Add GNN-Based Predictor with DAG Preprocessing

pyproject.toml

@@ -44,6 +44,10 @@ dependencies = [
     "numpy>=1.26; python_version >= '3.12'",
     "numpy>=1.24; python_version >= '3.11'",
     "numpy>=1.22",
+    "optuna>=4.0.0",
+    "torch-geometric>=2.7.0,<2.8.0",
+    "torch>=2.7.0,<2.8.0; sys_platform == 'darwin' and 'x86_64' in platform_machine and python_version < '3.13'",  # Restrict torch v2.3.0 for macOS x86 since it is not supported anymore.
+    "safetensors>=0.3.0",
     "typing-extensions>=4.1",  # for `assert_never`
 ]
 
@@ -116,6 +120,8 @@ log_level = "INFO"
 filterwarnings = [
     'error',
     'ignore:.*pytorch.*:UserWarning:',
+    "ignore:`torch_geometric.distributed` has been deprecated:DeprecationWarning:torch_geometric",
+    "ignore:.*'type_params' parameter of 'typing\\._eval_type'.*:DeprecationWarning:",
     'ignore:.*Values in x.*:RuntimeWarning:',
     'ignore:.*The least populated class in y has only 3 members, which is less than n_splits=5.*:UserWarning:',
     'ignore:.*divide by zero encountered in det.*:RuntimeWarning:',
@@ -162,9 +168,15 @@ implicit_reexport = true
 # recent versions of `gym` are typed, but stable-baselines3 pins a very old version of gym.
 # qiskit is not yet marked as typed, but is typed mostly.
 # the other libraries do not have type stubs.
-module = ["qiskit.*", "joblib.*", "sklearn.*", "matplotlib.*", "gymnasium.*", "mqt.bench.*", "sb3_contrib.*", "bqskit.*", "qiskit_ibm_runtime.*", "networkx.*", "stable_baselines3.*"]
+module = ["qiskit.*", "joblib.*", "sklearn.*", "matplotlib.*", "gymnasium.*", "mqt.bench.*", "sb3_contrib.*", "bqskit.*", "qiskit_ibm_runtime.*", "networkx.*", "stable_baselines3.*", "torch", "torch.*", "torch_geometric", "torch_geometric.*", "optuna.*", "safetensors", "safetensors.torch", "safetensors.*" ]
 ignore_missing_imports = true
 
+# Relax type checking for the GNN module since torch and torch_geometric expose
+# many APIs as Any due to incomplete type stubs; the GNN implementation must
+# work with these untyped returns and may need to subclass Any-typed classes.
+[[tool.mypy.overrides]]
+module = ["mqt.predictor.ml.gnn"]
+disallow_subclassing_any = false
 
 [tool.ruff]
 line-length = 120
@@ -252,6 +264,7 @@ wille = "wille"
 anc = "anc"
 aer = "aer"
 fom = "fom"
+TPE = "TPE"
 
 
 [tool.repo-review]

src/mqt/predictor/_version.py

@@ -0,0 +1,39 @@
+# Copyright (c) 2023 - 2025 Chair for Design Automation, TUM
+# Copyright (c) 2025 Munich Quantum Software Company GmbH
+# All rights reserved.
+#
+# SPDX-License-Identifier: MIT
+#
+# Licensed under the MIT License
+
+# file generated by setuptools-scm
+# don't change, don't track in version control
+
+__all__ = [
+    "__commit_id__",
+    "__version__",
+    "__version_tuple__",
+    "commit_id",
+    "version",
+    "version_tuple",
+]
+
+TYPE_CHECKING = False
+if TYPE_CHECKING:
+    VERSION_TUPLE = tuple[int | str, ...]
+    COMMIT_ID = str | None
+else:
+    VERSION_TUPLE = object
+    COMMIT_ID = object
+
+version: str
+__version__: str
+__version_tuple__: VERSION_TUPLE
+version_tuple: VERSION_TUPLE
+commit_id: COMMIT_ID
+__commit_id__: COMMIT_ID
+
+__version__ = version = "0.1.dev957+g1f2bf74aa.d20260107"
+__version_tuple__ = version_tuple = (0, 1, "dev957", "g1f2bf74aa.d20260107")
+
+__commit_id__ = commit_id = None

src/mqt/predictor/hellinger/utils.py

@@ -132,12 +132,13 @@ def calc_device_specific_features(
     return np.array(list(feature_dict.values()))
 
 
-def get_hellinger_model_path(device: Target) -> Path:
-    """Returns the path to the trained model folder resulting from the machine learning training."""
-    training_data_path = Path(str(resources.files("mqt.predictor"))) / "ml" / "training_data"
-    model_path = (
-        training_data_path
-        / "trained_model"
-        / ("trained_hellinger_distance_regressor_" + device.description + ".joblib")
+def get_hellinger_model_path(device: Target, gnn: bool = False) -> Path:
+    """Returns the path to the trained model file resulting from the machine learning training."""
+    training_data_path = Path(str(resources.files("mqt.predictor"))) / "ml" / "training_data" / "trained_model"
+    device_description = str(device.description)
+    filename = (
+        (f"trained_hellinger_distance_regressor_gnn_{device_description}.pth")
+        if gnn
+        else (f"trained_hellinger_distance_regressor_{device_description}.joblib")
     )
-    return Path(model_path)
+    return training_data_path / filename

src/mqt/predictor/ml/__init__.py

@@ -13,4 +13,9 @@
 from mqt.predictor.ml import helper
 from mqt.predictor.ml.predictor import Predictor, predict_device_for_figure_of_merit, setup_device_predictor
 
-__all__ = ["Predictor", "helper", "predict_device_for_figure_of_merit", "setup_device_predictor"]
+__all__ = [
+    "Predictor",
+    "helper",
+    "predict_device_for_figure_of_merit",
+    "setup_device_predictor",
+]