Shardul-Ghuge commited on
Commit
4d46349
1 Parent(s): e119d6c

Update Readme.md

Browse files
Files changed (1) hide show
  1. README.md +10 -6
README.md CHANGED
@@ -1,4 +1,8 @@
1
- ## Classifier-Bias-SG Model Card
 
 
 
 
2
 
3
  ## Model Details
4
  Classifier-Bias-SG is a proof of concept model designed to classify texts based on their bias levels. The model categorizes texts into 2 classes: "Biased", and "Non-Biased".
@@ -8,8 +12,8 @@ The model is built upon the distilbert-base-uncased architecture and has been fi
8
 
9
  ## Dataset
10
  The model was trained on a BABE dataset containing news articles from various sources, annotated with one of the 2 bias levels. The dataset contains:
11
- Biased: 1810 articles
12
- Non-Biased: 1810 articles
13
 
14
  ## Training Procedure
15
  The model was trained using the Adam optimizer for 15 epochs.
@@ -17,9 +21,9 @@ The model was trained using the Adam optimizer for 15 epochs.
17
  ## Performance
18
  On our validation set, the model achieved:
19
 
20
- Accuracy: 78%
21
- F1 Score (Biased): 79%
22
- F1 Score (Non-Biased): 77%
23
 
24
  ## How to Use
25
  To use this model for text classification, use the following code:
 
1
+ ---
2
+ license: openrail
3
+ ---
4
+
5
+ # Classifier-Bias-SG Model Card
6
 
7
  ## Model Details
8
  Classifier-Bias-SG is a proof of concept model designed to classify texts based on their bias levels. The model categorizes texts into 2 classes: "Biased", and "Non-Biased".
 
12
 
13
  ## Dataset
14
  The model was trained on a BABE dataset containing news articles from various sources, annotated with one of the 2 bias levels. The dataset contains:
15
+ - **Biased**: 1810 articles
16
+ - **Non-Biased**: 1810 articles
17
 
18
  ## Training Procedure
19
  The model was trained using the Adam optimizer for 15 epochs.
 
21
  ## Performance
22
  On our validation set, the model achieved:
23
 
24
+ - **Accuracy**: 78%
25
+ - **F1 Score (Biased)**: 79%
26
+ - **F1 Score (Non-Biased)**: 77%
27
 
28
  ## How to Use
29
  To use this model for text classification, use the following code: