apply lint fixes again

api/onnx_web/chain/blend_img2img.py

@@ -1,7 +1,7 @@
 from logging import getLogger
 
-import torch
 import numpy as np
+import torch
 from diffusers import OnnxStableDiffusionImg2ImgPipeline
 from PIL import Image
 
@@ -35,7 +35,7 @@ def blend_img2img(
         params.lpw,
     )
     if params.lpw:
-        logger.debug('using LPW pipeline for img2img')
+        logger.debug("using LPW pipeline for img2img")
         rng = torch.manual_seed(params.seed)
         result = pipe.img2img(
             prompt,

api/onnx_web/chain/blend_inpaint.py

@@ -1,8 +1,8 @@
 from logging import getLogger
 from typing import Callable, Tuple
 
-import torch
 import numpy as np
+import torch
 from diffusers import OnnxStableDiffusionInpaintPipeline
 from PIL import Image
 
@@ -70,7 +70,7 @@ def blend_inpaint(
         )
 
         if params.lpw:
-            logger.debug('using LPW pipeline for inpaint')
+            logger.debug("using LPW pipeline for inpaint")
             rng = torch.manual_seed(params.seed)
             result = pipe.inpaint(
                 params.prompt,

api/onnx_web/chain/source_txt2img.py

@@ -1,7 +1,7 @@
 from logging import getLogger
 
-import torch
 import numpy as np
+import torch
 from diffusers import OnnxStableDiffusionPipeline
 from PIL import Image
 
@@ -34,11 +34,15 @@ def source_txt2img(
 
     latents = get_latents_from_seed(params.seed, size)
     pipe = load_pipeline(
-        OnnxStableDiffusionPipeline, params.model, params.scheduler, job.get_device(), params.lpw
+        OnnxStableDiffusionPipeline,
+        params.model,
+        params.scheduler,
+        job.get_device(),
+        params.lpw,
     )
 
     if params.lpw:
-        logger.debug('using LPW pipeline for txt2img')
+        logger.debug("using LPW pipeline for txt2img")
         rng = torch.manual_seed(params.seed)
         result = pipe.text2img(
             prompt,

api/onnx_web/chain/upscale_outpaint.py

@@ -1,8 +1,8 @@
 from logging import getLogger
 from typing import Callable, Tuple
 
-import torch
 import numpy as np
+import torch
 from diffusers import OnnxStableDiffusionInpaintPipeline
 from PIL import Image, ImageDraw
 
@@ -75,7 +75,7 @@ def upscale_outpaint(
             params.lpw,
         )
         if params.lpw:
-            logger.debug('using LPW pipeline for inpaint')
+            logger.debug("using LPW pipeline for inpaint")
             rng = torch.manual_seed(params.seed)
             result = pipe.inpaint(
                 image,
@@ -102,10 +102,8 @@ def upscale_outpaint(
                 negative_prompt=params.negative_prompt,
                 num_inference_steps=params.steps,
                 width=size.width,
-
             )
 
-
         # once part of the image has been drawn, keep it
         draw_mask.rectangle((left, top, left + tile, top + tile), fill="black")
         return result.images[0]
@@ -116,7 +114,9 @@ def upscale_outpaint(
 
     if border.left == border.right and border.top == border.bottom:
         logger.debug("outpainting with an even border, using spiral tiling")
-        output = process_tile_spiral(source_image, SizeChart.auto, 1, [outpaint], overlap=overlap)
+        output = process_tile_spiral(
+            source_image, SizeChart.auto, 1, [outpaint], overlap=overlap
+        )
     else:
         logger.debug("outpainting with an uneven border, using grid tiling")
         output = process_tile_grid(source_image, SizeChart.auto, 1, [outpaint])

api/onnx_web/convert.py

@@ -60,7 +60,7 @@ base_models: Models = {
             "correction-codeformer",
             "https://github.com/sczhou/CodeFormer/releases/download/v0.1.0/codeformer.pth",
             1,
-        )
+        ),
     ],
     "upscaling": [
         (

api/onnx_web/diffusion/load.py

@@ -47,7 +47,11 @@ def get_tile_latents(
 
 
 def load_pipeline(
-    pipeline: DiffusionPipeline, model: str, scheduler: Any, device: DeviceParams, lpw: bool
+    pipeline: DiffusionPipeline,
+    model: str,
+    scheduler: Any,
+    device: DeviceParams,
+    lpw: bool,
 ):
     global last_pipeline_instance
     global last_pipeline_scheduler

api/onnx_web/diffusion/run.py

@@ -1,8 +1,8 @@
 from logging import getLogger
 from typing import Any
 
-import torch
 import numpy as np
+import torch
 from diffusers import OnnxStableDiffusionImg2ImgPipeline, OnnxStableDiffusionPipeline
 from PIL import Image, ImageChops
 
@@ -27,12 +27,16 @@ def run_txt2img_pipeline(
 ) -> None:
     latents = get_latents_from_seed(params.seed, size)
     pipe = load_pipeline(
-        OnnxStableDiffusionPipeline, params.model, params.scheduler, job.get_device(), params.lpw
+        OnnxStableDiffusionPipeline,
+        params.model,
+        params.scheduler,
+        job.get_device(),
+        params.lpw,
     )
     progress = job.get_progress_callback()
 
     if params.lpw:
-        logger.debug('using LPW pipeline for txt2img')
+        logger.debug("using LPW pipeline for txt2img")
         rng = torch.manual_seed(params.seed)
         result = pipe.text2img(
             params.prompt,
@@ -59,7 +63,6 @@ def run_txt2img_pipeline(
             callback=progress,
         )
 
-
     image = result.images[0]
     image = run_upscale_correction(
         job, server, StageParams(), params, image, upscale=upscale
@@ -89,11 +92,11 @@ def run_img2img_pipeline(
         params.model,
         params.scheduler,
         job.get_device(),
-        params.lpw
+        params.lpw,
     )
     progress = job.get_progress_callback()
     if params.lpw:
-        logger.debug('using LPW pipeline for img2img')
+        logger.debug("using LPW pipeline for img2img")
         rng = torch.manual_seed(params.seed)
         result = pipe.img2img(
             source_image,
@@ -118,7 +121,6 @@ def run_img2img_pipeline(
             callback=progress,
         )
 
-
     image = result.images[0]
     image = run_upscale_correction(
         job, server, StageParams(), params, image, upscale=upscale

api/onnx_web/serve.py

@@ -143,7 +143,7 @@ correction_models = []
 upscaling_models = []
 
 
-def get_config_value(key: str, subkey: str = "default", default = None):
+def get_config_value(key: str, subkey: str = "default", default=None):
     return config_params.get(key, {}).get(subkey, default)
 
 
@@ -234,7 +234,14 @@ def pipeline_from_request() -> Tuple[DeviceParams, ImageParams, Size]:
     )
 
     params = ImageParams(
-        model_path, scheduler, prompt, cfg, steps, seed, lpw=lpw, negative_prompt=negative_prompt
+        model_path,
+        scheduler,
+        prompt,
+        cfg,
+        steps,
+        seed,
+        lpw=lpw,
+        negative_prompt=negative_prompt,
     )
     size = Size(width, height)
     return (device, params, size)