use heuristic to select RAM cache size during headless install; blackified

installer/lib/installer.py

@@ -344,12 +344,12 @@ class InvokeAiInstance:
                 auto_install = True
         sys.argv = new_argv
 
-        import requests  # to catch download exceptions
         import messages
+        import requests  # to catch download exceptions
 
         auto_install = auto_install or messages.user_wants_auto_configuration()
         if auto_install:
-            sys.argv.append('--yes')
+            sys.argv.append("--yes")
         else:
             messages.introduction()
 

installer/lib/messages.py

@@ -7,7 +7,7 @@ import os
 import platform
 from pathlib import Path
 
-from prompt_toolkit import prompt, HTML
+from prompt_toolkit import HTML, prompt
 from prompt_toolkit.completion import PathCompleter
 from prompt_toolkit.validation import Validator
 from rich import box, print
@@ -97,13 +97,17 @@ def user_wants_auto_configuration() -> bool:
             padding=(1, 1),
         )
     )
-    choice = prompt(HTML("Choose <b>&lt;a&gt;</b>utomatic or <b>&lt;m&gt;</b>anual configuration [a/m] (a): "),
+    choice = (
+        prompt(
+            HTML("Choose <b>&lt;a&gt;</b>utomatic or <b>&lt;m&gt;</b>anual configuration [a/m] (a): "),
             validator=Validator.from_callable(
-                        lambda n: n=='' or n.startswith(('a', 'A', 'm', 'M')),
+                lambda n: n == "" or n.startswith(("a", "A", "m", "M")), error_message="Please select 'a' or 'm'"
-                        error_message="Please select 'a' or 'm'"
             ),
-                ) or 'a'
+        )
-    return choice.lower().startswith('a')
+        or "a"
+    )
+    return choice.lower().startswith("a")
+
 
 def dest_path(dest=None) -> Path:
     """

invokeai/app/invocations/math.py

@@ -199,7 +199,7 @@ class IntegerMathInvocation(BaseInvocation):
         elif self.operation == "DIV":
             return IntegerOutput(value=int(self.a / self.b))
         elif self.operation == "EXP":
-            return IntegerOutput(value=self.a**self.b)
+            return IntegerOutput(value=self.a ** self.b)
         elif self.operation == "MOD":
             return IntegerOutput(value=self.a % self.b)
         elif self.operation == "ABS":
@@ -273,7 +273,7 @@ class FloatMathInvocation(BaseInvocation):
         elif self.operation == "DIV":
             return FloatOutput(value=self.a / self.b)
         elif self.operation == "EXP":
-            return FloatOutput(value=self.a**self.b)
+            return FloatOutput(value=self.a ** self.b)
         elif self.operation == "SQRT":
             return FloatOutput(value=np.sqrt(self.a))
         elif self.operation == "ABS":

invokeai/backend/install/invokeai_configure.py

@@ -70,7 +70,6 @@ def get_literal_fields(field) -> list[Any]:
 config = InvokeAIAppConfig.get_config()
 
 Model_dir = "models"
-
 Default_config_file = config.model_conf_path
 SD_Configs = config.legacy_conf_path
 
@@ -458,7 +457,7 @@ Use cursor arrows to make a checkbox selection, and space to toggle.
         )
         self.add_widget_intelligent(
             npyscreen.TitleFixedText,
-            name="Model RAM cache size (GB). Make this at least large enough to hold a single full model.",
+            name="Model RAM cache size (GB). Make this at least large enough to hold a single full model (2GB for SD-1, 6GB for SDXL).",
             begin_entry_at=0,
             editable=False,
             color="CONTROL",
@@ -651,8 +650,19 @@ def edit_opts(program_opts: Namespace, invokeai_opts: Namespace) -> argparse.Nam
     return editApp.new_opts()
 
 
+def default_ramcache() -> float:
+    """Run a heuristic for the default RAM cache based on installed RAM."""
+
+    # Note that on my 64 GB machine, psutil.virtual_memory().total gives 62 GB,
+    # So we adjust everthing down a bit.
+    return (
+        15.0 if MAX_RAM >= 60 else 7.5 if MAX_RAM >= 30 else 4 if MAX_RAM >= 14 else 2.1
+    )  # 2.1 is just large enough for sd 1.5 ;-)
+
+
 def default_startup_options(init_file: Path) -> Namespace:
     opts = InvokeAIAppConfig.get_config()
+    opts.ram = default_ramcache()
     return opts
 
