VIBE Image Edit* and *UltraFlux*
 - Enhanced capabilities for *Anima*, *Ernie-Image*, *LTX*, *Flux.2* and *Chroma* models
 - UI improvements accross the board: *Main panels*, *Gallery*, *Kanvas*, and more...
 
@@ -14,7 +14,7 @@ For full details, see [ChangeLog](https://github.com/vladmandic/automatic/blob/m
 
 [ReadMe](https://github.com/vladmandic/automatic/blob/master/README.md) | [ChangeLog](https://github.com/vladmandic/automatic/blob/master/CHANGELOG.md) | [Docs](https://vladmandic.github.io/sdnext-docs/) | [WiKi](https://github.com/vladmandic/automatic/wiki) | [Discord](https://discord.com/invite/sd-next-federal-batch-inspectors-1101998836328697867) | [Sponsor](https://github.com/sponsors/vladmandic)  
 
-### Details for 2026-05-11
+### Details for 2026-05-10
 
 - **Models**
   - [HiDream-O1-Image](https://huggingface.co/HiDream-ai/HiDream-O1-Image) pixel-level unified transformer model support  
@@ -22,7 +22,6 @@ For full details, see [ChangeLog](https://github.com/vladmandic/automatic/blob/m
     includes both **HiDream-O1-Image** *(base)* and **HiDream-O1-Image-Dev** *(distilled*)* variants  
     includes *T2I* and *I2I edit* capabilities and resolutions up to 2048px   
     *note*: use steps:50 for base and steps:28 for dev variants  
-    *note*: when using quantization, make sure that quantized matmul is disabled, otherwise quality degrades significantly  
   - [JoyAI Image Edit](https://huggingface.co/jdopensource/JoyAI-Image-Edit-Diffusers) image-editing model support  
     includes multimodal conditioning using *Qwen3-VL* with a dedicated *JoyImageEdit* diffusion transformer  
     *note* this is a large model at 50GB so use of agressive quantization is recommended  
@@ -55,7 +54,6 @@ For full details, see [ChangeLog](https://github.com/vladmandic/automatic/blob/m
   - custom **VAE** loader for all pipelines  
     *note*: vae still needs to be compatible with the model  
   - **CivitAI** downloaded thumbnails now include metadata  
-  - **Installer** support for `git+http` style references
 - **UI**
   - **Networks** using networks to load model or auto-download a reference model will now be reflected in the UI  
   - ability to manually reorient *input/output* panels
@@ -81,7 +79,6 @@ For full details, see [ChangeLog](https://github.com/vladmandic/automatic/blob/m
   - remove obsolete `lora` stepwise and functional code, thanks @awsr
   - interrupt model loading between components
   - patch `rich` for cleaner exception logging
-  - stricter `ruff` linting
 - **Fixes**
   - add missing `jquery` and `sparkline` js scripts
   - save handle already decoded images

cli/api-samplers.py

@@ -4,6 +4,7 @@
 get list of all samplers and details of current sampler
 """
 
+import sys
 import logging
 import urllib3
 import requests

modules/api/endpoints.py

@@ -361,7 +361,7 @@ def get_deletefile(file: str):
         return {"deleted": f"{file}"}
     except Exception as e:
         log.error(f'Delete: file="{file}" error: {e}')
-        raise HTTPException(status_code=500, detail=f"error deleting file {file}: {e!s}") from e
+        raise HTTPException(status_code=500, detail=f"error deleting file {file}: {str(e)}") from e
 
 def get_deleteimage(file: str):
     import os
@@ -383,7 +383,7 @@ def get_deleteimage(file: str):
         return {"deleted": f"{file}"}
     except Exception as e:
         log.error(f'Delete: file="{file}" error: {e}')
-        raise HTTPException(status_code=500, detail=f"error deleting file {file}: {e!s}") from e
+        raise HTTPException(status_code=500, detail=f"error deleting file {file}: {str(e)}") from e
 
 def get_pnginfo(file: str):
     """Extract generation parameters from a image file path. Returns raw info string and parsed parameters dict."""

modules/api/middleware.py

@@ -44,7 +44,7 @@ def setup_middleware(app: FastAPI, cmd_opts):
             client = req.scope.get('client', ('0:0.0.0', 0))[0]
             token = req.cookies.get("access-token") or req.cookies.get("access-token-unsecure")
             validate_request(client, endpoint)
-            if cmd_opts.api_log:
+            if (cmd_opts.api_log):
                 if not validate_log(client, endpoint):
                     return res
                 log.info('API user={user} code={code} {prot}/{ver} {method} {endpoint} {client} {duration}'.format( # pylint: disable=consider-using-f-string, logging-format-interpolation

modules/api/validate.py

@@ -9,7 +9,6 @@ request_cost = {
     "/run/predict": 0,
     "/sdapi/v1/browser/thumb": 0,
     "/sdapi/v1/network/thumb": 0,
-    "/sdapi/v1/gpu-smi": 0,
     "/sdapi/v1/txt2img": 5,
     "/sdapi/v1/img2img": 5,
     "/sdapi/v1/control": 5,

modules/caption/joycaption.py

@@ -65,7 +65,7 @@ def load(repo: str | None = None):
     if llava_model is None or opts.repo != repo:
         opts.repo = repo
         llava_model = None
-        log.info(f'Caption: type=vlm model="JoyCaption" {opts!s}')
+        log.info(f'Caption: type=vlm model="JoyCaption" {str(opts)}')
         processor = AutoProcessor.from_pretrained(repo, max_pixels=1024*1024, cache_dir=shared.opts.hfcache_dir)
         quant_args = model_quant.create_config(module='LLM')
         llava_model = LlavaForConditionalGeneration.from_pretrained(

modules/caption/moondream3.py

@@ -402,7 +402,7 @@ def predict(question: str, image: Image.Image, repo: str, model_name: str | None
     except Exception as e:
         from modules import errors
         errors.display(e, 'Moondream3')
-        return f"Error: {e!s}"
+        return f"Error: {str(e)}"
     finally:
         offload_aux('moondream3')
 

modules/caption/vqa.py

@@ -13,7 +13,7 @@ from modules import shared, devices, errors, model_quant, sd_models, sd_models_c
 from modules.sd_offload_aux import register_aux, deregister_aux, move_aux_to_gpu, offload_aux
 from modules.logger import log, console
 from modules.caption import vqa_detection
-from modules.caption.models_def import vlm_models, vlm_prefill, vlm_prompt_mapping, vlm_prompt_placeholders, vlm_prompts_common, vlm_prompts_florence, vlm_prompts_moondream, vlm_prompts_moondream2, vlm_prompts_promptgen, get_vlm_repo
+from modules.caption.models_def import vlm_models, vlm_system, vlm_default, vlm_prefill, vlm_prompts, vlm_prompt_mapping, vlm_prompt_placeholders, vlm_prompts_common, vlm_prompts_florence, vlm_prompts_moondream, vlm_prompts_moondream2, vlm_prompts_promptgen, get_vlm_repo
 
 # Debug logging - function-based to avoid circular import
 debug_enabled = os.environ.get('SD_CAPTION_DEBUG', None) is not None

modules/cfgzero/cogview4_pipeline.py

@@ -191,7 +191,7 @@ class CogView4CFGZeroPipeline(DiffusionPipeline, CogView4LoraLoaderMixin):
 
     def _get_glm_embeds(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         max_sequence_length: int = 1024,
         device: Optional[torch.device] = None,
         dtype: Optional[torch.dtype] = None,

modules/cfgzero/flux_pipeline.py

@@ -217,7 +217,7 @@ class FluxCFGZeroPipeline(
 
     def _get_t5_prompt_embeds(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         num_images_per_prompt: int = 1,
         max_sequence_length: int = 512,
         device: Optional[torch.device] = None,
@@ -535,7 +535,7 @@ class FluxCFGZeroPipeline(
 
     @staticmethod
     def _unpack_latents(latents, height, width, vae_scale_factor):
-        batch_size, _num_patches, channels = latents.shape
+        batch_size, num_patches, channels = latents.shape
 
         # VAE applies 8x compression on images but we must also account for packing which requires
         # latent height and width to be divisible by 2.
@@ -637,9 +637,9 @@ class FluxCFGZeroPipeline(
     @replace_example_docstring(EXAMPLE_DOC_STRING)
     def __call__(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         prompt_2: Optional[Union[str, List[str]]] = None,
-        negative_prompt: Union[str, List[str]] | None = None,
+        negative_prompt: Union[str, List[str]] = None,
         negative_prompt_2: Optional[Union[str, List[str]]] = None,
         true_cfg_scale: float = 1.0,
         height: Optional[int] = None,

modules/cfgzero/hidream_pipeline.py

@@ -211,7 +211,7 @@ class HiDreamImageCFGZeroPipeline(DiffusionPipeline, HiDreamImageLoraLoaderMixin
 
     def _get_t5_prompt_embeds(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         max_sequence_length: int = 128,
         device: Optional[torch.device] = None,
         dtype: Optional[torch.dtype] = None,
@@ -285,7 +285,7 @@ class HiDreamImageCFGZeroPipeline(DiffusionPipeline, HiDreamImageLoraLoaderMixin
 
     def _get_llama3_prompt_embeds(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         max_sequence_length: int = 128,
         device: Optional[torch.device] = None,
         dtype: Optional[torch.dtype] = None,
@@ -545,7 +545,7 @@ class HiDreamImageCFGZeroPipeline(DiffusionPipeline, HiDreamImageLoraLoaderMixin
     @torch.no_grad()
     def __call__(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         prompt_2: Optional[Union[str, List[str]]] = None,
         prompt_3: Optional[Union[str, List[str]]] = None,
         prompt_4: Optional[Union[str, List[str]]] = None,

modules/cfgzero/hunyuan_t2v_pipeline.py

@@ -317,7 +317,7 @@ class HunyuanVideoCFGZeroPipeline(DiffusionPipeline, HunyuanVideoLoraLoaderMixin
     def encode_prompt(
         self,
         prompt: Union[str, List[str]],
-        prompt_2: Union[str, List[str]] | None = None,
+        prompt_2: Union[str, List[str]] = None,
         prompt_template: Dict[str, Any] = DEFAULT_PROMPT_TEMPLATE,
         num_videos_per_prompt: int = 1,
         prompt_embeds: Optional[torch.Tensor] = None,
@@ -481,15 +481,15 @@ class HunyuanVideoCFGZeroPipeline(DiffusionPipeline, HunyuanVideoLoraLoaderMixin
     @replace_example_docstring(EXAMPLE_DOC_STRING)
     def __call__(
         self,
-        prompt: Union[str, List[str]] | None = None,
-        prompt_2: Union[str, List[str]] | None = None,
-        negative_prompt: Union[str, List[str]] | None = None,
-        negative_prompt_2: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
+        prompt_2: Union[str, List[str]] = None,
+        negative_prompt: Union[str, List[str]] = None,
+        negative_prompt_2: Union[str, List[str]] = None,
         height: int = 720,
         width: int = 1280,
         num_frames: int = 129,
         num_inference_steps: int = 50,
-        sigmas: List[float] | None = None,
+        sigmas: List[float] = None,
         true_cfg_scale: float = 1.0,
         guidance_scale: float = 6.0,
         num_videos_per_prompt: Optional[int] = 1,

modules/cfgzero/sd3_pipeline.py

@@ -246,7 +246,7 @@ class StableDiffusion3CFGZeroPipeline(DiffusionPipeline, SD3LoraLoaderMixin, Fro
 
     def _get_t5_prompt_embeds(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         num_images_per_prompt: int = 1,
         max_sequence_length: int = 256,
         device: Optional[torch.device] = None,
@@ -786,7 +786,7 @@ class StableDiffusion3CFGZeroPipeline(DiffusionPipeline, SD3LoraLoaderMixin, Fro
     @replace_example_docstring(EXAMPLE_DOC_STRING)
     def __call__(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         prompt_2: Optional[Union[str, List[str]]] = None,
         prompt_3: Optional[Union[str, List[str]]] = None,
         height: Optional[int] = None,
@@ -813,7 +813,7 @@ class StableDiffusion3CFGZeroPipeline(DiffusionPipeline, SD3LoraLoaderMixin, Fro
         callback_on_step_end: Optional[Callable[[int, int, Dict], None]] = None,
         callback_on_step_end_tensor_inputs: List[str] = ["latents"],
         max_sequence_length: int = 256,
-        skip_guidance_layers: List[int] | None = None,
+        skip_guidance_layers: List[int] = None,
         skip_layer_guidance_scale: float = 2.8,
         skip_layer_guidance_stop: float = 0.2,
         skip_layer_guidance_start: float = 0.01,

modules/cfgzero/wan_t2v_pipeline.py

@@ -153,7 +153,7 @@ class WanCFGZeroPipeline(DiffusionPipeline, WanLoraLoaderMixin):
 
     def _get_t5_prompt_embeds(
         self,
-        prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
         num_videos_per_prompt: int = 1,
         max_sequence_length: int = 226,
         device: Optional[torch.device] = None,
@@ -374,8 +374,8 @@ class WanCFGZeroPipeline(DiffusionPipeline, WanLoraLoaderMixin):
     @replace_example_docstring(EXAMPLE_DOC_STRING)
     def __call__(
         self,
-        prompt: Union[str, List[str]] | None = None,
-        negative_prompt: Union[str, List[str]] | None = None,
+        prompt: Union[str, List[str]] = None,
+        negative_prompt: Union[str, List[str]] = None,
         height: int = 480,
         width: int = 832,
         num_frames: int = 81,

modules/civitai/metadata_civitai.py

@@ -93,7 +93,7 @@ def civit_update_metadata(raw: bool = False):
                         model.latest_name = f.get('name', '')
                         if model.vername == model.latest:
                             model.status = 'Latest version'
-                        elif any(map(lambda v: v in model.latest_hashes, all_hashes)): # pylint: disable=cell-var-from-loop
+                        elif any(map(lambda v: v in model.latest_hashes, all_hashes)): # pylint: disable=cell-var-from-loop # noqa: C417
                             model.status = 'Update downloaded'
                         else:
                             model.status = 'Update available'

modules/control/proc/canny.py

@@ -7,7 +7,7 @@ from modules.control.util import HWC3, resize_image
 class CannyDetector:
     def __call__(self, input_image=None, low_threshold=100, high_threshold=200, detect_resolution=512, image_resolution=512, output_type=None, **kwargs):
         if "img" in kwargs:
-            warnings.warn("img is deprecated, please use `input_image=...` instead.", DeprecationWarning, stacklevel=2)
+            warnings.warn("img is deprecated, please use `input_image=...` instead.", DeprecationWarning)
             input_image = kwargs.pop("img")
         if input_image is None:
             raise ValueError("input_image must be defined.")

modules/control/proc/depth_anything/util/transform.py

@@ -1,4 +1,7 @@
+import random
+from PIL import Image, ImageOps, ImageFilter
 import torch
+from torchvision import transforms
 import torch.nn.functional as F
 
 import numpy as np

modules/control/proc/depth_pro/__init__.py

@@ -16,7 +16,7 @@ class DepthProDetector:
         self.processor = processor
 
     @classmethod
-    def from_pretrained(cls, pretrained_model_or_path: str = "apple/DepthPro-hf", cache_dir: str | None = None, local_files_only = False) -> "DepthProDetector":
+    def from_pretrained(cls, pretrained_model_or_path: str = "apple/DepthPro-hf", cache_dir: str = None, local_files_only = False) -> "DepthProDetector":
         from transformers import AutoImageProcessor, DepthProForDepthEstimation
 
         processor = AutoImageProcessor.from_pretrained(pretrained_model_or_path, cache_dir=cache_dir, local_files_only=local_files_only)

modules/control/proc/edge.py

@@ -33,7 +33,7 @@ class EdgeDetector:
         params.PFmode = pf
         ed.setParams(params)
         if "img" in kwargs:
-            warnings.warn("img is deprecated, please use `input_image=...` instead.", DeprecationWarning, stacklevel=2)
+            warnings.warn("img is deprecated, please use `input_image=...` instead.", DeprecationWarning)
             input_image = kwargs.pop("img")
         if input_image is None:
             raise ValueError("input_image must be defined.")

modules/control/proc/leres/leres/Resnext_torch.py

@@ -5,7 +5,7 @@ import torch.nn as nn
 try:
     from urllib import urlretrieve
 except ImportError:
-    pass
+    from urllib.request import urlretrieve
 
 __all__ = ['resnext101_32x8d']
 

modules/control/proc/leres/leres/multi_depth_model_woauxi.py

@@ -1,3 +1,4 @@
+import torch
 import torch.nn as nn
 
 from . import network_auxi as network

modules/control/proc/leres/leres/network_auxi.py

@@ -384,7 +384,7 @@ class SenceUnderstand(nn.Module):
         self.initial_params()
 
     def forward(self, x):
-        n, _c, h, w = x.size()
+        n, c, h, w = x.size()
         x = self.conv1(x)
         x = self.pool(x)
         x = x.view(n, -1)

modules/control/proc/leres/pix2pix/options/base_options.py

@@ -1,8 +1,8 @@
 import argparse
 import os
-from ...pix2pix.util import util  # noqa: TID252
+from ...pix2pix.util import util
 # import torch
-from ...pix2pix import models  # noqa: TID252
+from ...pix2pix import models
 # import pix2pix.data
 import numpy as np
 

modules/control/proc/marigold/marigold_pipeline.py

@@ -113,7 +113,7 @@ class MarigoldPipeline(DiffusionPipeline):
         batch_size: int = 0,
         color_map: str = "Spectral",
         show_progress_bar: bool = True,
-        ensemble_kwargs: Dict | None = None,
+        ensemble_kwargs: Dict = None,
     ) -> MarigoldDepthOutput:
         """
         Function invoked when calling the pipeline.

modules/control/proc/marigold/util/ensemble.py

@@ -43,7 +43,7 @@ def ensemble_depths(
     max_iter: int = 2,
     tol: float = 1e-3,
     reduction: str = "median",
-    max_res: int | None = None,
+    max_res: int = None,
 ):
     """
     To ensemble multiple affine-invariant depth images (up to scale and shift),

modules/control/proc/marigold/util/seed_all.py

@@ -28,6 +28,6 @@ def seed_all(seed: int = 0):
     Set random seeds of all components.
     """
     random.seed(seed)
-    np.random.seed(seed)
+    np.random.seed(seed) # noqa
     torch.manual_seed(seed)
     torch.cuda.manual_seed_all(seed)

modules/control/proc/mediapipe_face.py

@@ -16,6 +16,7 @@ def check_dependencies():
         if not installed(pkg[1], quiet=True):
             install(pkg[0], pkg[1], ignore=False)
     try:
+        import mediapipe as mp # pylint: disable=unused-import
         checked_ok = True
         return True
     except Exception as e:

modules/control/proc/midas/api.py

@@ -2,6 +2,7 @@
 
 import cv2
 import os
+import torch
 import torch.nn as nn
 from torchvision.transforms import Compose
 

modules/control/proc/midas/midas/blocks.py

@@ -5,6 +5,7 @@ from .vit import (
     _make_pretrained_vitb_rn50_384,
     _make_pretrained_vitl16_384,
     _make_pretrained_vitb16_384,
+    forward_vit,
 )
 
 def _make_encoder(backbone, features, use_pretrained, groups=1, expand=False, exportable=True, hooks=None, use_vit_only=False, use_readout="ignore",):

modules/control/proc/midas/midas/dpt_depth.py

@@ -1,8 +1,10 @@
 import torch
 import torch.nn as nn
+import torch.nn.functional as F
 
 from .base_model import BaseModel
 from .blocks import (
+    FeatureFusionBlock,
     FeatureFusionBlock_custom,
     Interpolate,
     _make_encoder,

modules/control/proc/midas/midas/midas_net_custom.py

@@ -6,7 +6,7 @@ import torch
 import torch.nn as nn
 
 from .base_model import BaseModel
-from .blocks import FeatureFusionBlock_custom, Interpolate, _make_encoder
+from .blocks import FeatureFusionBlock, FeatureFusionBlock_custom, Interpolate, _make_encoder
 
 
 class MidasNet_small(BaseModel):

modules/control/proc/midas/midas/vit.py

@@ -54,7 +54,7 @@ class Transpose(nn.Module):
 
 
 def forward_vit(pretrained, x):
-    _b, _c, h, w = x.shape
+    b, c, h, w = x.shape
 
     pretrained.model.forward_flex(x)
 
@@ -115,7 +115,7 @@ def _resize_pos_embed(self, posemb, gs_h, gs_w):
 
 
 def forward_flex(self, x):
-    _b, _c, h, w = x.shape
+    b, c, h, w = x.shape
 
     pos_embed = self._resize_pos_embed(
         self.pos_embed, h // self.patch_size[1], w // self.patch_size[0]

modules/control/proc/midas/utils.py

@@ -75,7 +75,7 @@ def write_pfm(path, image, scale=1):
         if len(image.shape) == 3 and image.shape[2] == 3:  # color image
             color = True
         elif (
-            len(image.shape) == 2 or (len(image.shape) == 3 and image.shape[2] == 1)
+            len(image.shape) == 2 or len(image.shape) == 3 and image.shape[2] == 1
         ):  # greyscale
             color = False
         else:
@@ -86,7 +86,7 @@ def write_pfm(path, image, scale=1):
 
         endian = image.dtype.byteorder
 
-        if endian == "<" or (endian == "=" and sys.byteorder == "little"):
+        if endian == "<" or endian == "=" and sys.byteorder == "little":
             scale = -scale
 
         file.write("%f\n".encode() % scale)

modules/control/proc/mlsd/models/mbv2_mlsd_large.py

@@ -1,3 +1,5 @@
+import os
+import sys
 import torch
 import torch.nn as nn
 import torch.utils.model_zoo as model_zoo

modules/control/proc/mlsd/models/mbv2_mlsd_tiny.py

@@ -1,3 +1,5 @@
+import os
+import sys
 import torch
 import torch.nn as nn
 import torch.utils.model_zoo as model_zoo

modules/control/proc/mlsd/utils.py

@@ -9,6 +9,7 @@ Copyright 2021-present NAVER Corp.
 Apache License v2.0
 '''
 
+import os
 import numpy as np
 import cv2
 import torch
@@ -21,7 +22,7 @@ def deccode_output_score_and_ptss(tpMap, topk_n = 200, ksize = 5):
     center: tpMap[1, 0, :, :]
     displacement: tpMap[1, 1:5, :, :]
     '''
-    b, _c, _h, w = tpMap.shape
+    b, c, h, w = tpMap.shape
     assert  b==1, 'only support bsize==1'
     displacement = tpMap[:, 1:5, :, :][0]
     center = tpMap[:, 0, :, :]
@@ -470,9 +471,9 @@ def pred_squares(image,
                 square[end_idx]
 
                 # check whether outside or inside
-                _start_position, start_min, start_cover_param, start_peri_param = check_outside_inside(start_segments,
+                start_position, start_min, start_cover_param, start_peri_param = check_outside_inside(start_segments,
                                                                                                       connect_idx)
-                _end_position, end_min, end_cover_param, end_peri_param = check_outside_inside(end_segments, connect_idx)
+                end_position, end_min, end_cover_param, end_peri_param = check_outside_inside(end_segments, connect_idx)
 
                 cover += dist_segments[connect_idx] + start_cover_param * start_min + end_cover_param * end_min
                 perimeter += dist_segments[connect_idx] + start_peri_param * start_min + end_peri_param * end_min

modules/control/proc/openpose/__init__.py

@@ -194,14 +194,14 @@ class OpenposeDetector:
     def __call__(self, input_image, detect_resolution=512, image_resolution=512, include_body=True, include_hand=False, include_face=False, hand_and_face=None, output_type="pil", **kwargs):
         self.to(devices.device)
         if hand_and_face is not None:
-            warnings.warn("hand_and_face is deprecated. Use include_hand and include_face instead.", DeprecationWarning, stacklevel=2)
+            warnings.warn("hand_and_face is deprecated. Use include_hand and include_face instead.", DeprecationWarning)
             include_hand = hand_and_face
             include_face = hand_and_face
         if "return_pil" in kwargs:
-            warnings.warn("return_pil is deprecated. Use output_type instead.", DeprecationWarning, stacklevel=2)
+            warnings.warn("return_pil is deprecated. Use output_type instead.", DeprecationWarning)
             output_type = "pil" if kwargs["return_pil"] else "np"
         if type(output_type) is bool:
-            warnings.warn("Passing `True` or `False` to `output_type` is deprecated and will raise an error in future versions", stacklevel=2)
+            warnings.warn("Passing `True` or `False` to `output_type` is deprecated and will raise an error in future versions")
             if output_type:
                 output_type = "pil"
         if not isinstance(input_image, np.ndarray):

modules/control/proc/openpose/face.py

@@ -328,7 +328,7 @@ class Face(object):
 
     def __call__(self, face_img):
         device = next(iter(self.model.parameters())).device
-        H, W, _C = face_img.shape
+        H, W, C = face_img.shape
 
         w_size = 384
         x_data = torch.from_numpy(util.smart_resize(face_img, (w_size, w_size))).permute([2, 0, 1]) / 256.0 - 0.5

modules/control/proc/openpose/hand.py

@@ -32,7 +32,7 @@ class Hand(object):
         wsize = 128
         heatmap_avg = np.zeros((wsize, wsize, 22))
 
-        Hr, Wr, _Cr = oriImgRaw.shape
+        Hr, Wr, Cr = oriImgRaw.shape
 
         oriImg = cv2.GaussianBlur(oriImgRaw, (0, 0), 0.8)
 

modules/control/proc/segment_anything/__init__.py

@@ -53,7 +53,7 @@ class SamDetector:
 
     def __call__(self, input_image: Union[np.ndarray, Image.Image]=None, detect_resolution=512, image_resolution=512, output_type="pil", **kwargs) -> Image.Image:
         if "image" in kwargs:
-            warnings.warn("image is deprecated, please use `input_image=...` instead.", DeprecationWarning, stacklevel=2)
+            warnings.warn("image is deprecated, please use `input_image=...` instead.", DeprecationWarning)
             input_image = kwargs.pop("image")
         if input_image is None:
             raise ValueError("input_image must be defined.")

modules/control/proc/segment_anything/modeling/sam.py

@@ -25,8 +25,8 @@ class Sam(nn.Module):
         image_encoder: Union[ImageEncoderViT, TinyViT],
         prompt_encoder: PromptEncoder,
         mask_decoder: MaskDecoder,
-      pixel_mean: List[float] | None = None,
-      pixel_std: List[float] | None = None,
+        pixel_mean: List[float] = None,
+        pixel_std: List[float] = None,
     ) -> None:
         """
         SAM predicts object masks from an image and input prompts.

modules/control/proc/segment_anything/modeling/transformer.py

@@ -79,7 +79,7 @@ class TwoWayTransformer(nn.Module):
           torch.Tensor: the processed image_embedding
         """
         # BxCxHxW -> BxHWxC == B x N_image_tokens x C
-        _bs, _c, _h, _w = image_embedding.shape
+        bs, c, h, w = image_embedding.shape
         image_embedding = image_embedding.flatten(2).permute(0, 2, 1)
         image_pe = image_pe.flatten(2).permute(0, 2, 1)
 

modules/control/proc/segment_anything/utils/onnx.py

@@ -10,7 +10,7 @@ from torch.nn import functional as F
 
 from typing import Tuple
 
-from ..modeling import Sam  # noqa: TID252
+from ..modeling import Sam
 from .amg import calculate_stability_score
 
 

modules/control/proc/shuffle.py

@@ -10,10 +10,10 @@ from modules.control.util import HWC3, img2mask, make_noise_disk, resize_image
 class ContentShuffleDetector:
     def __call__(self, input_image, h=None, w=None, f=None, detect_resolution=512, image_resolution=512, output_type="pil", **kwargs):
         if "return_pil" in kwargs:
-            warnings.warn("return_pil is deprecated. Use output_type instead.", DeprecationWarning, stacklevel=2)
+            warnings.warn("return_pil is deprecated. Use output_type instead.", DeprecationWarning)
             output_type = "pil" if kwargs["return_pil"] else "np"
         if type(output_type) is bool:
-            warnings.warn("Passing `True` or `False` to `output_type` is deprecated and will raise an error in future versions", stacklevel=2)
+            warnings.warn("Passing `True` or `False` to `output_type` is deprecated and will raise an error in future versions")
             if output_type:
                 output_type = "pil"
 
@@ -49,7 +49,7 @@ class ContentShuffleDetector:
 class ColorShuffleDetector:
     def __call__(self, img):
         H, W, C = img.shape
-        F = np.random.randint(64, 384)
+        F = np.random.randint(64, 384) # noqa
         A = make_noise_disk(H, W, 3, F)
         B = make_noise_disk(H, W, 3, F)
         C = (A + B) / 2.0
@@ -82,11 +82,11 @@ class DownSampleDetector:
     def __call__(self, img, level=3, k=16.0):
         h = img.astype(np.float32)
         for _ in range(level):
-            h += np.random.normal(loc=0.0, scale=k, size=h.shape)
+            h += np.random.normal(loc=0.0, scale=k, size=h.shape) # noqa
             h = cv2.pyrDown(h)
         for _ in range(level):
             h = cv2.pyrUp(h)
-            h += np.random.normal(loc=0.0, scale=k, size=h.shape)
+            h += np.random.normal(loc=0.0, scale=k, size=h.shape) # noqa
         return h.clip(0, 255).astype(np.uint8)
 
 

modules/control/proc/zoe/zoedepth/models/base_models/midas_repo/midas/backbones/beit.py

@@ -66,7 +66,7 @@ def attention_forward(self, x, resolution, shared_rel_pos_bias: Optional[torch.T
     """
     Modification of timm.models.beit.py: Attention.forward to support arbitrary window sizes.
     """
-    B, N, _C = x.shape
+    B, N, C = x.shape
 
     qkv_bias = torch.cat((self.q_bias, self.k_bias, self.v_bias)) if self.q_bias is not None else None
     qkv = F.linear(input=x, weight=self.qkv.weight, bias=qkv_bias)

modules/control/proc/zoe/zoedepth/models/base_models/midas_repo/midas/backbones/utils.py

@@ -81,7 +81,7 @@ def forward_default(pretrained, x, function_name="forward_features"):
 
 
 def forward_adapted_unflatten(pretrained, x, function_name="forward_features"):
-    _b, _c, h, w = x.shape
+    b, c, h, w = x.shape
 
     exec(f"glob = pretrained.model.{function_name}(x)")
 

modules/control/proc/zoe/zoedepth/models/base_models/midas_repo/midas/backbones/vit.py

@@ -31,7 +31,7 @@ def _resize_pos_embed(self, posemb, gs_h, gs_w):
 
 
 def forward_flex(self, x):
-    _b, _c, h, w = x.shape
+    b, c, h, w = x.shape
 
     pos_embed = self._resize_pos_embed(
         self.pos_embed, h // self.patch_size[1], w // self.patch_size[0]

modules/control/proc/zoe/zoedepth/models/base_models/midas_repo/midas/blocks.py

@@ -5,6 +5,10 @@ from .backbones.beit import (
     _make_pretrained_beitl16_512,
     _make_pretrained_beitl16_384,
     _make_pretrained_beitb16_384,
+    forward_beit,
+)
+from .backbones.swin_common import (
+    forward_swin,
 )
 from .backbones.swin2 import (
     _make_pretrained_swin2l24_384,
@@ -16,11 +20,13 @@ from .backbones.swin import (
 )
 from .backbones.levit import (
     _make_pretrained_levit_384,
+    forward_levit,
 )
 from .backbones.vit import (
     _make_pretrained_vitb_rn50_384,
     _make_pretrained_vitl16_384,
     _make_pretrained_vitb16_384,
+    forward_vit,
 )
 
 def _make_encoder(backbone, features, use_pretrained, groups=1, expand=False, exportable=True, hooks=None,

modules/control/proc/zoe/zoedepth/models/base_models/midas_repo/midas/midas_net_custom.py

