diff --git a/examples/multi_gpu/pcqm4m_ogb.py b/examples/multi_gpu/pcqm4m_ogb.py
index 10e190251699..2bce247e5d0b 100644
--- a/examples/multi_gpu/pcqm4m_ogb.py
+++ b/examples/multi_gpu/pcqm4m_ogb.py
@@ -117,7 +117,7 @@ def __init__(self, num_layers, emb_dim, drop_ratio=0.5, JK="last",
             num_layers (int): number of GNN message passing layers.
             residual (bool): whether to add residual connection.
             drop_ratio (float): dropout ratio.
-            JK (str): "last" or "sum" to choose JK concat strat.
+            JK (str): "last" or "sum" to choose JK concat strata.
             residual (bool): Whether or not to add the residual
             gnn_type (str): Type of GNN to use.
         """
@@ -302,8 +302,8 @@ def __init__(
             virtual_node (bool): whether to add virtual node or not.
             residual (bool): Whether or not to add the residual
             drop_ratio (float): dropout ratio.
-            JK (str): "last" or "sum" to choose JK concat strat.
-            graph_pooling (str): Graph pooling strat to use.
+            JK (str): "last" or "sum" to choose JK concat strata.
+            graph_pooling (str): Graph pooling strata to use.
         """
         super().__init__()
         if num_layers < 2:
diff --git a/examples/randlanet_classification.py b/examples/randlanet_classification.py
index 213bf957bbdb..0d814974e3a8 100644
--- a/examples/randlanet_classification.py
+++ b/examples/randlanet_classification.py
@@ -62,9 +62,9 @@ def message(self, x_j: Tensor, pos_i: Tensor, pos_j: Tensor,
         """Local Spatial Encoding (locSE) and attentive pooling of features.
 
         Args:
-            x_j (Tensor): neighboors features (K,d)
+            x_j (Tensor): neighbors features (K,d)
             pos_i (Tensor): centroid position (repeated) (K,3)
-            pos_j (Tensor): neighboors positions (K,3)
+            pos_j (Tensor): neighbors positions (K,3)
             index (Tensor): index of centroid positions
                 (e.g. [0,...,0,1,...,1,...,N,...,N])
 
diff --git a/graphgym/custom_graphgym/optimizer/example.py b/graphgym/custom_graphgym/optimizer/example.py
index 0b2c88a4c86b..dd5c869755b5 100644
--- a/graphgym/custom_graphgym/optimizer/example.py
+++ b/graphgym/custom_graphgym/optimizer/example.py
@@ -13,7 +13,7 @@ def adagrad_optimizer(params: Iterator[Parameter], base_lr: float,
     return Adagrad(params, lr=base_lr, weight_decay=weight_decay)
 
 
-@register.register_scheduler('pleateau')
+@register.register_scheduler('plateau')
 def plateau_scheduler(optimizer: Optimizer, patience: int,
                       lr_decay: float) -> ReduceLROnPlateau:
     return ReduceLROnPlateau(optimizer, patience=patience, factor=lr_decay)
diff --git a/torch_geometric/data/lightning/datamodule.py b/torch_geometric/data/lightning/datamodule.py
index 7f0783d07398..499a6d52e981 100644
--- a/torch_geometric/data/lightning/datamodule.py
+++ b/torch_geometric/data/lightning/datamodule.py
@@ -124,7 +124,7 @@ def __init__(
 
         if loader in ['neighbor', 'link_neighbor']:
 
-            # Define a new `NeighborSampler` to be re-used across data loaders:
+            # Define a new `NeighborSampler` to be reused across data loaders:
             sampler_kwargs, self.loader_kwargs = split_kwargs(
                 self.kwargs,
                 NeighborSampler,
diff --git a/torch_geometric/graphgym/models/gnn.py b/torch_geometric/graphgym/models/gnn.py
index a011b149523d..8048f3f32dd5 100644
--- a/torch_geometric/graphgym/models/gnn.py
+++ b/torch_geometric/graphgym/models/gnn.py
@@ -119,7 +119,7 @@ def __init__(self, dim_in: int):
                         has_bias=False,
                         cfg=cfg,
                     ))
-            # Update `dim_in` to reflect the new dimension fo the node features
+            # Update `dim_in` to reflect the new dimension of the node features
             self.dim_in = cfg.gnn.dim_inner
         if cfg.dataset.edge_encoder:
             # Encode integer edge features via `torch.nn.Embedding`: