[data.montgomery.loader] Use load_pil() instead of load_pil_baw() since...

[data.montgomery.loader] Use load_pil() instead of load_pil_baw() since Montgomery images are already in grayscale PNG (not black-and-white!)

src/ptbench/data/montgomery/loader.py

@@ -10,11 +10,11 @@ This configuration:
 
 * Raw data input (on disk):
 
-    * PNG images 12 bit grayscale
+    * PNG images 8 bit grayscale
     * resolution: 4020 x 4892 px or 4892 x 4020 px
 
 * Output image:
-    
+
     * Transforms:
 
         * Load raw PNG with :py:mod:`PIL`
@@ -25,7 +25,7 @@ This configuration:
     * Final specifications
 
         * Fixed resolution: 512 x 512 px
-        * Color RGB encoding
+        * Grayscale (single channel), 8 bits
 """
 
 import os
@@ -33,7 +33,7 @@ import os
 import torchvision.transforms
 
 from ...utils.rc import load_rc
-from ..image_utils import RemoveBlackBorders, load_pil_baw
+from ..image_utils import RemoveBlackBorders, load_pil
 from ..typing import RawDataLoader as _BaseRawDataLoader
 from ..typing import Sample
 
@@ -87,9 +87,7 @@ class RawDataLoader(_BaseRawDataLoader):
         sample
             The sample representation
         """
-        tensor = self.transform(
-            load_pil_baw(os.path.join(self.datadir, sample[0]))
-        )
+        tensor = self.transform(load_pil(os.path.join(self.datadir, sample[0])))
 
         return tensor, dict(label=sample[1], name=sample[0])  # type: ignore[arg-type]