 

invokeai/backend/ip_adapter/resampler.py

@@ -33,7 +33,7 @@ def reshape_tensor(x, heads):
 class PerceiverAttention(nn.Module):
     def __init__(self, *, dim, dim_head=64, heads=8):
         super().__init__()
-        self.scale = dim_head**-0.5
+        self.scale = dim_head ** -0.5
         self.dim_head = dim_head
         self.heads = heads
         inner_dim = dim_head * heads
@@ -91,7 +91,7 @@ class Resampler(nn.Module):
     ):
         super().__init__()
 
-        self.latents = nn.Parameter(torch.randn(1, num_queries, dim) / dim**0.5)
+        self.latents = nn.Parameter(torch.randn(1, num_queries, dim) / dim ** 0.5)
 
         self.proj_in = nn.Linear(embedding_dim, dim)
 

invokeai/backend/stable_diffusion/diffusion/cross_attention_control.py

@@ -261,7 +261,7 @@ class InvokeAICrossAttentionMixin:
         if q.shape[1] <= 4096:  # (512x512) max q.shape[1]: 4096
             return self.einsum_lowest_level(q, k, v, None, None, None)
         else:
-            slice_size = math.floor(2**30 / (q.shape[0] * q.shape[1]))
+            slice_size = math.floor(2 ** 30 / (q.shape[0] * q.shape[1]))
             return self.einsum_op_slice_dim1(q, k, v, slice_size)
 
     def einsum_op_mps_v2(self, q, k, v):

invokeai/backend/stable_diffusion/diffusion/shared_invokeai_diffusion.py

@@ -175,10 +175,7 @@ class InvokeAIDiffuserComponent:
                                 dim=0,
                             ),
                         }
-                    (
+                    (encoder_hidden_states, encoder_attention_mask,) = self._concat_conditionings_for_batch(
-                        encoder_hidden_states,
-                        encoder_attention_mask,
-                    ) = self._concat_conditionings_for_batch(
                         conditioning_data.unconditioned_embeddings.embeds,
                         conditioning_data.text_embeddings.embeds,
                     )
@@ -240,10 +237,7 @@ class InvokeAIDiffuserComponent:
         wants_cross_attention_control = len(cross_attention_control_types_to_do) > 0
 
         if wants_cross_attention_control:
-            (
+            (unconditioned_next_x, conditioned_next_x,) = self._apply_cross_attention_controlled_conditioning(
-                unconditioned_next_x,
-                conditioned_next_x,
-            ) = self._apply_cross_attention_controlled_conditioning(
                 sample,
                 timestep,
                 conditioning_data,
@@ -251,10 +245,7 @@ class InvokeAIDiffuserComponent:
                 **kwargs,
             )
         elif self.sequential_guidance:
-            (
+            (unconditioned_next_x, conditioned_next_x,) = self._apply_standard_conditioning_sequentially(
-                unconditioned_next_x,
-                conditioned_next_x,
-            ) = self._apply_standard_conditioning_sequentially(
                 sample,
                 timestep,
                 conditioning_data,
@@ -262,10 +253,7 @@ class InvokeAIDiffuserComponent:
             )
 
         else:
-            (
+            (unconditioned_next_x, conditioned_next_x,) = self._apply_standard_conditioning(
-                unconditioned_next_x,
-                conditioned_next_x,
-            ) = self._apply_standard_conditioning(
                 sample,
                 timestep,
                 conditioning_data,

invokeai/backend/training/textual_inversion_training.py

@@ -470,10 +470,7 @@ class TextualInversionDataset(Dataset):
 
         if self.center_crop:
             crop = min(img.shape[0], img.shape[1])
-            (
+            (h, w,) = (
-                h,
-                w,
-            ) = (
                 img.shape[0],
                 img.shape[1],
             )

invokeai/backend/util/mps_fixes.py

@@ -203,7 +203,7 @@ class ChunkedSlicedAttnProcessor:
         if attn.upcast_attention:
             out_item_size = 4
 
-        chunk_size = 2**29
+        chunk_size = 2 ** 29
 
         out_size = query.shape[1] * key.shape[1] * out_item_size
         chunks_count = min(query.shape[1], math.ceil((out_size - 1) / chunk_size))

invokeai/backend/util/util.py

@@ -207,7 +207,7 @@ def parallel_data_prefetch(
         return gather_res
 
 
-def rand_perlin_2d(shape, res, device, fade=lambda t: 6 * t**5 - 15 * t**4 + 10 * t**3):
+def rand_perlin_2d(shape, res, device, fade=lambda t: 6 * t ** 5 - 15 * t ** 4 + 10 * t ** 3):
     delta = (res[0] / shape[0], res[1] / shape[1])
     d = (shape[0] // res[0], shape[1] // res[1])