(43909856) ViT (Visual Transformer) for image classification of ADNI dataset

.gitignore

@@ -1,2 +1,2 @@
 
-*.pyc
+*.pyc

recognition/TRANSFORMER_43909856/dataset/README.md

@@ -0,0 +1,22 @@
+# dataset:
+
+The ADNI dataset should be placed in this directory.
+
+To ensure that the model and dataset loading works correctly, the dataset should
+be added such that it conforms to the following directory structure:
+
+```
+    - dataset/:
+        - 'test/'
+            - 'AD/
+                - ...
+            - 'NC/'
+                - ...
+        - 'train/'
+            - 'AD/'
+                - ...
+            - 'NC/'
+                - ...
+```
+
+'...' represents the ADNI image JPEG files.

recognition/TRANSFORMER_43909856/models/README.md

@@ -0,0 +1,9 @@
+# models
+
+After model training has completed, the trained model will be saved to this directory location.
+
+This location will also contain saved metrics for the train, validation, and test set performance.
+
+These metrics include the train and validation loss throughout the training process, 
+the validation set accuracy, and the predictions and observed classes for the validation
+and test set.