Update README.md

README.md

@@ -57,6 +57,20 @@ The task is to predict whether a food image belongs to **Asian** or **Western**
 
 ---
 
+### Search Space and Hyperparameters
+AutoGluon automatically searched across the following dimensions:
+- **Architectures (depth/width):** ResNet18 (shallow), MobileNetV3-Small (compact width), EfficientNet-B0 (deeper, wider baseline).  
+- **Optimizers:** Variants of AdamW and SGD.  
+- **Learning rate / weight decay:** Schedules in range ~1e-3 to 1e-5 with decay applied.  
+- **Regularization:** Implicit dropout layers (in backbones) and weight decay.  
+- **Augmentation:**  
+  - Random crops and flips (default torchvision pipeline).  
+  - RandAugment (with random distortions).  
+  - Mixup (interpolated samples).  
+- **Early stopping:** Triggered automatically when validation metric stops improving.
+
+---
+
 ## Results
 - **Best model (AutoGluon selected):** *efficientnet_b0*  
 - **Validation Accuracy:** *0.96875*