updated readme

README.md

@@ -14,11 +14,31 @@ safetensors:
   total: 1
   format: safetensors
   weight_dtype: float32
-  size_in_bytes: 80000000   # ≈ 82 MB (you can adjust to your real file size)
+  size_in_bytes: 80000000   
 
 model-index:
 - name: Deepfake Detection with Improved EfficientViT
-  results: []
+  results:
+  - task:
+      type: image-classification
+      name: Deepfake Detection
+    dataset:
+      type: custom
+      name: FaceForensics++,Celeb-DF
+    metrics:
+      - name: Accuracy
+        type: accuracy
+        value: 0.8864
+      - name: Precision
+        type: precision
+        value: 0.8920
+      - name: Recall
+        type: recall
+        value: 0.8792
+      - name: F1-score
+        type: f1
+        value: 0.8856
+
   config: config.json
   metadata:
     model_type: EfficientViT
@@ -90,19 +110,21 @@ pip install -r requirements.txt
 ```python
 
 from huggingface_hub import hf_hub_download
+from safetensors.torch import load_file
 import torch
 from model import ImprovedEfficientViT
-from inference import predict_vedio  # your inference function
+from inference import predict_vedio 
 
 # 1️⃣ Download the checkpoint from Hugging Face
 checkpoint_path = hf_hub_download(
     repo_id="faisalishfaq2005/deepfake-detection-efficientnet-vit",  
-    filename="model.pth"
+    filename="model.safetensors"
 )
 
-# 2️⃣ Load the model
+# 2️⃣ Load the model weights safely
+state_dict = load_file(checkpoint_path)
 model = ImprovedEfficientViT()
-model.load_state_dict(torch.load(checkpoint_path, map_location="cpu"))
+model.load_state_dict(state_dict)
 model.eval()
 
 # 3️⃣ Run inference on a video
@@ -111,7 +133,6 @@ result = predict_vedio(video_path, model)
 print(result)
 # Example Output: {'class': 1}
 
-
 ```
 # 2. Manual Download