fix import ordering, remove code I reverted that the resync added back

Author: Vargol

invokeai/backend/model_manager/load/model_cache/cached_model/cached_model_only_full_load.py

@@ -1,8 +1,9 @@
-from invokeai.backend.quantization.gguf.ggml_tensor import GGMLTensor
 from typing import Any
 
 import torch
 
+from invokeai.backend.quantization.gguf.ggml_tensor import GGMLTensor
+
 
 class CachedModelOnlyFullLoad:
     """A wrapper around a PyTorch model to handle full loads and unloads between the CPU and the compute device.
@@ -78,8 +79,7 @@ def full_load_to_vram(self) -> int:
                 new_state_dict[k] = v.to(self._compute_device, copy=True)
             self._model.load_state_dict(new_state_dict, assign=True)
 
-        
-        check_for_gguf = hasattr(self._model, 'state_dict') and self._model.state_dict().get("img_in.weight")
+        check_for_gguf = hasattr(self._model, "state_dict") and self._model.state_dict().get("img_in.weight")
         if isinstance(check_for_gguf, GGMLTensor):
             old_value = torch.__future__.get_overwrite_module_params_on_conversion()
             torch.__future__.set_overwrite_module_params_on_conversion(True)
@@ -103,7 +103,7 @@ def full_unload_from_vram(self) -> int:
         if self._cpu_state_dict is not None:
             self._model.load_state_dict(self._cpu_state_dict, assign=True)
 
-        check_for_gguf = hasattr(self._model, 'state_dict') and self._model.state_dict().get("img_in.weight")
+        check_for_gguf = hasattr(self._model, "state_dict") and self._model.state_dict().get("img_in.weight")
         if isinstance(check_for_gguf, GGMLTensor):
             old_value = torch.__future__.get_overwrite_module_params_on_conversion()
             torch.__future__.set_overwrite_module_params_on_conversion(True)

invokeai/backend/quantization/gguf/ggml_tensor.py

@@ -119,19 +119,6 @@ def size(self, dim: int | None = None):
             return self.tensor_shape[dim]
         return self.tensor_shape
 
-    @overload
-    def to(self, *args, **kwargs) -> torch.Tensor: ...
-
-    def to(self, *args, **kwargs):
-        for func_arg in args:
-            if isinstance(func_arg, torch.dtype) and func_arg != self.quantized_data.dtype:
-                raise ValueError("Operation changed the dtype of GGMLTensor unexpectedly.")
-        if "dtype" in kwargs.keys():
-            if kwargs["dtype"] != self.quantized_data.dtype:
-                raise ValueError("Operation changed the dtype of GGMLTensor unexpectedly.")
-        self.quantized_data = self.quantized_data.to(*args, **kwargs)
-        return self
-
     @property
     def shape(self) -> torch.Size:  # pyright: ignore[reportIncompatibleVariableOverride] pyright doesn't understand this for some reason.
         """The shape of the tensor after dequantization. I.e. the shape that will be used in any math ops."""