Update model links and training configuration details

README.md

@@ -14,18 +14,18 @@ short_description: 'Block 2 '
 
 Compare three classification approaches on CIFAR-10 images:
 
-- **Fine-tuned ViT** ([adisaljusi/vit-base-cifar10](https://huggingface.co/adisaljusi/vit-base-cifar10)) — transfer learning model trained on CIFAR-10
+- **Fine-tuned ViT** ([adisaljusi/cifar10-vit](https://huggingface.co/adisaljusi/cifar10-vit)) — transfer learning model trained on CIFAR-10
 - **CLIP Zero-Shot** (`openai/clip-vit-large-patch14`) — open-source zero-shot classification
 - **OpenAI GPT-4.1-mini** — closed-source vision model via API
 
 ## Dataset
 
-**CIFAR-10** — 60,000 32x32 color images in 10 classes (6,000 images per class):
+**CIFAR-10** — 60,000 32x32 color images in 10 classes (6,000 images per class). For training we use a smaller subset to keep runtime manageable on a laptop:
 
-| Split | Images |
-|-------|--------|
-| Train | 50,000 |
-| Test  | 10,000 |
+| Split | Full Dataset | Used for Training |
+|-------|-------------:|------------------:|
+| Train | 50,000 | 8,000 |
+| Test  | 10,000 | 2,000 |
 
 **Classes:** airplane, automobile, bird, cat, deer, dog, frog, horse, ship, truck
 
@@ -45,7 +45,7 @@ Applied using `AutoImageProcessor` from `google/vit-base-patch16-224`.
 
 **Transfer learning approach:** All layers frozen except the final classification head (10 outputs for CIFAR-10 classes). Only 7,690 of 85.8M parameters are trainable.
 
-**Training config:** 5 epochs, batch size 16, learning rate 3e-4, AdamW optimizer.
+**Training config:** 4 epochs, batch size 32, learning rate 2e-4, warmup ratio 0.1, weight decay 0.01, AdamW optimizer.
 
 ### Training Results
 
@@ -55,7 +55,7 @@ Applied using `AutoImageProcessor` from `google/vit-base-patch16-224`.
 
 ## Links
 
-- **Model:** [adisaljusi/vit-base-cifar10](https://huggingface.co/adisaljusi/vit-base-cifar10)
+- **Model:** [adisaljusi/cifar10-vit](https://huggingface.co/adisaljusi/cifar10-vit)
 - **App:** [adisaljusi/computer-vision-classification-model-comparison](https://huggingface.co/spaces/adisaljusi/computer-vision-classification-model-comparison)
 
 ## Comparison Results