Update README.md

README.md

@@ -1,23 +1,31 @@
----
-license: apache-2.0
-language:
-- en
-base_model:
-- google/vit-base-patch16-224-in21k
-pipeline_tag: image-classification
-library_name: transformers
-tags:
-- Deepfake
-- Quality
-- Assess
----
-      Classification report:
-      
-                             precision    recall  f1-score   support
-      
-          Issue In Deepfake     0.7962    0.8067    0.8014      1500
-      High Quality Deepfake     0.7877    0.7767    0.7822      1500
-      
-                   accuracy                         0.7940      3000
-                  macro avg     0.7920    0.7917    0.7918      3000
+---
+license: apache-2.0
+language:
+- en
+base_model:
+- google/vit-base-patch16-224-in21k
+pipeline_tag: image-classification
+library_name: transformers
+tags:
+- Deepfake
+- Quality
+- Assess
+---
+# **Deepfake-QualityAssess-85M**
+
+Deepfake-QualityAssess-85M is an image classification model for quality assessment of good and bad quality deepfakes. It is based on Google's ViT model (`google/vit-base-patch16-224-in21k`).  
+
+A reasonable number of training samples were used to achieve good efficiency in the final training process and its efficiency metrics. Since this task involves classifying deepfake images with varying quality levels, the model was trained accordingly. Future improvements will be made based on the complexity of the task.
+
+
+
+      Classification report:
+      
+                             precision    recall  f1-score   support
+      
+          Issue In Deepfake     0.7962    0.8067    0.8014      1500
+      High Quality Deepfake     0.7877    0.7767    0.7822      1500
+      
+                   accuracy                         0.7940      3000
+                  macro avg     0.7920    0.7917    0.7918      3000
                weighted avg     0.7920    0.7917    0.7918      3000