Add EMNIST VIT model with metrics

README.md

@@ -0,0 +1,18 @@
+
+# EMNIST ResNet Model
+
+This model is a ResNet-based CNN trained on the EMNIST Balanced dataset.
+
+## Evaluation Metrics
+
+- Test Accuracy: 88.47%
+- Test Loss: 0.3291
+
+## Training Details
+
+- Epochs: 7
+- Best Validation Accuracy: 88.47%
+
+## Confusion Matrix
+
+You can generate and visualize the confusion matrix with the provided notebook.

vit_emnist_model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:da5fec4a57fd1320e016c8a5f24f870047fb769384a66b3accef42ae6faf6469
+size 1101712