feat(nodes): add resize and scale latents nodes

- this resize/scale latents is what is needed for hires fix
- also remove unused `seed` from t2l

invokeai/app/invocations/latent.py

@@ -146,7 +146,6 @@ class TextToLatentsInvocation(BaseInvocation):
     # TODO: consider making prompt optional to enable providing prompt through a link
     # fmt: off
     prompt: Optional[str] = Field(description="The prompt to generate an image from")
-    seed:        int = Field(default=-1,ge=-1, le=np.iinfo(np.uint32).max, description="The seed to use (-1 for a random seed)", )
     noise: Optional[LatentsField] = Field(description="The noise to use")
     steps:       int = Field(default=10, gt=0, description="The number of steps to use to generate the image")
     width:       int = Field(default=512, multiple_of=64, gt=0, description="The width of the resulting image", )
@@ -363,9 +362,87 @@ class LatentsToImageInvocation(BaseInvocation):
                 session_id=context.graph_execution_state_id, node=self
             )
 
+            torch.cuda.empty_cache()
+
             context.services.images.save(image_type, image_name, image, metadata)
             return build_image_output(
-                image_type=image_type,
-                image_name=image_name,
-                image=image
+                image_type=image_type, image_name=image_name, image=image
             )
+
+
+LATENTS_INTERPOLATION_MODE = Literal[
+    "nearest", "linear", "bilinear", "bicubic", "trilinear", "area", "nearest-exact"
+]
+
+
+class ResizeLatentsInvocation(BaseInvocation):
+    """Resizes latents to explicit width/height."""
+
+    type: Literal["lresize"] = "lresize"
+
+    # Inputs
+    latents: Optional[LatentsField] = Field(description="The latents to resize")
+    width: int = Field(ge=64, multiple_of=8, description="The width to resize to")
+    height: int = Field(ge=64, multiple_of=8, description="The height to resize to")
+    downsample: int = Field(
+        default=8, ge=1, description="The downsampling factor (leave at 8 for SD)"
+    )
+    mode: LATENTS_INTERPOLATION_MODE = Field(
+        default="bilinear", description="The interpolation mode"
+    )
+
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        latents = context.services.latents.get(self.latents.latents_name)
+        # resizing
+        resized_latents = torch.nn.functional.interpolate(
+            latents,
+            size=(
+                self.height // self.downsample,
+                self.width // self.downsample,
+            ),
+            mode=self.mode,
+        )
+
+        # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        torch.cuda.empty_cache()
+
+        name = f"{context.graph_execution_state_id}__{self.id}"
+        context.services.latents.set(name, resized_latents)
+        return LatentsOutput(latents=LatentsField(latents_name=name))
+
+
+class ScaleLatentsInvocation(BaseInvocation):
+    """Scales latents by a given factor."""
+
+    type: Literal["lscale"] = "lscale"
+
+    # Inputs
+    latents: Optional[LatentsField] = Field(description="The latents to resize")
+    scale: int = Field(
+        default=2, ge=1, description="The factor by which to scale the latents"
+    )
+    mode: LATENTS_INTERPOLATION_MODE = Field(
+        default="bilinear", description="The interpolation mode"
+    )
+
+    def invoke(self, context: InvocationContext) -> LatentsOutput:
+        latents = context.services.latents.get(self.latents.latents_name)
+
+        (_, _, h, w) = latents.size()
+
+        # resizing
+        resized_latents = torch.nn.functional.interpolate(
+            latents,
+            size=(
+                h * self.scale,
+                w * self.scale,
+            ),
+            mode=self.mode,
+        )
+
+        # https://discuss.huggingface.co/t/memory-usage-by-later-pipeline-stages/23699
+        torch.cuda.empty_cache()
+
+        name = f"{context.graph_execution_state_id}__{self.id}"
+        context.services.latents.set(name, resized_latents)
+        return LatentsOutput(latents=LatentsField(latents_name=name))