Text Classification
Transformers
Safetensors
roberta
Generated from Trainer
cedricbonhomme commited on
Commit
c43ab1f
·
verified ·
1 Parent(s): a0837d7

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +42 -8
README.md CHANGED
@@ -9,29 +9,63 @@ metrics:
9
  model-index:
10
  - name: vulnerability-severity-classification-roberta-base
11
  results: []
 
 
12
  ---
13
 
14
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
15
  should probably proofread and complete it, then remove this comment. -->
16
 
17
- # vulnerability-severity-classification-roberta-base
 
 
 
 
 
 
 
 
 
 
 
 
 
18
 
19
- This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on an unknown dataset.
20
  It achieves the following results on the evaluation set:
21
  - Loss: 0.5065
22
  - Accuracy: 0.8293
23
 
24
- ## Model description
 
 
 
 
 
 
 
 
 
 
 
 
 
25
 
26
- More information needed
 
 
27
 
28
- ## Intended uses & limitations
 
 
 
29
 
30
- More information needed
 
 
 
 
31
 
32
- ## Training and evaluation data
33
 
34
- More information needed
35
 
36
  ## Training procedure
37
 
 
9
  model-index:
10
  - name: vulnerability-severity-classification-roberta-base
11
  results: []
12
+ datasets:
13
+ - CIRCL/vulnerability-scores
14
  ---
15
 
16
  <!-- This model card has been generated automatically according to the information the Trainer had access to. You
17
  should probably proofread and complete it, then remove this comment. -->
18
 
19
+ # VLAI: A RoBERTa-Based Model for Automated Vulnerability Severity Classification
20
+
21
+ # Severity classification
22
+
23
+ This model is a fine-tuned version of [roberta-base](https://huggingface.co/roberta-base) on the dataset [CIRCL/vulnerability-scores](https://huggingface.co/datasets/CIRCL/vulnerability-scores).
24
+
25
+ The model was presented in the paper [VLAI: A RoBERTa-Based Model for Automated Vulnerability Severity Classification](https://huggingface.co/papers/2507.03607) [[arXiv](https://arxiv.org/abs/2507.03607)].
26
+
27
+ **Abstract:** VLAI is a transformer-based model that predicts software vulnerability severity levels directly from text descriptions. Built on RoBERTa, VLAI is fine-tuned on over 600,000 real-world vulnerabilities and achieves over 82% accuracy in predicting severity categories, enabling faster and more consistent triage ahead of manual CVSS scoring. The model and dataset are open-source and integrated into the Vulnerability-Lookup service.
28
+
29
+ You can read [this page](https://www.vulnerability-lookup.org/user-manual/ai/) for more information.
30
+
31
+
32
+ ## Model description
33
 
 
34
  It achieves the following results on the evaluation set:
35
  - Loss: 0.5065
36
  - Accuracy: 0.8293
37
 
38
+ It is a classification model and is aimed to assist in classifying vulnerabilities by severity based on their descriptions.
39
+
40
+ ## How to get started with the model
41
+
42
+ ```python
43
+ from transformers import AutoModelForSequenceClassification, AutoTokenizer
44
+ import torch
45
+
46
+ labels = ["low", "medium", "high", "critical"]
47
+
48
+ model_name = "CIRCL/vulnerability-severity-classification-roberta-base"
49
+ tokenizer = AutoTokenizer.from_pretrained(model_name)
50
+ model = AutoModelForSequenceClassification.from_pretrained(model_name)
51
+ model.eval()
52
 
53
+ test_description = "SAP NetWeaver Visual Composer Metadata Uploader is not protected with a proper authorization, allowing unauthenticated agent to upload potentially malicious executable binaries \
54
+ that could severely harm the host system. This could significantly affect the confidentiality, integrity, and availability of the targeted system."
55
+ inputs = tokenizer(test_description, return_tensors="pt", truncation=True, padding=True)
56
 
57
+ # Run inference
58
+ with torch.no_grad():
59
+ outputs = model(**inputs)
60
+ predictions = torch.nn.functional.softmax(outputs.logits, dim=-1)
61
 
62
+ # Print results
63
+ print("Predictions:", predictions)
64
+ predicted_class = torch.argmax(predictions, dim=-1).item()
65
+ print("Predicted severity:", labels[predicted_class])
66
+ ```
67
 
 
68
 
 
69
 
70
  ## Training procedure
71