@@ -6,7 +6,7 @@ import torch
 import torch.nn as nn
 
 from .base_model import BaseModel
-from .blocks import FeatureFusionBlock_custom, Interpolate, _make_encoder
+from .blocks import FeatureFusionBlock, FeatureFusionBlock_custom, Interpolate, _make_encoder
 
 
 class MidasNet_small(BaseModel):

modules/control/proc/zoe/zoedepth/models/layers/attractor.py

@@ -100,7 +100,7 @@ class AttractorLayer(nn.Module):
         A = self._net(x)
         eps = 1e-3
         A = A + eps
-        n, _c, h, w = A.shape
+        n, c, h, w = A.shape
         A = A.view(n, self.n_attractors, 2, h, w)
         A_normed = A / A.sum(dim=2, keepdim=True)  # n, a, 2, h, w
         A_normed = A[:, :, 0, ...]  # n, na, h, w
@@ -177,7 +177,7 @@ class AttractorLayerUnnormed(nn.Module):
             x = x + prev_b_embedding
 
         A = self._net(x)
-        _n, _c, h, w = A.shape
+        n, c, h, w = A.shape
 
         b_prev = nn.functional.interpolate(
             b_prev, (h, w), mode='bilinear', align_corners=True)

modules/control/proc/zoe/zoedepth/models/layers/localbins_layers.py

@@ -146,7 +146,7 @@ class LinearSplitter(nn.Module):
         S = self._net(x)
         eps = 1e-3
         S = S + eps
-        n, _c, h, w = S.shape
+        n, c, h, w = S.shape
         S = S.view(n, self.prev_nbins, self.split_factor, h, w)
         S_normed = S / S.sum(dim=2, keepdim=True)  # fractional splits
 

modules/control/proc/zoe/zoedepth/models/zoedepth/zoedepth_v1.py

@@ -26,13 +26,13 @@ import itertools
 
 import torch
 import torch.nn as nn
-from ..depth_model import DepthModel  # noqa: TID252
-from ..base_models.midas import MidasCore  # noqa: TID252
-from ..layers.attractor import AttractorLayer, AttractorLayerUnnormed  # noqa: TID252
-from ..layers.dist_layers import ConditionalLogBinomial  # noqa: TID252
-from ..layers.localbins_layers import (Projector, SeedBinRegressor,  # noqa: TID252
+from ..depth_model import DepthModel
+from ..base_models.midas import MidasCore
+from ..layers.attractor import AttractorLayer, AttractorLayerUnnormed
+from ..layers.dist_layers import ConditionalLogBinomial
+from ..layers.localbins_layers import (Projector, SeedBinRegressor,
                                             SeedBinRegressorUnnormed)
-from ..model_io import load_state_from_resource  # noqa: TID252
+from ..model_io import load_state_from_resource
 
 
 class ZoeDepth(DepthModel):
@@ -139,7 +139,7 @@ class ZoeDepth(DepthModel):
                 - probs (torch.Tensor): Output probability distribution of shape (B, n_bins, H, W). Present only if return_probs is True
 
         """
-        b, _c, h, w = x.shape
+        b, c, h, w = x.shape
         # print("input shape ", x.shape)
         self.orig_input_width = w
         self.orig_input_height = h

modules/control/proc/zoe/zoedepth/models/zoedepth_nk/zoedepth_nk_v1.py

@@ -27,14 +27,14 @@ import itertools
 import torch
 import torch.nn as nn
 
-from ..depth_model import DepthModel  # noqa: TID252
-from ..base_models.midas import MidasCore  # noqa: TID252
-from ..layers.attractor import AttractorLayer, AttractorLayerUnnormed  # noqa: TID252
-from ..layers.dist_layers import ConditionalLogBinomial  # noqa: TID252
-from ..layers.localbins_layers import (Projector, SeedBinRegressor,  # noqa: TID252
+from ..depth_model import DepthModel
+from ..base_models.midas import MidasCore
+from ..layers.attractor import AttractorLayer, AttractorLayerUnnormed
+from ..layers.dist_layers import ConditionalLogBinomial
+from ..layers.localbins_layers import (Projector, SeedBinRegressor,
                                             SeedBinRegressorUnnormed)
-from ..layers.patch_transformer import PatchTransformerEncoder  # noqa: TID252
-from ..model_io import load_state_from_resource  # noqa: TID252
+from ..layers.patch_transformer import PatchTransformerEncoder
+from ..model_io import load_state_from_resource
 
 class ZoeDepthNK(DepthModel):
     def __init__(self, core,  bin_conf, bin_centers_type="softplus", bin_embedding_dim=128,
@@ -173,10 +173,10 @@ class ZoeDepthNK(DepthModel):
                 - "bin_centers": Bin centers of shape (B, N, H, W). Present only if return_final_centers is True
                 - "probs": Bin probabilities of shape (B, N, H, W). Present only if return_probs is True
         """
-        b, _c, h, w = x.shape
+        b, c, h, w = x.shape
         self.orig_input_width = w
         self.orig_input_height = h
-        _rel_depth, out = self.core(x, denorm=denorm, return_rel_depth=True)
+        rel_depth, out = self.core(x, denorm=denorm, return_rel_depth=True)
 
         outconv_activation = out[0]
         btlnck = out[1]

modules/control/units/controlnet.py

@@ -5,7 +5,7 @@ from typing import Union
 from diffusers import StableDiffusionPipeline, StableDiffusionXLPipeline, FluxPipeline, StableDiffusion3Pipeline, ControlNetModel
 from modules.control.units import detect
 from modules.shared import log, opts, cmd_opts, state, listdir
-from modules import errors, sd_models, devices
+from modules import errors, sd_models, devices, model_quant
 from modules.processing import StableDiffusionProcessingControl
 
 
@@ -163,7 +163,7 @@ def find_models():
 find_models()
 
 
-def api_list_models(model_type: str | None = None):
+def api_list_models(model_type: str = None):
     import modules.shared
     model_type = model_type or modules.shared.sd_model_type
     model_list = []
@@ -215,7 +215,7 @@ def list_models(refresh=False):
 
 
 class ControlNet():
-    def __init__(self, model_id: str | None = None, device = None, dtype = None, load_config = None):
+    def __init__(self, model_id: str = None, device = None, dtype = None, load_config = None):
         self.model: ControlNetModel = None
         self.model_id: str = model_id
         self.device = device
@@ -311,7 +311,7 @@ class ControlNet():
             self.load_config['original_config_file '] = config_path
         self.model = cls.from_single_file(model_path, config=config, **self.load_config)
 
-    def load(self, model_id: str | None = None, force: bool = False) -> str:
+    def load(self, model_id: str = None, force: bool = False) -> str:
         with load_lock:
             try:
                 t0 = time.time()

modules/control/units/lite.py

@@ -63,7 +63,7 @@ def list_models(refresh=False):
 
 
 class ControlLLLite():
-    def __init__(self, model_id: str | None = None, device = None, dtype = None, load_config = None):
+    def __init__(self, model_id: str = None, device = None, dtype = None, load_config = None):
         self.model: ControlNetLLLite = None
         self.model_id: str = model_id
         self.device = device
@@ -83,7 +83,7 @@ class ControlLLLite():
         self.model = None
         self.model_id = None
 
-    def load(self, model_id: str | None = None, force: bool = True) -> str:
+    def load(self, model_id: str = None, force: bool = True) -> str:
         with load_lock:
             try:
                 t0 = time.time()

modules/control/units/t2iadapter.py

@@ -71,7 +71,7 @@ class AdapterModel(T2IAdapter):
 
 
 class Adapter():
-    def __init__(self, model_id: str | None = None, device = None, dtype = None, load_config = None):
+    def __init__(self, model_id: str = None, device = None, dtype = None, load_config = None):
         self.model: AdapterModel = None
         self.model_id: str = model_id
         self.device = device
@@ -91,7 +91,7 @@ class Adapter():
         self.model = None
         self.model_id = None
 
-    def load(self, model_id: str | None = None, force: bool = True) -> str:
+    def load(self, model_id: str = None, force: bool = True) -> str:
         with load_lock:
             try:
                 t0 = time.time()

modules/control/units/xs.py

@@ -59,7 +59,7 @@ def list_models(refresh=False):
 
 
 class ControlNetXS():
-    def __init__(self, model_id: str | None = None, device = None, dtype = None, load_config = None):
+    def __init__(self, model_id: str = None, device = None, dtype = None, load_config = None):
         self.model: ControlNetXSModel = None
         self.model_id: str = model_id
         self.device = device
@@ -79,7 +79,7 @@ class ControlNetXS():
         self.model = None
         self.model_id = None
 
