Create readme.md

Updated temp readme

readme.md

@@ -0,0 +1,67 @@
+# Cat Breed Classification & Model Comparison
+
+## Project Overview
+
+This project presents a computer vision application that classifies images of different cat breeds.
+
+The goal is to compare three approaches to image classification:
+	1.	A fine-tuned Vision Transformer (ViT) model trained on a custom dataset
+	2.	A zero-shot CLIP model (open-source)
+	3.	An OpenAI vision model (closed-source)
+
+The application is deployed as a Hugging Face Space and allows users to upload images or select example images.
+
+
+## Dataset Description
+
+### The dataset consists of images from seven cat breeds:
+- Sphynx
+- Russian Blue
+- Maine Coon
+- Ragdoll
+- Bengal
+- Singapura
+- Calico Cat
+
+### Dataset characteristics:
+- Number of classes: 7
+- Images per class: ~[fill in]
+- Total images: ~[fill in]
+
+### Data sources:
+- Public datasets (Kaggle / Hugging Face)
+- Manually collected images
+
+### Split:
+- Training: 80%
+- Validation/Test: 20%
+
+
+## Preprocessing Steps
+- Resize images to 224 × 224
+- Convert to RGB
+- Remove corrupted images
+- Normalize using model-specific values
+
+## Data Augmentation
+- Random horizontal flip
+- Random rotation
+- Optional brightness/contrast adjustments
+
+
+## Model and Training
+
+### Fine-Tuned Model
+- Base model: google/vit-base-patch16-224
+- Approach: Transfer learning + fine-tuning
+- Output classes: 7
+
+### Training settings:
+- Epochs: [e.g. 5–10]
+- Batch size: [e.g. 16]
+- Learning rate: [e.g. 2e-5]
+
+
+### Links
+- Hugging Face Space: [ADD LINK]
+- Hugging Face Model: [ADD LINK]