auto switch mps device to cpu device

lama_cleaner/helper.py

@@ -8,10 +8,18 @@ import cv2
 from PIL import Image, ImageOps
 import numpy as np
 import torch
+from lama_cleaner.const import MPS_SUPPORT_MODELS
 from loguru import logger
 from torch.hub import download_url_to_file, get_dir
 
 
+def switch_mps_device(model_name, device):
+    if model_name not in MPS_SUPPORT_MODELS and (device == "mps" or device == torch.device('mps')):
+        logger.info(f"{model_name} not support mps, switch to cpu")
+        return torch.device('cpu')
+    return device
+
+
 def get_cache_path_by_url(url):
     parts = urlparse(url)
     hub_dir = get_dir()

lama_cleaner/model/base.py

@@ -6,11 +6,12 @@ import torch
 import numpy as np
 from loguru import logger
 
-from lama_cleaner.helper import boxes_from_mask, resize_max_size, pad_img_to_modulo
+from lama_cleaner.helper import boxes_from_mask, resize_max_size, pad_img_to_modulo, switch_mps_device
 from lama_cleaner.schema import Config, HDStrategy
 
 
 class InpaintModel:
+    name = "base"
     min_size: Optional[int] = None
     pad_mod = 8
     pad_to_square = False
@@ -21,6 +22,7 @@ class InpaintModel:
         Args:
             device:
         """
+        device = switch_mps_device(self.name, device)
         self.device = device
         self.init_model(device, **kwargs)
 

lama_cleaner/model/fcf.py

@@ -1131,6 +1131,7 @@ FCF_MODEL_URL = os.environ.get(
 
 
 class FcF(InpaintModel):
+    name = "fcf"
     min_size = 512
     pad_mod = 512
     pad_to_square = True

lama_cleaner/model/instruct_pix2pix.py

@@ -9,6 +9,7 @@ from lama_cleaner.schema import Config
 
 
 class InstructPix2Pix(DiffusionInpaintModel):
+    name = "instruct_pix2pix"
     pad_mod = 8
     min_size = 512
 

lama_cleaner/model/lama.py

@@ -16,6 +16,7 @@ LAMA_MODEL_URL = os.environ.get(
 
 
 class LaMa(InpaintModel):
+    name = "lama"
     pad_mod = 8
 
     def init_model(self, device, **kwargs):

lama_cleaner/model/ldm.py

@@ -225,6 +225,7 @@ class LatentDiffusion(DDPM):
 
 
 class LDM(InpaintModel):
+    name = "ldm"
     pad_mod = 32
 
     def __init__(self, device, fp16: bool = True, **kwargs):

lama_cleaner/model/manga.py

@@ -76,6 +76,7 @@ MANGA_LINE_MODEL_URL = os.environ.get(
 
 
 class Manga(InpaintModel):
+    name = "manga"
     pad_mod = 16
 
     def init_model(self, device, **kwargs):

lama_cleaner/model/mat.py

@@ -1401,6 +1401,7 @@ MAT_MODEL_URL = os.environ.get(
 
 
 class MAT(InpaintModel):
+    name = "mat"
     min_size = 512
     pad_mod = 512
     pad_to_square = True

lama_cleaner/model/opencv2.py

@@ -2,12 +2,11 @@ import cv2
 from lama_cleaner.model.base import InpaintModel
 from lama_cleaner.schema import Config
 
-flag_map = {
+flag_map = {"INPAINT_NS": cv2.INPAINT_NS, "INPAINT_TELEA": cv2.INPAINT_TELEA}
-    "INPAINT_NS": cv2.INPAINT_NS,
+
-    "INPAINT_TELEA": cv2.INPAINT_TELEA
-}
 
 class OpenCV2(InpaintModel):
+    name = "cv2"
     pad_mod = 1
 
     @staticmethod
@@ -20,5 +19,10 @@ class OpenCV2(InpaintModel):
         mask: [H, W, 1]
         return: BGR IMAGE
         """
-        cur_res = cv2.inpaint(image[:,:,::-1], mask, inpaintRadius=config.cv2_radius, flags=flag_map[config.cv2_flag])
+        cur_res = cv2.inpaint(
+            image[:, :, ::-1],
+            mask,
+            inpaintRadius=config.cv2_radius,
+            flags=flag_map[config.cv2_flag],
+        )
         return cur_res

lama_cleaner/model/paint_by_example.py

@@ -11,6 +11,7 @@ from lama_cleaner.schema import Config
 
 
 class PaintByExample(DiffusionInpaintModel):
+    name = "paint_by_example"
     pad_mod = 8
     min_size = 512
 

lama_cleaner/model/sd.py

@@ -160,8 +160,10 @@ class SD(DiffusionInpaintModel):
 
 
 class SD15(SD):
+    name = "sd1.5"
     model_id_or_path = "runwayml/stable-diffusion-inpainting"
 
 
 class SD2(SD):
+    name = "sd2"
     model_id_or_path = "stabilityai/stable-diffusion-2-inpainting"

lama_cleaner/model/zits.py

@@ -203,6 +203,7 @@ def to_device(data, device):
 
 
 class ZITS(InpaintModel):
+    name = "zits"
     min_size = 256
     pad_mod = 32
     pad_to_square = True

lama_cleaner/model_manager.py

@@ -1,6 +1,7 @@
 import torch
 import gc
 
+from lama_cleaner.helper import switch_mps_device
 from lama_cleaner.model.fcf import FcF
 from lama_cleaner.model.lama import LaMa
 from lama_cleaner.model.ldm import LDM
@@ -13,8 +14,19 @@ from lama_cleaner.model.zits import ZITS
 from lama_cleaner.model.opencv2 import OpenCV2
 from lama_cleaner.schema import Config
 
-models = {"lama": LaMa, "ldm": LDM, "zits": ZITS, "mat": MAT, "fcf": FcF, "sd1.5": SD15, "cv2": OpenCV2, "manga": Manga,
+models = {
-          "sd2": SD2, "paint_by_example": PaintByExample, "instruct_pix2pix": InstructPix2Pix}
+    "lama": LaMa,
+    "ldm": LDM,
+    "zits": ZITS,
+    "mat": MAT,
+    "fcf": FcF,
+    "sd1.5": SD15,
+    "cv2": OpenCV2,
+    "manga": Manga,
+    "sd2": SD2,
+    "paint_by_example": PaintByExample,
+    "instruct_pix2pix": InstructPix2Pix,
+}
 
 
 class ModelManager:
@@ -44,13 +56,15 @@ class ModelManager:
         if new_name == self.name:
             return
         try:
-            if (torch.cuda.memory_allocated() > 0):
+            if torch.cuda.memory_allocated() > 0:
                 # Clear current loaded model from memory
                 torch.cuda.empty_cache()
                 del self.model
                 gc.collect()
 
-            self.model = self.init_model(new_name, self.device, **self.kwargs)
+            self.model = self.init_model(
+                new_name, switch_mps_device(new_name, self.device), **self.kwargs
+            )
             self.name = new_name
         except NotImplementedError as e:
             raise e

lama_cleaner/parse_args.py

@@ -143,12 +143,6 @@ def parse_args():
                 "torch.cuda.is_available() is False, please use --device cpu or check your pytorch installation"
             )
 
-    if args.device == "mps":
-        if args.model not in MPS_SUPPORT_MODELS:
-            parser.error(
-                f"mps only support: {MPS_SUPPORT_MODELS}, but got {args.model}"
-            )
-
     if args.model_dir and args.model_dir is not None:
         if os.path.isfile(args.model_dir):
             parser.error(f"invalid --model-dir: {args.model_dir} is a file")

lama_cleaner/web_config.py

@@ -22,7 +22,8 @@ from lama_cleaner.const import (
     DEFAULT_MODEL,
     DEFAULT_DEVICE,
     NO_GUI_AUTO_CLOSE_HELP,
-    DEFAULT_MODEL_DIR, MPS_SUPPORT_MODELS,
+    DEFAULT_MODEL_DIR,
+    MPS_SUPPORT_MODELS,
 )
 
 _config_file = None
@@ -115,7 +116,7 @@ def main(config_file: str):
         with gr.Row():
             model = gr.Radio(AVAILABLE_MODELS, label="Model", value=init_config.model)
             device = gr.Radio(
-                AVAILABLE_DEVICES, label=f"Device(mps supports {MPS_SUPPORT_MODELS})", value=init_config.device
+                AVAILABLE_DEVICES, label="Device", value=init_config.device
             )
         gui = gr.Checkbox(init_config.gui, label=f"{GUI_HELP}")
         no_gui_auto_close = gr.Checkbox(