-    def load(self, model_id: str | None = None, time_embedding_mix: float = 0.0, force: bool = True) -> str:
+    def load(self, model_id: str = None, time_embedding_mix: float = 0.0, force: bool = True) -> str:
         with load_lock:
             try:
                 t0 = time.time()

modules/control/units/xs_model.py

@@ -64,9 +64,9 @@ class ControlNetXSOutput(BaseOutput):
 class ControlNetConditioningEmbedding(nn.Module):
     """
     Quoting from https://arxiv.org/abs/2302.05543: "Stable Diffusion uses a pre-processing method similar to VQ-GAN
-    [11] to convert the entire dataset of 512 x 512 images into smaller 64 x 64 “latent images” for stabilized
-    training. This requires ControlNets to convert image-based conditions to 64 x 64 feature space to match the
-    convolution size. We use a tiny network E(·) of four convolution layers with 4 x 4 kernels and 2 x 2 strides
+    [11] to convert the entire dataset of 512 × 512 images into smaller 64 × 64 “latent images” for stabilized
+    training. This requires ControlNets to convert image-based conditions to 64 × 64 feature space to match the
+    convolution size. We use a tiny network E(·) of four convolution layers with 4 × 4 kernels and 2 × 2 strides
     (activated by ReLU, channels are 16, 32, 64, 128, initialized with Gaussian weights, trained jointly with the full
     model) to encode image-space conditions ... into feature maps ..."
     """
@@ -657,7 +657,7 @@ class ControlNetXSModel(ModelMixin, ConfigMixin):
             if base_model.config.addition_embed_type == "text":
                 aug_emb = base_model.add_embedding(encoder_hidden_states)
             elif base_model.config.addition_embed_type == "text_image":
-                raise NotImplementedError
+                raise NotImplementedError()
             elif base_model.config.addition_embed_type == "text_time":
                 # SDXL - style
                 if "text_embeds" not in added_cond_kwargs:
@@ -676,9 +676,9 @@ class ControlNetXSModel(ModelMixin, ConfigMixin):
                 add_embeds = add_embeds.to(temb.dtype)
                 aug_emb = base_model.add_embedding(add_embeds)
             elif base_model.config.addition_embed_type == "image":
-                raise NotImplementedError
+                raise NotImplementedError()
             elif base_model.config.addition_embed_type == "image_hint":
-                raise NotImplementedError
+                raise NotImplementedError()
 
         temb = temb + aug_emb if aug_emb is not None else temb
 

modules/control/units/xs_pipe.py

@@ -518,8 +518,8 @@ class StableDiffusionXLControlNetXSPipeline(
         )
         if (
             isinstance(self.controlnet, ControlNetXSModel)
-            or (is_compiled
-            and isinstance(self.controlnet._orig_mod, ControlNetXSModel))
+            or is_compiled
+            and isinstance(self.controlnet._orig_mod, ControlNetXSModel)
         ):
             self.check_image(image, prompt, prompt_embeds)
         else:
@@ -528,8 +528,8 @@ class StableDiffusionXLControlNetXSPipeline(
         # Check `controlnet_conditioning_scale`
         if (
             isinstance(self.controlnet, ControlNetXSModel)
-            or (is_compiled
-            and isinstance(self.controlnet._orig_mod, ControlNetXSModel))
+            or is_compiled
+            and isinstance(self.controlnet._orig_mod, ControlNetXSModel)
         ):
             if not isinstance(controlnet_conditioning_scale, float):
                 raise TypeError("For single controlnet: `controlnet_conditioning_scale` must be type `float`.")
@@ -1521,8 +1521,8 @@ class StableDiffusionControlNetXSPipeline(
         )
         if (
             isinstance(self.controlnet, ControlNetXSModel)
-            or (is_compiled
-            and isinstance(self.controlnet._orig_mod, ControlNetXSModel))
+            or is_compiled
+            and isinstance(self.controlnet._orig_mod, ControlNetXSModel)
         ):
             self.check_image(image, prompt, prompt_embeds)
         else:
@@ -1531,8 +1531,8 @@ class StableDiffusionControlNetXSPipeline(
         # Check `controlnet_conditioning_scale`
         if (
             isinstance(self.controlnet, ControlNetXSModel)
-            or (is_compiled
-            and isinstance(self.controlnet._orig_mod, ControlNetXSModel))
+            or is_compiled
+            and isinstance(self.controlnet._orig_mod, ControlNetXSModel)
         ):
             if not isinstance(controlnet_conditioning_scale, float):
                 raise TypeError("For single controlnet: `controlnet_conditioning_scale` must be type `float`.")

modules/errorlimiter.py

@@ -2,8 +2,10 @@ from __future__ import annotations
 
 from contextlib import contextmanager
 from threading import Lock
-from typing import ClassVar
+from typing import TYPE_CHECKING, ClassVar
 
+if TYPE_CHECKING:
+    from collections.abc import Iterable
 
 _instance_id = 0
 _lock = Lock()

modules/framepack/pipeline/hunyuan_video_packed.py

@@ -832,7 +832,7 @@ class HunyuanVideoTransformer3DModelPacked(ModelMixin, ConfigMixin, PeftAdapterM
             clean_latents_4x=None, clean_latent_4x_indices=None
     ):
         hidden_states = self.gradient_checkpointing_method(self.x_embedder.proj, latents)
-        B, _C, T, H, W = hidden_states.shape
+        B, C, T, H, W = hidden_states.shape
 
         if latent_indices is None:
             latent_indices = torch.arange(0, T).unsqueeze(0).expand(B, -1)
@@ -897,7 +897,7 @@ class HunyuanVideoTransformer3DModelPacked(ModelMixin, ConfigMixin, PeftAdapterM
         if attention_kwargs is None:
             attention_kwargs = {}
 
-        batch_size, _num_channels, num_frames, height, width = hidden_states.shape
+        batch_size, num_channels, num_frames, height, width = hidden_states.shape
         p, p_t = self.config['patch_size'], self.config['patch_size_t']
         post_patch_num_frames = num_frames // p_t
         post_patch_height = height // p

modules/framepack/pipeline/uni_pc_fm.py

@@ -18,7 +18,7 @@ torch_linalg_solve = None
 
 
 def test_solver():
-    from modules import devices
+    from modules import devices, shared
     try:
         a = torch.randn(50, 50).to(device=devices.device, dtype=torch.float32)
         b = torch.randn(50, 2).to(device=devices.device, dtype=torch.float32)

modules/hidiffusion/hidiffusion.py

@@ -107,7 +107,7 @@ def make_diffusers_transformer_block(block_class: Type[torch.nn.Module]) -> Type
             encoder_hidden_states: Optional[torch.FloatTensor] = None,
             encoder_attention_mask: Optional[torch.FloatTensor] = None,
             timestep: Optional[torch.LongTensor] = None,
-            cross_attention_kwargs: Dict[str, Any] | None = None,
+            cross_attention_kwargs: Dict[str, Any] = None,
             class_labels: Optional[torch.LongTensor] = None,
             added_cond_kwargs: Optional[Dict[str, torch.Tensor]] = None,
         ) -> torch.FloatTensor:

modules/hidiffusion/hidiffusion_controlnet.py

@@ -272,7 +272,7 @@ def make_diffusers_sdxl_contrtolnet_ppl(block_class):
         @torch.no_grad()
         def __call__(
             self,
-            prompt: Union[str, List[str]] | None = None,
+            prompt: Union[str, List[str]] = None,
             prompt_2: Optional[Union[str, List[str]]] = None,
             image: PipelineImageInput = None,
             control_image: PipelineImageInput = None,
@@ -298,9 +298,9 @@ def make_diffusers_sdxl_contrtolnet_ppl(block_class):
             guess_mode: bool = False,
             control_guidance_start: Union[float, List[float]] = 0.0,
             control_guidance_end: Union[float, List[float]] = 1.0,
-            original_size: Tuple[int, int] | None = None,
+            original_size: Tuple[int, int] = None,
             crops_coords_top_left: Tuple[int, int] = (0, 0),
-            target_size: Tuple[int, int] | None = None,
+            target_size: Tuple[int, int] = None,
             negative_original_size: Optional[Tuple[int, int]] = None,
             negative_crops_coords_top_left: Tuple[int, int] = (0, 0),
             negative_target_size: Optional[Tuple[int, int]] = None,

modules/images.py

@@ -6,21 +6,21 @@ from modules.image.grid import Grid, image_grid, check_grid_size, get_grid_size,
 from modules.image.util import draw_text, flatten
 
 __all__ = [
-    'FilenameGenerator',
-    'Grid',
     'check_grid_size',
     'combine_grid',
     'draw_grid_annotations',
     'draw_prompt_matrix',
-    'draw_text',
-    'flatten',
-    'get_font',
+    'FilenameGenerator',
     'get_grid_size',
-    'get_next_sequence_number',
+    'Grid',
     'image_data',
     'image_grid',
     'read_info_from_image',
     'resize_image',
     'sanitize_filename_part',
     'save_image',
+    'get_font',
+    'get_next_sequence_number',
+    'draw_text',
+    'flatten',
 ]

modules/intel/ipex/diffusers.py

@@ -50,7 +50,7 @@ def hidream_rope(pos: torch.Tensor, dim: int, theta: int) -> torch.Tensor:
     scale = torch.arange(0, dim, 2, dtype=torch.float64, device=pos.device) / dim
     omega = 1.0 / (theta**scale)
 
-    batch_size, _seq_length = pos.shape
+    batch_size, seq_length = pos.shape
     out = torch.einsum("...n,d->...nd", pos, omega)
     cos_out = torch.cos(out)
     sin_out = torch.sin(out)

modules/intel/ipex/hijacks.py

@@ -52,7 +52,7 @@ def autocast_init(self, device_type=None, dtype=None, enabled=True, cache_enable
 
 original_grad_scaler_init = torch.amp.grad_scaler.GradScaler.__init__
 @wraps(torch.amp.grad_scaler.GradScaler.__init__)
-def GradScaler_init(self, device: str | None = None, init_scale: float = 2.0**16, growth_factor: float = 2.0, backoff_factor: float = 0.5, growth_interval: int = 2000, enabled: bool = True):
+def GradScaler_init(self, device: str = None, init_scale: float = 2.0**16, growth_factor: float = 2.0, backoff_factor: float = 0.5, growth_interval: int = 2000, enabled: bool = True):
     if device is None or check_cuda(device):
         return original_grad_scaler_init(self, device=return_xpu(device), init_scale=init_scale, growth_factor=growth_factor, backoff_factor=backoff_factor, growth_interval=growth_interval, enabled=enabled)
     else:

modules/loader.py

@@ -75,7 +75,7 @@ except Exception:
     pass
 
 try:
-    pass # pylint: disable=unused-import,ungrouped-imports
+    import torch.distributed.distributed_c10d as _c10d # pylint: disable=unused-import,ungrouped-imports
 except Exception:
     log.warning('Loader: torch is not built with distributed support')
 
@@ -96,6 +96,7 @@ warnings.filterwarnings(action="ignore", category=UserWarning, module="torchvisi
 torchvision = None
 try:
     import torchvision # pylint: disable=W0611,C0411
+    import pytorch_lightning # pytorch_lightning should be imported after torch, but it re-enables warnings on import so import once to disable them # pylint: disable=W0611,C0411
 except Exception as e:
     report(f'torchvision=={torchvision.__version__ if torchvision is not None else None}', e)
 
@@ -126,6 +127,7 @@ if ".dev" in torch.__version__ or "+git" in torch.__version__:
 timer.startup.record("torch")
 
 try:
+    import bitsandbytes # pylint: disable=unused-import
     _bnb = True
 except Exception:
     _bnb = False
@@ -209,9 +211,10 @@ except Exception as e:
     sys.exit(1)
 
 try:
-    pass # pylint: disable=W0611,C0411
+    import pillow_jxl # pylint: disable=W0611,C0411
 except Exception:
     pass
+from PIL import Image # pylint: disable=W0611,C0411
 timer.startup.record("pillow")
 
