pykeen · mberr · Jun 3, 2022 · Aug 27, 2022 · Aug 27, 2022 · Aug 29, 2022
diff --git a/docs/source/byo/data.rst b/docs/source/byo/data.rst
@@ -47,14 +47,14 @@ The remainder of the examples will be for :func:`pykeen.pipeline.pipeline`, but
 for :func:`pykeen.hpo.hpo_pipeline`.
 
 If you want to add dataset-wide arguments, you can use the ``dataset_kwargs`` argument
-to the :class:`pykeen.pipeline.pipeline` to enable options like ``create_inverse_triples=True``.
+to the :class:`pykeen.pipeline.pipeline` to enable options like ``use_inverse_relations=True``.
 
 >>> from pykeen.pipeline import pipeline
 >>> from pykeen.datasets.nations import NATIONS_TRAIN_PATH, NATIONS_TEST_PATH
 >>> result = pipeline(
 ...     training=NATIONS_TRAIN_PATH,
 ...     testing=NATIONS_TEST_PATH,
-...     dataset_kwargs={'create_inverse_triples': True},
+...     dataset_kwargs={'use_inverse_relations': True},
 ...     model='TransE',
 ...     epochs=5,  # short epochs for testing - you should go higher
 ... )
@@ -89,21 +89,21 @@ TSV files, you can use the :class:`pykeen.triples.TriplesFactory` interface.
     the wrong identifiers in the training set during evaluation, and we'd get nonsense results.
 
 The ``dataset_kwargs`` argument is ignored when passing your own :class:`pykeen.triples.TriplesFactory`, so be
-sure to include the ``create_inverse_triples=True`` in the instantiation of those classes if that's your
+sure to include the ``use_inverse_relations=True`` in the instantiation of those classes if that's your
 desired behavior as in:
 
 >>> from pykeen.triples import TriplesFactory
 >>> from pykeen.pipeline import pipeline
 >>> from pykeen.datasets.nations import NATIONS_TRAIN_PATH, NATIONS_TEST_PATH
 >>> training = TriplesFactory.from_path(
 ...     NATIONS_TRAIN_PATH,
-...     create_inverse_triples=True,
+...     use_inverse_relations=True,
 ... )
 >>> testing = TriplesFactory.from_path(
 ...     NATIONS_TEST_PATH,
 ...     entity_to_id=training.entity_to_id,
 ...     relation_to_id=training.relation_to_id,
-...     create_inverse_triples=True,
+...     use_inverse_relations=True,
 ... )
 >>> result = pipeline(
 ...     training=training,

diff --git a/docs/source/tutorial/inductive_lp.rst b/docs/source/tutorial/inductive_lp.rst
@@ -91,7 +91,7 @@ Let's create a basic `InductiveNodePiece` using one of the `InductiveFB15k237` d
     from pykeen.models.inductive import InductiveNodePiece
     from pykeen.losses import NSSALoss
 
-    dataset = InductiveFB15k237(version="v1", create_inverse_triples=True)
+    dataset = InductiveFB15k237(version="v1", use_inverse_relations=True)
 
     model = InductiveNodePiece(
         triples_factory=dataset.transductive_training,  # training factory, used to tokenize training nodes
@@ -110,7 +110,7 @@ Creating a message-passing version of NodePiece is pretty much the same:
     from pykeen.models.inductive import InductiveNodePieceGNN
     from pykeen.losses import NSSALoss
 
-    dataset = InductiveFB15k237(version="v1", create_inverse_triples=True)
+    dataset = InductiveFB15k237(version="v1", use_inverse_relations=True)
 
     model = InductiveNodePieceGNN(
         triples_factory=dataset.transductive_training,  # training factory, will be also used for a GNN
@@ -166,7 +166,7 @@ Let's create a training loop and validation / test evaluators:
     from pykeen.evaluation.rank_based_evaluator import SampledRankBasedEvaluator
     from pykeen.losses import NSSALoss
 
-    dataset = InductiveFB15k237(version="v1", create_inverse_triples=True)
+    dataset = InductiveFB15k237(version="v1", use_inverse_relations=True)
 
     model = ...  # model init here, one of InductiveNodePiece
     optimizer = ...  # some optimizer
@@ -207,7 +207,7 @@ in the sLCWA mode with 32 negative samples per positive, with NSSALoss, and Samp
 
     from torch.optim import Adam
 
-    dataset = InductiveFB15k237(version="v1", create_inverse_triples=True)
+    dataset = InductiveFB15k237(version="v1", use_inverse_relations=True)
 
     model = InductiveNodePieceGNN(
         triples_factory=dataset.transductive_training,  # training factory, will be also used for a GNN

diff --git a/docs/source/tutorial/node_piece.rst b/docs/source/tutorial/node_piece.rst
@@ -18,7 +18,7 @@ throughout the following examples.
    from pykeen.datasets import FB15k237
 
