migrate pydantic (#295)

* migrate pydantic 

* abstract n_folds

* tell bandit to ignore torch save/load warnings

* Update notebook tests

.github/workflows/test_tutorials.yml

@@ -40,4 +40,4 @@ jobs:
           python -m pip install ipykernel
           python -m ipykernel install --user
       - name: Run the tutorials
-        run: python tests/nb_eval.py --nb_dir tutorials/ --tutorial_tests minimal_tests
+        run: python tests/nb_eval.py --nb_dir tutorials/ --tutorial_tests minimal_tests --timeout 3600

setup.cfg

@@ -47,7 +47,7 @@ install_requires =
     tenacity
     tqdm
     loguru
-    pydantic<2.0
+    pydantic
     cloudpickle
     scipy
     xgboost<3.0.0

src/synthcity/benchmark/__init__.py

@@ -57,6 +57,7 @@ def evaluate(
         strict_augmentation: bool = False,
         ad_hoc_augment_vals: Optional[Dict] = None,
         use_metric_cache: bool = True,
+        n_eval_folds: int = 5,
         **generate_kwargs: Any,
     ) -> pd.DataFrame:
         """Benchmark the performance of several algorithms.
@@ -102,6 +103,8 @@ def evaluate(
                 A dictionary containing the number of each class to augment the real data with. This is only required if using the rule="ad-hoc" option. Defaults to None.
             use_metric_cache: bool
                 If the current metric has been previously run and is cached, it will be reused for the experiments. Defaults to True.
+            n_eval_folds: int
+                the KFolds used by MetricEvaluators in the benchmarks. Defaults to 5.
             plugin_kwargs:
                 Optional kwargs for each algorithm. Example {"adsgan": {"n_iter": 10}},
         """
@@ -295,6 +298,7 @@ def evaluate(
                     task_type=task_type,
                     workspace=workspace,
                     use_cache=use_metric_cache,
+                    n_folds=n_eval_folds,
                 )
 
                 mean_score = evaluation["mean"].to_dict()

src/synthcity/metrics/_utils.py

@@ -332,7 +332,7 @@ def f() -> None:
                     "epoch": epoch,
                 },
                 workspace / "DomiasMIA_bnaf_checkpoint.pt",
-            )
+            )  # nosec B614
 
     return f
 
@@ -348,7 +348,7 @@ def f() -> None:
 
         log.info("Loading model..")
         if (workspace / "checkpoint.pt").exists():
-            checkpoint = torch.load(workspace / "checkpoint.pt")
+            checkpoint = torch.load(workspace / "checkpoint.pt")  # nosec B614
             model.load_state_dict(checkpoint["model"])
             optimizer.load_state_dict(checkpoint["optimizer"])
 
@@ -453,7 +453,7 @@ def train(
                 "epoch": epoch,
             },
             workspace / "checkpoint.pt",
-        )
+        )  # nosec B614
         log.debug(
             f"""
 ###### Stop training after {epoch + 1} epochs!

src/synthcity/metrics/eval.py

@@ -119,6 +119,7 @@ def evaluate(
         random_state: int = 0,
         workspace: Path = Path("workspace"),
         use_cache: bool = True,
+        n_folds: int = 5,
     ) -> pd.DataFrame:
         """Core evaluation logic for the metrics
 
@@ -238,6 +239,7 @@ def evaluate(
                         random_state=random_state,
                         workspace=workspace,
                         use_cache=use_cache,
+                        n_folds=n_folds,
                     ),
                     X_gt,
                     X_augmented,
@@ -251,6 +253,7 @@ def evaluate(
                         random_state=random_state,
                         workspace=workspace,
                         use_cache=use_cache,
+                        n_folds=n_folds,
                     ),
                     X_gt,
                     X_syn,
@@ -267,6 +270,7 @@ def evaluate(
                         random_state=random_state,
                         workspace=workspace,
                         use_cache=use_cache,
+                        n_folds=n_folds,
                     ),
                     X_gt.sample(eval_cnt),
                     X_syn.sample(eval_cnt),

src/synthcity/plugins/core/constraints.py

@@ -4,11 +4,13 @@
 # third party
 import numpy as np
 import pandas as pd
-from pydantic import BaseModel, validate_arguments, validator
+from pydantic import BaseModel, field_validator, validate_arguments
 
 # synthcity absolute
 import synthcity.logger as log
 
+Rule = Tuple[str, str, Any]  # Define a type alias for clarity
+
 
 class Constraints(BaseModel):
     """
@@ -41,10 +43,10 @@ class Constraints(BaseModel):
             and thresh is the threshold or data type.
     """
 
-    rules: list = []
+    rules: list[Rule] = []
 
-    @validator("rules")
-    def _validate_rules(cls: Any, rules: List, values: dict, **kwargs: Any) -> List:
+    @field_validator("rules", mode="before")
+    def _validate_rules(cls: Any, rules: List) -> List:
         supported_ops: list = [
             "<",
             ">=",