 

modules/lora/lora_load.py

@@ -354,7 +354,7 @@ def network_load(names, te_multipliers=None, unet_multipliers=None, dyn_dims=Non
             sd_model.set_adapters(adapter_names=lora_diffusers.diffuser_loaded, adapter_weights=lora_diffusers.diffuser_scales)
         except Exception as e:
             if str(e) not in exclude_errors:
-                log.error(f'Network load: type=LoRA action=strength {e!s}')
+                log.error(f'Network load: type=LoRA action=strength {str(e)}')
             if l.debug:
                 errors.display(e, 'LoRA')
         try:
@@ -363,7 +363,7 @@ def network_load(names, te_multipliers=None, unet_multipliers=None, dyn_dims=Non
                 sd_model.unload_lora_weights()
             l.timer.activate += time.time() - t1
         except Exception as e:
-            log.error(f'Network load: type=LoRA action=fuse {e!s}')
+            log.error(f'Network load: type=LoRA action=fuse {str(e)}')
             if l.debug:
                 errors.display(e, 'LoRA')
         shared.sd_model = sd_models.apply_balanced_offload(shared.sd_model, force=True, silent=True) # some layers may end up on cpu without hook

modules/ltx/ltx_process.py

@@ -149,7 +149,7 @@ def run_ltx(task_id,
             extra_networks.deactivate(p)
         shared.state.end()
         progress.finish_task(task_id)
-        yield None, f'LTX Error: {e!s}'
+        yield None, f'LTX Error: {str(e)}'
 
     if model is None or len(model) == 0 or model == 'None':
         yield from abort('Video: no model selected', ok=True)

modules/merging/merge_methods.py

@@ -4,18 +4,18 @@ import torch
 from torch import Tensor
 
 __all__ = [
+    "weighted_sum",
+    "weighted_subtraction",
+    "tensor_sum",
     "add_difference",
-    "distribution_crossover",
+    "sum_twice",
+    "triple_sum",
     "euclidean_add_difference",
     "multiply_difference",
-    "similarity_add_difference",
-    "sum_twice",
-    "tensor_sum",
-    "ties_add_difference",
     "top_k_tensor_sum",
-    "triple_sum",
-    "weighted_subtraction",
-    "weighted_sum",
+    "similarity_add_difference",
+    "distribution_crossover",
+    "ties_add_difference",
 ]
 
 

modules/onnx_impl/execution_providers.py

@@ -96,7 +96,7 @@ def get_provider() -> tuple:
 
 def install_execution_provider(ep: ExecutionProvider):
     import importlib  # pylint: disable=deprecated-module
-    from installer import install, uninstall
+    from installer import installed, install, uninstall
     res = "<br><pre>"
     res += uninstall(["onnxruntime", "onnxruntime-directml", "onnxruntime-gpu", "onnxruntime-training", "onnxruntime-openvino"], quiet=True)
     packages = ["onnxruntime"] # Failed to load olive: cannot import name '__version__' from 'onnxruntime'

modules/postprocess/swinir_model_arch.py

@@ -238,7 +238,7 @@ class SwinTransformerBlock(nn.Module):
 
     def forward(self, x, x_size):
         H, W = x_size
-        B, _L, C = x.shape
+        B, L, C = x.shape
         # assert L == H * W, "input feature has wrong size"
 
         shortcut = x
@@ -559,7 +559,7 @@ class PatchUnEmbed(nn.Module):
         self.embed_dim = embed_dim
 
     def forward(self, x, x_size):
-        B, _HW, _C = x.shape
+        B, HW, C = x.shape
         x = x.transpose(1, 2).view(B, self.embed_dim, x_size[0], x_size[1])  # B Ph*Pw C
         return x
 

modules/postprocess/swinir_model_arch_v2.py

@@ -266,7 +266,7 @@ class SwinTransformerBlock(nn.Module):
 
     def forward(self, x, x_size):
         H, W = x_size
-        B, _L, C = x.shape
+        B, L, C = x.shape
         #assert L == H * W, "input feature has wrong size"
 
         shortcut = x
@@ -476,7 +476,7 @@ class PatchEmbed(nn.Module):
             self.norm = None
 
     def forward(self, x):
-        _B, _C, _H, _W = x.shape
+        B, C, H, W = x.shape
         # FIXME look at relaxing size constraints
         # assert H == self.img_size[0] and W == self.img_size[1],
         #     f"Input image size ({H}*{W}) doesn't match model ({self.img_size[0]}*{self.img_size[1]})."
@@ -591,7 +591,7 @@ class PatchUnEmbed(nn.Module):
         self.embed_dim = embed_dim
 
     def forward(self, x, x_size):
-        B, _HW, _C = x.shape
+        B, HW, C = x.shape
         x = x.transpose(1, 2).view(B, self.embed_dim, x_size[0], x_size[1])  # B Ph*Pw C
         return x
 

modules/processing.py

@@ -8,6 +8,10 @@ from modules.logger import log
 from modules.sd_hijack_hypertile import context_hypertile_vae, context_hypertile_unet
 from modules.processing_class import ( # pylint: disable=unused-import
     StableDiffusionProcessing,
+    StableDiffusionProcessingTxt2Img,
+    StableDiffusionProcessingImg2Img,
+    StableDiffusionProcessingVideo,
+    StableDiffusionProcessingControl,
 )
 from modules.processing_info import create_infotext
 

modules/ras/__init__.py

@@ -19,7 +19,7 @@ def apply(pipe, p: processing.StableDiffusionProcessing):
     MANAGER.width = p.width
     MANAGER.height = p.height
     MANAGER.error_reset_steps = [int(1*p.steps/3), int(2*p.steps/3)]
-    log.info(f'RAS: scheduler={pipe.scheduler.__class__.__name__} {MANAGER!s}')
+    log.info(f'RAS: scheduler={pipe.scheduler.__class__.__name__} {str(MANAGER)}')
     MANAGER.reset_cache()
     MANAGER.generate_skip_token_list()
     pipe.transformer.old_forward = pipe.transformer.forward

modules/res4lyf/res_unified_scheduler.py

@@ -304,7 +304,7 @@ class RESUnifiedScheduler(SchedulerMixin, ConfigMixin):
             return SchedulerOutput(prev_sample=x_next)
 
         # GET COEFFICIENTS
-        b, _h_val = self._get_coefficients(sigma, sigma_next)
+        b, h_val = self._get_coefficients(sigma, sigma_next)
 
         if len(b) == 1:
             res = b[0] * x0

modules/schedulers/scheduler_dc.py

@@ -126,7 +126,7 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
             Any other scheduler that if specified, the algorithm becomes `solver_p + UniC`.
         use_karras_sigmas (`bool`, *optional*, defaults to `False`):
             Whether to use Karras sigmas for step sizes in the noise schedule during the sampling process. If `True`,
-            the sigmas are determined according to a sequence of noise levels {sigma_i}.
+            the sigmas are determined according to a sequence of noise levels {σi}.
         timestep_spacing (`str`, defaults to `"linspace"`):
             The way the timesteps should be scaled. Refer to Table 2 of the [Common Diffusion Noise Schedules and
             Sample Steps are Flawed](https://huggingface.co/papers/2305.08891) for more information.
@@ -449,7 +449,7 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
         model_output: torch.FloatTensor = None,
         *args,
         sample: torch.FloatTensor = None,
-        order: int | None = None,
+        order: int = None,
         **kwargs,
     ) -> torch.FloatTensor:
         """
@@ -488,12 +488,13 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
             )
         model_output_list = self.model_outputs
 
+        s0 = self.timestep_list[-1]
         m0 = model_output_list[-1]
         assert m0 is not None
         x = sample
 
         if self.solver_p:
-            raise NotImplementedError
+            raise NotImplementedError()
 
         sigma_t, sigma_s0 = self.sigmas[self.step_index + 1], self.sigmas[self.step_index]
         alpha_t, sigma_t = self._sigma_to_alpha_sigma_t(sigma_t)
@@ -533,7 +534,7 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
         elif self.config.solver_type == "bh2":
             B_h = torch.expm1(hh)
         else:
-            raise NotImplementedError
+            raise NotImplementedError()
 
         for i in range(1, order + 1):
             R.append(torch.pow(rks, i - 1))
@@ -578,7 +579,7 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
         *args,
         last_sample: torch.FloatTensor = None,
         this_sample: torch.FloatTensor = None,
-        order: int | None = None,
+        order: int = None,
         **kwargs,
     ) -> torch.FloatTensor:
         """
@@ -668,7 +669,7 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
         elif self.config.solver_type == "bh2":
             B_h = torch.expm1(hh)
         else:
-            raise NotImplementedError
+            raise NotImplementedError()
 
         for i in range(1, order + 1):
             R.append(torch.pow(rks, i - 1))
@@ -810,7 +811,7 @@ class DCSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
             return loss
 
         optimizer = torch.optim.AdamW([ratio_param], lr=0.1)
-        for _ in range(self.num_iters):
+        for iter_ in range(self.num_iters):
             optimizer.zero_grad()
             loss = closure(ratio_param)
             loss.backward()

modules/schedulers/scheduler_dpm_flowmatch.py

@@ -170,7 +170,7 @@ class FlowMatchDPMSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
         from installer import install
         install('torchsde==0.2.6', 'torchsde', quiet=True)
         try:
-            pass
+            import torchsde
         except Exception as e:
             raise ImportError("Failed to import torchsde. Please make sure it is installed correctly.") from e
 
