update

lama_cleaner/model/instruct_pix2pix.py

@@ -66,9 +66,9 @@ class InstructPix2Pix(DiffusionInpaintModel):
             image=PIL.Image.fromarray(image),
             prompt=config.prompt,
             negative_prompt=config.negative_prompt,
-            num_inference_steps=config.p2p_steps,
+            num_inference_steps=config.sd_steps,
             image_guidance_scale=config.p2p_image_guidance_scale,
-            guidance_scale=config.p2p_guidance_scale,
+            guidance_scale=config.sd_guidance_scale,
             output_type="np",
             generator=torch.manual_seed(config.sd_seed),
         ).images[0]

lama_cleaner/model/paint_by_example.py

@@ -19,7 +19,7 @@ class PaintByExample(DiffusionInpaintModel):
         fp16 = not kwargs.get("no_half", False)
         use_gpu = device == torch.device("cuda") and torch.cuda.is_available()
         torch_dtype = torch.float16 if use_gpu and fp16 else torch.float32
-        model_kwargs = {"local_files_only": kwargs.get("local_files_only", False)}
+        model_kwargs = {}
 
         if kwargs["disable_nsfw"] or kwargs.get("cpu_offload", False):
             logger.info("Disable Paint By Example Model NSFW checker")
@@ -58,24 +58,17 @@ class PaintByExample(DiffusionInpaintModel):
             image=PIL.Image.fromarray(image),
             mask_image=PIL.Image.fromarray(mask[:, :, -1], mode="L"),
             example_image=config.paint_by_example_example_image,
-            num_inference_steps=config.paint_by_example_steps,
+            num_inference_steps=config.sd_steps,
+            guidance_scale=config.sd_guidance_scale,
+            negative_prompt="out of frame, lowres, error, cropped, worst quality, low quality, jpeg artifacts, ugly, duplicate, morbid, mutilated, out of frame, mutation, deformed, blurry, dehydrated, bad anatomy, bad proportions, extra limbs, disfigured, gross proportions, malformed limbs, watermark, signature",
             output_type="np.array",
-            generator=torch.manual_seed(config.paint_by_example_seed),
+            generator=torch.manual_seed(config.sd_seed),
         ).images[0]
 
         output = (output * 255).round().astype("uint8")
         output = cv2.cvtColor(output, cv2.COLOR_RGB2BGR)
         return output
 
-    def forward_post_process(self, result, image, mask, config):
-        if config.paint_by_example_match_histograms:
-            result = self._match_histograms(result, image[:, :, ::-1], mask)
-
-        if config.paint_by_example_mask_blur != 0:
-            k = 2 * config.paint_by_example_mask_blur + 1
-            mask = cv2.GaussianBlur(mask, (k, k), 0)
-        return result, image, mask
-
     @staticmethod
     def is_downloaded() -> bool:
         # model will be downloaded when app start, and can't switch in frontend settings

lama_cleaner/model/utils.py

@@ -1,3 +1,4 @@
+import gc
 import math
 import random
 from typing import Any
@@ -913,6 +914,7 @@ def torch_gc():
     if torch.cuda.is_available():
         torch.cuda.empty_cache()
         torch.cuda.ipc_collect()
+    gc.collect()
 
 
 def set_seed(seed: int):