diff --git a/graph_weather/models/gencast/graph/graph_builder.py b/graph_weather/models/gencast/graph/graph_builder.py
index 68b57678..8a48aeb0 100644
--- a/graph_weather/models/gencast/graph/graph_builder.py
+++ b/graph_weather/models/gencast/graph/graph_builder.py
@@ -9,23 +9,37 @@
 """
 
 import gc
+from dataclasses import dataclass
 
 import numpy as np
 import torch
+from dacite import from_dict
 from torch_geometric.data import Data, HeteroData
 
 # from torch_geometric.transforms import TwoHop
 from graph_weather.models.gencast.graph import grid_mesh_connectivity, icosahedral_mesh, model_utils
 
-# Some configs from graphcast:
-_spatial_features_kwargs = dict(
-    add_node_positions=False,
-    add_node_latitude=True,
-    add_node_longitude=True,
-    add_relative_positions=True,
-    relative_longitude_local_coordinates=True,
-    relative_latitude_local_coordinates=True,
-)
+
+@dataclass
+class SpatialFeaturesConfig:
+    add_node_positions: bool = False
+    add_node_latitude: bool = True
+    add_node_longitude: bool = True
+    add_relative_positions: bool = True
+    relative_longitude_local_coordinates: bool = True
+    relative_latitude_local_coordinates: bool = True
+
+
+config_dict = {
+    "add_node_positions": False,
+    "add_node_latitude": True,
+    "add_node_longitude": True,
+    "add_relative_positions": True,
+    "relative_longitude_local_coordinates": True,
+    "relative_latitude_local_coordinates": True,
+}
+
+_spatial_features_kwargs = from_dict(data_class=SpatialFeaturesConfig, data=config_dict)
 
 # radius_query_fraction_edge_length: Scalar that will be multiplied by the
 #   length of the longest edge of the finest mesh to define the radius of
diff --git a/graph_weather/models/gencast/train.py b/graph_weather/models/gencast/train.py
index a6bdb127..0526c36c 100644
--- a/graph_weather/models/gencast/train.py
+++ b/graph_weather/models/gencast/train.py
@@ -7,10 +7,14 @@
 os.environ["CUDA_DEVICE_ORDER"] = "PCI_BUS_ID"
 os.environ["CUDA_VISIBLE_DEVICES"] = "0,3"
 
+from dataclasses import dataclass  # noqa: E402
+from typing import List  # noqa: E402
+
 import lightning as L  # noqa: E402
 import matplotlib.pyplot as plt  # noqa: E402
 import numpy as np  # noqa: E402
 import torch  # noqa: E402
+from dacite import from_dict  # noqa: E402
 from lightning.pytorch.callbacks import Callback, LearningRateMonitor, ModelCheckpoint  # noqa: E402
 from lightning.pytorch.loggers import WandbLogger  # noqa: E402
 from torch.utils.data import DataLoader  # noqa: E402
@@ -37,7 +41,21 @@
 
 # model configs
 CHECKPOINT_PATH = "checkpoints/epoch=3-step=10776.ckpt"
-CFG = {
+
+
+@dataclass
+class Config:
+    hidden_dims: List[int]
+    num_blocks: int
+    num_heads: int
+    splits: int
+    num_hops: int
+    sparse: bool
+    use_edges_features: bool
+    scale_factor: float
+
+
+config = {
     "hidden_dims": [512, 512],
     "num_blocks": 16,
     "num_heads": 4,
@@ -47,6 +65,7 @@
     "use_edges_features": False,
     "scale_factor": 1.0,
 }
+CFG = from_dict(data_class=Config, data=config)
 
 # dataset configs
 atmospheric_features = [
diff --git a/requirements.txt b/requirements.txt
index 79d45e1b..7612a6a6 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -10,3 +10,4 @@ xarray
 setuptools
 pydantic
 safetensors
+dacite