update

Author: CodeByStella

README.md

@@ -6,6 +6,7 @@
 | 4   | [Transformer: Attention Overview](./diagrams/transformer-attention-overview.md)         |
 | 5   | [LLM Fine-tuning Pipeline](./diagrams/llm-finetuning-pipeline.md)                       |
 | 6   | [Neural Network Overview](./diagrams/neural-network-overview.md)                        |
+| 7   | [ML Data Split: Train, Validation, Test](./diagrams/ml-data-split.md)                  |
 
 # My Diagrams
 

diagrams/ml-data-split.md

@@ -0,0 +1,22 @@
+# ML Data Split: Train, Validation, Test
+
+```mermaid
+flowchart TD
+    A[Raw Dataset] --> B[Split]
+    B --> C[Train Set]
+    B --> D[Validation Set]
+    B --> E[Test Set]
+
+    C --> F[Model Training]
+    D --> G[Hyperparameter Tuning]
+    E --> H[Final Evaluation]
+
+    F --> I[Trained Model]
+    G -.-> F
+    I --> H
+```
+
+**Description:**  
+Split data into train (fit model), validation (tune hyperparameters, early stopping), and test (unseen holdout for final metrics). Never tune on test.
+
+**Flow:** Raw data → split → train/val/test → train on train set, tune on validation, report on test set.

diagrams/neural-network-overview.md

@@ -8,7 +8,7 @@ flowchart LR
         I3[x₃]
     end
 
-    subgraph hidden[Hidden Layer(s)]
+    subgraph hidden[Hidden Layers]
         H1
         H2
         H3