fixing aspect ratio on hires

ldm/invoke/generator/txt2img2img.py

@@ -26,29 +26,28 @@ def get_make_image(self,prompt,sampler,steps,cfg_scale,ddim_eta,
         """
         uc, c, extra_conditioning_info = conditioning
 
-        trained_square = 512 * 512
-        actual_square = width * height
-        scale = math.sqrt(trained_square / actual_square)
-
-        init_width = math.ceil(scale * width / 64) * 64
-        init_height = math.ceil(scale * height / 64) * 64
-
         @torch.no_grad()
         def make_image(x_T):
 
+            scale_dim = min(width, height)
+            scale = 512 / scale_dim
+
+            init_width = math.ceil(scale * width / 64) * 64
+            init_height = math.ceil(scale * height / 64) * 64
+
             shape = [
                 self.latent_channels,
                 init_height // self.downsampling_factor,
                 init_width // self.downsampling_factor,
             ]
-            
+
             sampler.make_schedule(
                     ddim_num_steps=steps, ddim_eta=ddim_eta, verbose=False
             )
-            
+
             #x = self.get_noise(init_width, init_height)
             x = x_T
-            
+
             if self.free_gpu_mem and self.model.model.device != self.model.device:
                 self.model.model.to(self.model.device)
 
@@ -65,15 +64,15 @@ def make_image(x_T):
                 img_callback                 = step_callback,
                 extra_conditioning_info      = extra_conditioning_info
             )
-            
+
             print(
                   f"\n>> Interpolating from {init_width}x{init_height} to {width}x{height} using DDIM sampling"
                  )
-            
+
             # resizing
             samples = torch.nn.functional.interpolate(
-                samples, 
-                size=(height // self.downsampling_factor, width // self.downsampling_factor), 
+                samples,
+                size=(height // self.downsampling_factor, width // self.downsampling_factor),
                 mode="bilinear"
             )
 
@@ -162,7 +161,7 @@ def get_noise(self,width,height,scale = True):
         else:
             scaled_width = width
             scaled_height = height
-            
+
         device      = self.model.device
         if self.use_mps_noise or device.type == 'mps':
             return torch.randn([1,