🎨 Refactor ModelABC to Help Use Default Torch Models

tests/models/test_arch_vanilla.py

@@ -4,7 +4,7 @@
 import pytest
 import torch
 
-from tiatoolbox.models.architecture.vanilla import CNNModel, TimmModel
+from tiatoolbox.models.architecture.vanilla import CNNModel, TimmModel, infer_batch
 from tiatoolbox.models.models_abc import model_to
 
 ON_GPU = False
@@ -45,7 +45,7 @@ def test_functional() -> None:
         for backbone in backbones:
             model = CNNModel(backbone, num_classes=1)
             model_ = model_to(device=device, model=model)
-            model.infer_batch(model_, samples, device=device)
+            infer_batch(model_, samples, device=device)
     except ValueError as exc:
         msg = f"Model {backbone} failed."
         raise AssertionError(msg) from exc
@@ -72,7 +72,7 @@ def test_timm_functional() -> None:
         for backbone in backbones:
             model = TimmModel(backbone=backbone, num_classes=1, pretrained=False)
             model_ = model_to(device=device, model=model)
-            model.infer_batch(model_, samples, device=device)
+            infer_batch(model_, samples, device=device)
     except ValueError as exc:
         msg = f"Model {backbone} failed."
         raise AssertionError(msg) from exc

tiatoolbox/models/architecture/vanilla.py

@@ -90,6 +90,40 @@ def _get_architecture(
     return model.features
 
 
+def infer_batch(
+    model: nn.Module,
+    batch_data: torch.Tensor,
+    *,
+    device: str = "cpu",
+) -> dict[str, np.ndarray]:
+    """Run inference on an input batch.
+
+    Contains logic for forward operation as well as i/o aggregation.
+
+    Args:
+        model (nn.Module):
+            PyTorch defined model.
+        batch_data (torch.Tensor):
+            A batch of data generated by
+            `torch.utils.data.DataLoader`.
+        device (str):
+            Transfers model to the specified device. Default is "cpu".
+
+    """
+    img_patches_device = batch_data.to(device).type(
+        torch.float32,
+    )  # to NCHW
+    img_patches_device = img_patches_device.permute(0, 3, 1, 2).contiguous()
+
+    # Inference mode
+    model.eval()
+    # Do not compute the gradient (not training)
+    with torch.inference_mode():
+        output = model(img_patches_device)
+    # Output should be a single tensor or scalar
+    return {"probabilities": output.cpu().numpy()}
+
+
 def _get_timm_architecture(
     arch_name: str,
     *,

tiatoolbox/models/engine/engine_abc.py

@@ -18,6 +18,7 @@
 from tiatoolbox import DuplicateFilter, logger, rcParam
 from tiatoolbox.models.architecture import get_pretrained_model
 from tiatoolbox.models.architecture.utils import compile_model
+from tiatoolbox.models.architecture.vanilla import infer_batch
 from tiatoolbox.models.dataset.dataset_abc import PatchDataset, WSIPatchDataset
 from tiatoolbox.models.models_abc import load_torch_model
 from tiatoolbox.utils.misc import (
@@ -560,7 +561,7 @@ def infer_patches(
             zarr_group = zarr.open(save_path, mode="w")
 
         for _, batch_data in enumerate(dataloader):
-            batch_output = self.model.infer_batch(
+            batch_output = infer_batch(
                 self.model,
                 batch_data["image"],
                 device=self.device,

tiatoolbox/models/models_abc.py

@@ -101,30 +101,6 @@ def forward(
         """Torch method, this contains logic for using layers defined in init."""
         ...  # pragma: no cover
 
-    @staticmethod
-    @abstractmethod
-    def infer_batch(model: nn.Module, batch_data: np.ndarray, *, device: str) -> dict:
-        """Run inference on an input batch.
-
-        Contains logic for forward operation as well as I/O aggregation.
-
-        Args:
-            model (nn.Module):
-                PyTorch defined model.
-            batch_data (np.ndarray):
-                A batch of data generated by
-                `torch.utils.data.DataLoader`.
-            device (str):
-                Transfers model to the specified device. Default is "cpu".
-
-        Returns:
-            dict:
-                Returns a dictionary of predictions and other expected outputs
-                depending on the network architecture.
-
-        """
-        ...  # pragma: no cover
-
     @staticmethod
     def preproc(image: np.ndarray) -> np.ndarray:
         """Define the pre-processing of this class of model."""