bhavitvyamalik commited on
Commit
130d1ed
1 Parent(s): 0c06b9c

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +35 -19
README.md CHANGED
@@ -1,20 +1,36 @@
1
- ---
2
- license: mit
3
- ---
4
- ### Dataset used
5
- [Fake and real news dataset](https://www.kaggle.com/datasets/clmentbisaillon/fake-and-real-news-dataset)
6
-
7
- ### Labels
8
- Fake news: 1 </br>
9
- Real news: 0
10
-
11
- ### Performance on test data
12
- ```json
13
- 'test/accuracy': 0.9977836608886719,
14
- 'test/aucroc': 0.9999998807907104,
15
- 'test/f1': 0.9976308941841125,
16
- 'test/loss': 0.00828308891505003
17
- ```
18
-
19
- ### Run can be tracked here
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
20
  [Wandb project for Fake news classifier](https://wandb.ai/bhavitvya/Fake%20news%20classifier?workspace=user-bhavitvya)
 
1
+ ---
2
+ license: mit
3
+ ---
4
+ ### Dataset used
5
+ [Fake and real news dataset](https://www.kaggle.com/datasets/clmentbisaillon/fake-and-real-news-dataset)
6
+
7
+ ### Labels
8
+ Fake news: 1 </br>
9
+ Real news: 0
10
+
11
+ ### Usage
12
+ ```python
13
+ from transformers import AutoModelForSequenceClassification, AutoTokenizer, AutoConfig
14
+ import torch
15
+
16
+ config = AutoConfig.from_pretrained("bhavitvyamalik/fake-news_xtremedistil-l6-h256-uncased")
17
+ model = AutoModelForSequenceClassification.from_pretrained("bhavitvyamalik/fake-news_xtremedistil-l6-h256-uncased", config=config)
18
+ tokenizer = AutoTokenizer.from_pretrained("microsoft/xtremedistil-l6-h256-uncased", usefast=True)
19
+
20
+ text = "According to reports by Fox News, Biden is the President of the USA"
21
+ encode = tokenizer(text, max_length=512, truncation=True, padding="max_length", return_tensors="pt")
22
+
23
+ output = model(**encode)
24
+ print(torch.argmax(output["logits"]))
25
+ ```
26
+
27
+ ### Performance on test data
28
+ ```json
29
+ 'test/accuracy': 0.9977836608886719,
30
+ 'test/aucroc': 0.9999998807907104,
31
+ 'test/f1': 0.9976308941841125,
32
+ 'test/loss': 0.00828308891505003
33
+ ```
34
+
35
+ ### Run can be tracked here
36
  [Wandb project for Fake news classifier](https://wandb.ai/bhavitvya/Fake%20news%20classifier?workspace=user-bhavitvya)