@@ -234,7 +234,7 @@ class FlowMatchDPMSolverMultistepScheduler(SchedulerMixin, ConfigMixin):
         return math.exp(mu) / (math.exp(mu) + (1 / t - 1) ** sigma)
 
     def set_timesteps(self,
-        num_inference_steps: int | None = None,
+        num_inference_steps: int = None,
         device: Union[str, torch.device] = None,
         sigmas: Optional[List[float]] = None,
         mu: Optional[float] = None,

modules/schedulers/scheduler_tdd.py

@@ -395,7 +395,7 @@ class TDDScheduler(DPMSolverSinglestepScheduler):
         model_output_list: List[torch.FloatTensor],
         *args,
         sample: torch.FloatTensor = None,
-        order: int | None = None,
+        order: int = None,
         **kwargs,
     ) -> torch.FloatTensor:
         timestep_list = args[0] if len(args) > 0 else kwargs.pop("timestep_list", None)

modules/schedulers/scheduler_unipc_flowmatch.py

@@ -54,7 +54,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             Any other scheduler that if specified, the algorithm becomes `solver_p + UniC`.
         use_karras_sigmas (`bool`, *optional*, defaults to `False`):
             Whether to use Karras sigmas for step sizes in the noise schedule during the sampling process. If `True`,
-            the sigmas are determined according to a sequence of noise levels {sigma_i}.
+            the sigmas are determined according to a sequence of noise levels {σi}.
         use_exponential_sigmas (`bool`, *optional*, defaults to `False`):
             Whether to use exponential sigmas for step sizes in the noise schedule during the sampling process.
         timestep_spacing (`str`, defaults to `"linspace"`):
@@ -311,7 +311,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             )
 
         sigma = self.sigmas[self.step_index]
-        _alpha_t, sigma_t = self._sigma_to_alpha_sigma_t(sigma)
+        alpha_t, sigma_t = self._sigma_to_alpha_sigma_t(sigma)
 
         if self.predict_x0:
             if self.config.prediction_type == "flow_prediction":
@@ -350,7 +350,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         model_output: torch.Tensor,
         *args,
         sample: torch.Tensor = None,
-        order: int | None = None,  # pyright: ignore
+        order: int = None,  # pyright: ignore
         **kwargs,
     ) -> torch.Tensor:
         """
@@ -439,7 +439,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         elif self.config.solver_type == "bh2":
             B_h = torch.expm1(hh)
         else:
-            raise NotImplementedError
+            raise NotImplementedError()
 
         for i in range(1, order + 1):
             R.append(torch.pow(rks, i - 1))
@@ -487,7 +487,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         *args,
         last_sample: torch.Tensor = None,
         this_sample: torch.Tensor = None,
-        order: int | None = None,  # pyright: ignore
+        order: int = None,  # pyright: ignore
         **kwargs,
     ) -> torch.Tensor:
         """
@@ -582,7 +582,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         elif self.config.solver_type == "bh2":
             B_h = torch.expm1(hh)
         else:
-            raise NotImplementedError
+            raise NotImplementedError()
 
         for i in range(1, order + 1):
             R.append(torch.pow(rks, i - 1))

modules/sd_models.py

@@ -15,9 +15,9 @@ from modules import timer, paths, shared, shared_items, modelloader, devices, sc
 from modules.memstats import memory_stats
 from modules.shared_helpers import walk_files
 from modules.modeldata import model_data
-from modules.sd_checkpoint import CheckpointInfo, select_checkpoint, list_models, checkpoint_titles, get_closest_checkpoint_match # pylint: disable=unused-import
+from modules.sd_checkpoint import CheckpointInfo, select_checkpoint, list_models, checkpoint_titles, get_closest_checkpoint_match, update_model_hashes, write_metadata, checkpoints_list # pylint: disable=unused-import
 from modules.sd_offload import get_module_names, disable_offload, set_diffuser_offload, apply_balanced_offload, set_accelerate # pylint: disable=unused-import
-from modules.sd_models_utils import NoWatermark, get_signature, path_to_repo, apply_function_to_model # pylint: disable=unused-import
+from modules.sd_models_utils import NoWatermark, get_signature, get_call, path_to_repo, apply_function_to_model, read_state_dict, get_state_dict_from_checkpoint # pylint: disable=unused-import
 
 
 model_dir = "Stable-diffusion"
@@ -331,6 +331,7 @@ def load_diffuser_initial(diffusers_load_config, op='model'):
 
 
 def load_diffuser_force(detected_model_type, checkpoint_info, diffusers_load_config, op='model'):
+    from modules import sdnq # pylint: disable=unused-import
     sd_model = None
     global allow_post_quant # pylint: disable=global-statement
     unload_model_weights(op=op)

modules/sdnq/layers/__init__.py

@@ -28,7 +28,7 @@ class SDNQLayer(torch.nn.Module):
         return self.forward_func(self, *args, **kwargs)
 
     def __repr__(self):
-        return f"{self.__class__.__name__}(original_class={self.original_class} forward_func={self.forward_func} sdnq_dequantizer={getattr(self, 'sdnq_dequantizer', None)!r})"
+        return f"{self.__class__.__name__}(original_class={self.original_class} forward_func={self.forward_func} sdnq_dequantizer={repr(getattr(self, 'sdnq_dequantizer', None))})"
 
 
 class SDNQLinear(SDNQLayer, torch.nn.Linear):

modules/seedvr/src/utils/color_fix.py

@@ -2,8 +2,8 @@ import torch
 from PIL import Image
 from torch import Tensor
 from torch.nn import functional as F
+from ..common.half_precision_fixes import safe_pad_operation, safe_interpolate_operation
 from torchvision.transforms import ToTensor, ToPILImage
-from modules.seedvr.src.common.half_precision_fixes import safe_pad_operation, safe_interpolate_operation
 
 def adain_color_fix(target: Image.Image, source: Image.Image):
     # Convert images to tensors

modules/shared.py

@@ -15,17 +15,21 @@ log.debug('Initializing: shared module')
 import modules.memmon
 import modules.paths as paths
 from modules.json_helpers import readfile # pylint: disable=W0611
-from modules.shared_helpers import listdir # pylint: disable=W0611
-from modules import errors, devices, shared_state, cmd_args, history, files_cache # pylint: disable=unused-import
+from modules.shared_helpers import listdir, req # pylint: disable=W0611
+from modules import errors, devices, shared_state, cmd_args, theme, history, files_cache # pylint: disable=unused-import
 from modules.shared_defaults import get_default_modes
 from modules.memstats import memory_stats # pylint: disable=unused-import
 
 log.debug('Initializing: pipelines')
+from modules import shared_items # pylint: disable=unused-import
+from modules.caption.openclip import get_clip_models, refresh_clip_models # pylint: disable=unused-import
+from modules.caption.vqa import vlm_models, vlm_prompts, vlm_system, vlm_default # pylint: disable=unused-import
 
 
 if TYPE_CHECKING:
     # Behavior modified by __future__.annotations
     from diffusers import DiffusionPipeline
+    from modules.shared_legacy import LegacyOption
     from modules.ui_extra_networks import ExtraNetworksPage
 
 
@@ -77,6 +81,7 @@ data_path = paths.data_path
 backend = Backend.DIFFUSERS
 if cmd_opts.use_openvino: # override for openvino
     os.environ.setdefault('PYTORCH_TRACING_MODE', 'TORCHFX')
+    from modules.intel.openvino import get_device_list as get_openvino_device_list # pylint: disable=ungrouped-imports,unused-import
 elif cmd_opts.use_ipex or devices.has_xpu():
     from modules.intel.ipex import ipex_init
     ok, e = ipex_init()
@@ -152,6 +157,7 @@ startup_offload_mode, startup_offload_min_gpu, startup_offload_max_gpu, startup_
 
 log.debug('Initializing: settings')
 from modules import ui_definitions
+from modules.ui_definitions import OptionInfo, options_section # pylint: disable=unused-import
 options_templates = ui_definitions.create_settings(cmd_opts)
 from modules.shared_legacy import get_legacy_options
 options_templates.update(get_legacy_options())

modules/sharpfin/cms.py

@@ -131,7 +131,7 @@ def apply_srgb(
                             flags=flags
                         )
                     else:
-                        img = cast('Image', profileToProfile(
+                        img = cast(Image, profileToProfile(
                             img,
                             profile,
                             _SRGB,

modules/teacache/teacache_hidream.py

@@ -1,6 +1,6 @@
 from typing import Any, Dict, List, Optional, Tuple
 from diffusers.models.modeling_outputs import Transformer2DModelOutput
-from diffusers.utils import deprecate, USE_PEFT_BACKEND, logging, scale_lora_layers, unscale_lora_layers
+from diffusers.utils import logging, deprecate, USE_PEFT_BACKEND, logging, scale_lora_layers, unscale_lora_layers
 
 import torch
 import numpy as np

modules/ui_models_load.py

@@ -222,7 +222,7 @@ def create_ui(gr_status, gr_file):
             if param.name == 'self' or param.name == 'args' or param.name == 'kwargs':
                 continue
             component = Component(param)
-            debug_log(f'Model component: {component!s}')
+            debug_log(f'Model component: {str(component)}')
             components.append(component)
         return components
 

modules/video_models/google_veo.py

@@ -6,7 +6,7 @@ import sys
 sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..')))
 
 from PIL import Image
-from installer import install
+from installer import install, reload
 from modules.logger import log
 
 

modules/video_models/video_load.py

@@ -30,6 +30,7 @@ def load_custom(model_name: str):
 
 
 def load_model(selected: models_def.Model):
+    from modules import sdnq # pylint: disable=unused-import
     if selected is None or selected.repo is None:
         return ''
     global loaded_model # pylint: disable=global-statement

modules/video_models/video_ui.py

@@ -64,7 +64,7 @@ def run_video(*args):
     selected = get_selected(engine, model)
     if not selected or engine is None or model is None or engine == 'None' or model == 'None':
         return video_utils.queue_err('model not selected')
-    debug(f'Video run: {selected!s}')
+    debug(f'Video run: {str(selected)}')
     if selected and 'Hunyuan' in selected.name:
         return video_run.generate(*args)
     elif selected and 'LTX' in selected.name:

pipelines/bria/bria_pipeline.py

@@ -622,7 +622,7 @@ class BriaPipeline(FluxPipeline):
 
     @staticmethod
     def _unpack_latents(latents, height, width, vae_scale_factor):
-        batch_size, _num_patches, channels = latents.shape
+        batch_size, num_patches, channels = latents.shape
 
         height = height // vae_scale_factor
         width = width // vae_scale_factor

pipelines/f_lite/f_lite.model.py

@@ -233,7 +233,7 @@ class PatchEmbed(nn.Module):
         self.patch_size = patch_size
 
     def forward(self, x):
-        _B, _C, _H, _W = x.shape
+        B, C, H, W = x.shape
         x = self.patch_proj(x)
         x = rearrange(x, "b c h w -> b (h w) c")
         return x
@@ -380,7 +380,7 @@ class DiT(ModelMixin, ConfigMixin, FromOriginalModelMixin, PeftAdapterMixin):  #
 
     @apply_forward_hook
     def forward(self, x, context, timesteps):
-        b, _c, h, w = x.shape
+        b, c, h, w = x.shape
         x = self.patch_embed(x)  # b, T, d
 
         x = torch.cat([self.register_tokens.repeat(b, 1, 1), x], 1)  # b, T + N, d

pipelines/f_lite/model.py

@@ -233,7 +233,7 @@ class PatchEmbed(nn.Module):
         self.patch_size = patch_size
 
     def forward(self, x):
-        _B, _C, _H, _W = x.shape
+        B, C, H, W = x.shape
         x = self.patch_proj(x)
         x = rearrange(x, "b c h w -> b (h w) c")
         return x
@@ -380,7 +380,7 @@ class DiT(ModelMixin, ConfigMixin, FromOriginalModelMixin, PeftAdapterMixin):  #
 
