docs: Update model card with latest evaluation results

README.md

@@ -13,31 +13,32 @@ pipeline_tag: image-classification
 
 # VeriChain Deepfake Detection Model - ViT
 
-This repository contains the Vision Transformer (ViT) model fine-tuned for the task of Deepfake detection, developed as part of the VeriChain project.
+This repository contains the artifacts for the Vision Transformer (ViT) model fine-tuned for the task of Deepfake detection, developed as part of the VeriChain project.
 
-The model is trained to classify an image into one of three categories: **Real**, **AI-Generated**, or **Deepfake**. The model files are located in the `vit-deepfake-model/` directory.
+The model is trained to classify an image into one of three categories: **Real**, **AI-Generated**, or **Deepfake**.
 
-## Model Details
+## Repository Structure
 
--   **Base Model:** `google/vit-base-patch16-224-in21k`
--   **Fine-tuning Dataset:** [einrafh/verichain-deepfake-data](https://huggingface.co/datasets/einrafh/verichain-deepfake-data)
--   **Framework:** PyTorch
--   **Architecture:** A Vision Transformer with a classification head fine-tuned for 3 labels.
+The model artifacts in this repository are organized as follows:
 
-## How to Use
+-   **/models/vit-deepfake-model/**: Contains the final, fine-tuned PyTorch model files, ready to be loaded with the `transformers` library.
+-   **/models/onnx/**: Contains the model converted to the ONNX format, optimized for production deployment and inference.
+-   **/assets/**: Contains visual assets for documentation, such as the confusion matrix.
 
-This model can be used for inference directly with the `pipeline` function from the `transformers` library.
+## How to Use (PyTorch Model)
+
+You can use the fine-tuned PyTorch model directly with the `pipeline` function from the `transformers` library. Make sure to specify the correct `subfolder`.
 
 ```python
 from transformers import pipeline
 from PIL import Image
 
 # Load the image classification pipeline with your model
-# 'subfolder' points to the directory containing the model files
+# The 'subfolder' parameter points to the directory containing the model files
 classifier = pipeline(
     "image-classification", 
     model="einrafh/verichain-deepfake-models", 
-    subfolder="vit-deepfake-model"
+    subfolder="models/vit-deepfake-model"
 )
 
 # Load an image you want to classify
@@ -50,24 +51,32 @@ except FileNotFoundError:
     print("Please provide a valid path to an image file.")
 
 # Example Output:
-# [{'label': 'Deepfake', 'score': 0.9978}, {'label': 'AI Generated', 'score': 0.0012}, {'label': 'Real', 'score': 0.0010}]
+# [{'label': 'Deepfake', 'score': 0.9985}, {'label': 'AI Generated', 'score': 0.0010}, {'label': 'Real', 'score': 0.0005}]
 ```
 
 ## Evaluation Results
 
-The model was evaluated on a held-out test set of 2,000 images, achieving excellent performance.
+The model was evaluated on a held-out test set of 2,000 images, achieving near-perfect performance.
 
 | Metric               | Score   |
 |----------------------|---------|
-| **Test Accuracy** | **0.9985** |
-| **F1-Score (Macro)** | **0.9985** |
-| Test Loss            | 0.01906 |
+| **Test Accuracy** | **0.9990** |
+| **F1-Score (Macro)** | **0.9990** |
+| Test Loss            | 0.0202  |
+
+### Classification Report
+
+| Class          | Precision | Recall | F1-Score |
+|----------------|-----------|--------|----------|
+| AI Generated   | 1.0000    | 0.9970 | 0.9985   |
+| Deepfake       | 0.9970    | 1.0000 | 0.9985   |
+| Real           | 1.0000    | 1.0000 | 1.0000   |
 
 ### Confusion Matrix
 
 The confusion matrix below shows the model's high precision and recall across all classes, with very few misclassifications.
 
-![Confusion Matrix](https://huggingface.co/einrafh/verichain-deepfake-models/resolve/main/vit-deepfake-model/confusion_matrix.png)
+![Confusion Matrix](https://huggingface.co/einrafh/verichain-deepfake-models/resolve/main/assets/confusion_matrix.png)
 
 ## Citation