    # inverses are necessary for the current version of NodePiece
-   dataset = FB15k237(create_inverse_triples=True)
+   dataset = FB15k237(use_inverse_relations=True)
 
 In the simplest usage of :class:`pykeen.models.NodePiece`, we'll only
 use relations for tokenization. We can do this by with the following
@@ -286,7 +286,7 @@ Let's pack the last NodePiece model into the pipeline:
    result = pipeline(
        dataset="fb15k237",
        dataset_kwargs=dict(
-           create_inverse_triples=True,
+           use_inverse_relations=True,
        ),
        model=NodePiece,
        model_kwargs=dict(
@@ -498,7 +498,7 @@ pipeline:
    result = pipeline(
        dataset="fb15k237",
        dataset_kwargs=dict(
-           create_inverse_triples=True,
+           use_inverse_relations=True,
        ),
        model=NodePiece,
        model_kwargs=dict(
@@ -590,7 +590,7 @@ Let's use the new tokenizer for the Wikidata5M graph of 5M nodes and 20M edges.
 
     from pykeen.datasets import Wikidata5M
 
-    dataset = Wikidata5M(create_inverse_triples=True)
+    dataset = Wikidata5M(use_inverse_relations=True)
 
     model = NodePiece(
         triples_factory=dataset.training,

diff --git a/docs/source/tutorial/running_ablation.rst b/docs/source/tutorial/running_ablation.rst
@@ -68,7 +68,7 @@ as ``title`` are special and used by PyKEEN and :mod:`optuna`.
     ... )
 
 As mentioned above, we also want to measure the effect of explicitly modeling inverse relations on the model's
-performance. Therefore, we extend the ablation study by including the ``create_inverse_triples`` argument:
+performance. Therefore, we extend the ablation study by including the ``use_inverse_relations`` argument:
 
 .. code-block:: python
 
@@ -82,7 +82,7 @@ performance. Therefore, we extend the ablation study by including the ``create_i
     ...     training_loops=["LCWA"],
     ...     optimizers=["Adam"],
     ...     # Add inverse triples with
-    ...     create_inverse_triples=[True, False],
+    ...     use_inverse_relations=[True, False],
     ...     # Fast testing configuration, make bigger in prod
     ...     epochs=1,
     ...     n_trials=1,
@@ -91,10 +91,10 @@ performance. Therefore, we extend the ablation study by including the ``create_i
 .. note::
 
     Unlike ``models``, ``datasets``, ``losses``, ``training_loops``, and ``optimizers``,
-    ``create_inverse_triples`` has a default value, which is ``False``.
+    ``use_inverse_relations`` has a default value, which is ``False``.
 
 If there is only one value for either the ``models``, ``datasets``, ``losses``, ``training_loops``, ``optimizers``,
-or ``create_inverse_triples`` argument, it can be given as a single value instead of the list.
+or ``use_inverse_relations`` argument, it can be given as a single value instead of the list.
 
 .. code-block:: python
 
@@ -107,7 +107,7 @@ or ``create_inverse_triples`` argument, it can be given as a single value instea
     ...     losses=["BCEAfterSigmoidLoss", "MarginRankingLoss"],
     ...     training_loops="LCWA",
     ...     optimizers="Adam",
-    ...     create_inverse_triples=[True, False],
+    ...     use_inverse_relations=[True, False],
     ...     # Fast testing configuration, make bigger in prod
     ...     epochs=1,
     ...     n_trials=1,
@@ -200,7 +200,7 @@ the best model of each ablation-experiment using the argument ``best_replicates`
     ...     losses=["BCEAfterSigmoidLoss", "MarginRankingLoss"],
     ...     training_loops=["LCWA"],
     ...     optimizers=["Adam"],
-    ...     create_inverse_triples=[True, False],
+    ...     use_inverse_relations=[True, False],
     ...     stopper="early",
     ...     stopper_kwargs={
     ...         "frequency": 5,
@@ -384,7 +384,7 @@ Now that we defined our own hyper-parameter values/ranges, let's have a look at
     >>> losses = ["BCEAfterSigmoidLoss"]
     >>> training_loops = ["lcwa"]
     >>> optimizers = ["adam"]
-    >>> create_inverse_triples= [True, False]
+    >>> use_inverse_relations= [True, False]
     >>> stopper = "early"
     >>> stopper_kwargs = {
     ...    "frequency": 5,
@@ -513,7 +513,7 @@ defined within our program would look as follows:
             "losses": ["BCEAfterSigmoidLoss", "CrossEntropyLoss"]
             "training_loops": ["lcwa"],
             "optimizers": ["adam"],
-            "create_inverse_triples": [true,false],
+            "use_inverse_relations": [true,false],
             "stopper": "early"
             "stopper_kwargs": {
                 "frequency": 5,