     @apply_forward_hook
     def forward(self, x, context, timesteps):
-        b, _c, h, w = x.shape
+        b, c, h, w = x.shape
         x = self.patch_embed(x)  # b, T, d
 
         x = torch.cat([self.register_tokens.repeat(b, 1, 1), x], 1)  # b, T + N, d

pipelines/hidream/scheduler_flashfloweuler.py

@@ -194,7 +194,7 @@ class FlashFlowMatchEulerDiscreteScheduler(SchedulerMixin, ConfigMixin):
 
     def set_timesteps(
             self,
-            num_inference_steps: int | None = None,
+            num_inference_steps: int = None,
             device: Union[str, torch.device] = None,
             sigmas: Optional[List[float]] = None,
             mu: Optional[float] = None,

pipelines/hidream/scheduler_flowunipc.py

@@ -53,7 +53,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             Any other scheduler that if specified, the algorithm becomes `solver_p + UniC`.
         use_karras_sigmas (`bool`, *optional*, defaults to `False`):
             Whether to use Karras sigmas for step sizes in the noise schedule during the sampling process. If `True`,
-            the sigmas are determined according to a sequence of noise levels {sigma_i}.
+            the sigmas are determined according to a sequence of noise levels {σi}.
         use_exponential_sigmas (`bool`, *optional*, defaults to `False`):
             Whether to use exponential sigmas for step sizes in the noise schedule during the sampling process.
         timestep_spacing (`str`, defaults to `"linspace"`):
@@ -294,7 +294,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             `torch.Tensor`:
                 The converted model output.
         """
-        _ = args[0] if len(args) > 0 else kwargs.pop("timestep", None)
+        timestep = args[0] if len(args) > 0 else kwargs.pop("timestep", None)
         if sample is None:
             if len(args) > 1:
                 sample = args[1]
@@ -303,7 +303,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
                     "missing `sample` as a required keyward argument")
 
         sigma = self.sigmas[self.step_index]
-        _alpha_t, sigma_t = self._sigma_to_alpha_sigma_t(sigma)
+        alpha_t, sigma_t = self._sigma_to_alpha_sigma_t(sigma)
 
         if self.predict_x0:
             if self.config.prediction_type == "flow_prediction":
@@ -342,7 +342,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         model_output: torch.Tensor,
         *args,
         sample: torch.Tensor = None,
-        order: int | None = None,  # pyright: ignore
+        order: int = None,  # pyright: ignore
         **kwargs,
     ) -> torch.Tensor:
         """
@@ -362,7 +362,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             `torch.Tensor`:
                 The sample tensor at the previous timestep.
         """
-        _ = args[0] if len(args) > 0 else kwargs.pop(
+        prev_timestep = args[0] if len(args) > 0 else kwargs.pop(
             "prev_timestep", None)
         if sample is None:
             if len(args) > 1:
@@ -473,7 +473,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
         *args,
         last_sample: torch.Tensor = None,
         this_sample: torch.Tensor = None,
-        order: int | None = None,  # pyright: ignore
+        order: int = None,  # pyright: ignore
         **kwargs,
     ) -> torch.Tensor:
         """
@@ -495,7 +495,7 @@ class FlowUniPCMultistepScheduler(SchedulerMixin, ConfigMixin):
             `torch.Tensor`:
                 The corrected sample tensor at the current timestep.
         """
-        _ = args[0] if len(args) > 0 else kwargs.pop(
+        this_timestep = args[0] if len(args) > 0 else kwargs.pop(
             "this_timestep", None)
         if last_sample is None:
             if len(args) > 1:

pipelines/lumina_dimmo/lumina_dimoo.py

@@ -17,7 +17,8 @@ import sys
 from abc import abstractmethod
 from dataclasses import dataclass, fields
 from enum import Enum
-from typing import Callable, Dict, Iterable, List, NamedTuple, Optional, Sequence, Tuple, Union, cast
+from typing import Any, Callable, Dict, Iterable, List, NamedTuple, Optional, Sequence, Tuple, Union, cast
+from accelerate import init_empty_weights
 
 from tqdm.rich import tqdm
 import numpy as np
@@ -34,6 +35,7 @@ from diffusers import DiffusionPipeline, VQModel
 from diffusers.utils import BaseOutput, logging, replace_example_docstring
 from diffusers.image_processor import PipelineImageInput, VaeImageProcessor
 
+from diffusers.pipelines.pipeline_utils import ImagePipelineOutput
 
 
 logger = logging.get_logger(__name__)  # pylint: disable=invalid-name
@@ -51,7 +53,7 @@ class StrEnum(str, Enum):
         return self.value
 
     def __repr__(self) -> str:
-        return f"'{self!s}'"
+        return f"'{str(self)}'"
 
 
 class LayerNormType(StrEnum):
@@ -943,7 +945,7 @@ class LLaDALlamaBlock(LLaDABlock):
         cat="cond",
         to_compute_mask=None,
     ) -> Tuple[torch.Tensor, Optional[Tuple[torch.Tensor, torch.Tensor]]]:
-        _B, _T, D = x.shape
+        B, T, D = x.shape
 
         x_normed = self.attn_norm(x)
         q = self.q_proj(x_normed)
@@ -1610,7 +1612,7 @@ def get_num_transfer_tokens(mask_index, steps):
 
 
 def mask_by_random_topk(keep_n, probs, temperature=1.0, generator=None):
-    B, _S = probs.shape
+    B, S = probs.shape
     noise = gumbel_noise(probs, generator=generator)
 
     conf = probs / temperature + noise
@@ -2047,7 +2049,7 @@ class LuminaDiMOOPipeline(DiffusionPipeline):
         """
         device = next(model.parameters()).device
         prompt = prompt.to(device)
-        B, _P = prompt.shape
+        B, P = prompt.shape
         assert B == 1, "batch>1 not supported - wrap in loop if needed"
 
         x = prompt
@@ -2163,7 +2165,7 @@ class LuminaDiMOOPipeline(DiffusionPipeline):
 
         device = next(model.parameters()).device
         prompt = prompt.to(device)
-        B, _P = prompt.shape
+        B, P = prompt.shape
         assert B == 1, "batch>1 not supported - wrap in loop if needed"
 
         x = prompt
@@ -2492,7 +2494,7 @@ class LuminaDiMOOPipeline(DiffusionPipeline):
         uncon_prompt_token = self.tokenizer(uncon_prompt)["input_ids"]
 
         if painting_mode:
-            img_mask_token, _img_vis = encode_img_with_paint(
+            img_mask_token, img_vis = encode_img_with_paint(
                 painting_image,
                 vqvae=self.vqvae,
                 mask_h_ratio=mask_h_ratio,
@@ -2572,7 +2574,7 @@ class LuminaDiMOOPipeline(DiffusionPipeline):
         processed_image = var_center_crop(image, crop_size_list=crop_size_list)
 
         image_width, image_height = processed_image.size
-        _seq_len, _newline_every, _token_grid_height, _token_grid_width = calculate_vq_params(
+        seq_len, newline_every, _token_grid_height, _token_grid_width = calculate_vq_params(
             image_height, image_width, self.vae_scale_factor
         )
 

pipelines/meissonic/pipeline_img2img.py

@@ -17,7 +17,7 @@ import torch
 from transformers import CLIPTextModelWithProjection, CLIPTokenizer
 
 from diffusers.image_processor import PipelineImageInput, VaeImageProcessor
-from diffusers.models import VQModel
+from diffusers.models import UVit2DModel, VQModel
 # from diffusers.schedulers import AmusedScheduler
 from .scheduler import Scheduler
 from diffusers.utils import replace_example_docstring
@@ -276,7 +276,7 @@ class MeissonicImg2ImgPipeline(DiffusionPipeline):
             self.vqvae.float()
 
         latents = self.vqvae.encode(image.to(dtype=self.vqvae.dtype, device=self._execution_device)).latents
-        latents_bsz, _channels, latents_height, latents_width = latents.shape
+        latents_bsz, channels, latents_height, latents_width = latents.shape
         latents = self.vqvae.quantize(latents)[2][2].reshape(latents_bsz, latents_height, latents_width)
         latents = self.scheduler.add_noise(
             latents, self.scheduler.timesteps[start_timestep_idx - 1], generator=generator

pipelines/meissonic/pipeline_inpaint.py

@@ -289,7 +289,7 @@ class MeissonicInpaintPipeline(DiffusionPipeline):
             self.vqvae.float()
 
         latents = self.vqvae.encode(image.to(dtype=self.vqvae.dtype, device=self._execution_device)).latents
-        latents_bsz, _channels, latents_height, latents_width = latents.shape
+        latents_bsz, channels, latents_height, latents_width = latents.shape
         latents = self.vqvae.quantize(latents)[2][2].reshape(latents_bsz, latents_height, latents_width)
 
         mask = self.mask_processor.preprocess(

pipelines/meissonic/transformer.py

@@ -22,7 +22,7 @@ import torch.nn.functional as F
 
 from diffusers.configuration_utils import ConfigMixin, register_to_config
 from diffusers.loaders import FromOriginalModelMixin, PeftAdapterMixin
-from diffusers.models.attention import FeedForward, SkipFFTransformerBlock
+from diffusers.models.attention import FeedForward, BasicTransformerBlock, SkipFFTransformerBlock
 from diffusers.models.attention_processor import (
     Attention,
     AttentionProcessor,
@@ -30,7 +30,7 @@ from diffusers.models.attention_processor import (
     # FusedFluxAttnProcessor2_0,
 )
 from diffusers.models.modeling_utils import ModelMixin
-from diffusers.models.normalization import AdaLayerNormZero, AdaLayerNormZeroSingle, GlobalResponseNorm, RMSNorm
+from diffusers.models.normalization import AdaLayerNormContinuous, AdaLayerNormZero, AdaLayerNormZeroSingle, GlobalResponseNorm, RMSNorm
 from diffusers.utils import USE_PEFT_BACKEND, is_torch_version, logging, scale_lora_layers, unscale_lora_layers
 from diffusers.utils.torch_utils import maybe_allow_in_graph
 from diffusers.models.embeddings import CombinedTimestepGuidanceTextProjEmbeddings, CombinedTimestepTextProjEmbeddings,TimestepEmbedding, get_timestep_embedding #,FluxPosEmbed