File size: 2,109 Bytes
22f66bb
 
e6ae7b7
22f66bb
e6ae7b7
 
22f66bb
 
 
0e7e7a1
22f66bb
 
 
 
e6ae7b7
22f66bb
 
 
 
 
c6f9f98
 
22f66bb
 
 
 
 
 
 
 
 
 
 
0e7e7a1
12463e7
 
c0154e7
22f66bb
 
a99c182
22f66bb
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
---
tags:
- security
- summarization
- shorts
- infosec
language:
- unk
widget:
- text: "LockBit 30 Ransomware Cybersecurity News Platform Followed by 376 million Resources About Site ContactTip Us Follow Us On Social Media Aug 26 2023THNEndpoint Security Cyber Threat The leak of the builder last year has led to threat actors abusing the tool to spawn new variants Russian cybersecurity company Kaspersky said it detected a ransomware intrusion that deployed a version of LockBit but with a markedly different ransom demand procedure The attacker behind this incident decided to use a different ransom note with a headline related to a previously unknown group called NATIONAL HAZARD AGENCY security researchers Eduardo Ovalle and Francesco Figurelli The revamped ransom note directly specified the amount to be paid"
co2_eq_emissions:
  emissions: 38.597620011381935
---

# Model Trained on Cyber Security Content

- Problem type: Summarization
- Model ID: 85203142751
- CO2 Emissions (in grams): 38.5976

Try model here - https://huggingface.co/spaces/venkycs/securityShots

## Validation Metrics

- Loss: 1.693
- Rouge1: 49.395
- Rouge2: 25.760
- RougeL: 36.111
- RougeLsum: 44.288
- Gen Len: 100.144

## Usage


Staying up-to-date in the CyberSecurity domain is a nightmare, as there are over a thousand security-related updates per day. It is essential for us to focus on relevant content despite the surrounding noise. I have trained a model based on Facebook BART to efficiently summarise security news.  The model has been trained on security-related issues and can summarise cyber security news effectively. Use cURL to access this model. The dataset utilised by the model is private, and I may take time to make it accessible to the public.

LinkedIn post about the model https://www.linkedin.com/pulse/cybersecurity-feed-summarisation-context-using-ai-venkatesh-siddi :

```
$ curl -X POST -H "Authorization: Bearer YOUR_HUGGINGFACE_API_KEY" -H "Content-Type: application/json" -d '{"inputs": "Security News or Article here"}' https://api-inference.huggingface.co/models/venkycs/securityShots
```