resolve conflicts

ldm/generate.py

@@ -292,6 +292,7 @@ class Generate:
             # Set this True to handle KeyboardInterrupt internally
             catch_interrupts = False,
             hires_fix        = False,
+            use_mps_noise    = False,
             **args,
     ):   # eat up additional cruft
         """
@@ -433,6 +434,7 @@ class Generate:
             generator.set_variation(
                 self.seed, variation_amount, with_variations
             )
+            generator.use_mps_noise = use_mps_noise
 
             checker = {
                 'checker':self.safety_checker,

ldm/invoke/args.py

@@ -816,6 +816,13 @@ class Args(object):
             type=str,
             help='list of variations to apply, in the format `seed:weight,seed:weight,...'
         )
+        render_group.add_argument(
+            '--use_mps_noise',
+            action='store_true',
+            dest='use_mps_noise',
+            help='Simulate noise on M1 systems to get the same results'
+        )
+
         return parser
 
 def format_metadata(**kwargs):

ldm/invoke/generator/base.py

@@ -28,6 +28,7 @@ class Generator():
         self.threshold = 0
         self.variation_amount = 0
         self.with_variations = []
+        self.use_mps_noise       = False
 
     # this is going to be overridden in img2img.py, txt2img.py and inpaint.py
     def get_make_image(self,prompt,**kwargs):

ldm/invoke/generator/txt2img.py

@@ -59,7 +59,7 @@ class Txt2Img(Generator):
     # returns a tensor filled with random numbers from a normal distribution
     def get_noise(self,width,height):
         device         = self.model.device
-        if device.type == 'mps':
+        if self.use_mps_noise or device.type == 'mps':
             x = torch.randn([1,
                                 self.latent_channels,
                                 height // self.downsampling_factor,

ldm/invoke/generator/txt2img2img.py

@@ -118,7 +118,7 @@ class Txt2Img2Img(Generator):
             scaled_height = height
             
         device      = self.model.device
-        if device.type == 'mps':
+        if self.use_mps_noise or device.type == 'mps':
             return torch.randn([1,
                                 self.latent_channels,
                                 scaled_height // self.downsampling_factor,