---
library_name: transformers
license: cc-by-4.0
base_model: roberta-base
metrics:
- accuracy
tags:
- generated_from_trainer
- text-classification
- classification
- nlp
- vulnerability
model-index:
- name: vulnerability-severity-classification-roberta-base
  results: []
datasets:
- CIRCL/vulnerability-scores
---


# VLAI: A RoBERTa-Based Model for Automated Vulnerability Severity Classification

# Severity classification

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).

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)].

**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.

You can read [this page](https://www.vulnerability-lookup.org/user-manual/ai/) for more information.


## Model description

It is a classification model and is aimed to assist in classifying vulnerabilities by severity based on their descriptions.

## How to get started with the model

```python
from transformers import AutoModelForSequenceClassification, AutoTokenizer
import torch

labels = ["low", "medium", "high", "critical"]

model_name = "CIRCL/vulnerability-severity-classification-roberta-base"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
model.eval()

print("Model revision:", model.config._commit_hash)

test_description = "SAP NetWeaver Visual Composer Metadata Uploader is not protected with a proper authorization, allowing unauthenticated agent to upload potentially malicious executable binaries \
that could severely harm the host system. This could significantly affect the confidentiality, integrity, and availability of the targeted system."
inputs = tokenizer(test_description, return_tensors="pt", truncation=True, padding=True)

# Run inference
with torch.no_grad():
    outputs = model(**inputs)
    predictions = torch.nn.functional.softmax(outputs.logits, dim=-1)

# Print results
print("Predictions:", predictions)
predicted_class = torch.argmax(predictions, dim=-1).item()
print("Predicted severity:", labels[predicted_class])
```

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- learning_rate: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 5

It achieves the following results on the evaluation set:
- Loss: 1.9981
- Accuracy: 0.8188
- F1 Macro: 0.7502
- Low Precision: 0.6490
- Low Recall: 0.5113
- Low F1: 0.572
- Medium Precision: 0.8488
- Medium Recall: 0.8681
- Medium F1: 0.8584
- High Precision: 0.8148
- High Recall: 0.8129
- High F1: 0.8138
- Critical Precision: 0.7592
- Critical Recall: 0.7543
- Critical F1: 0.7567

### Training results

| Training Loss | Epoch | Step  | Validation Loss | Accuracy | F1 Macro | Low Precision | Low Recall | Low F1 | Medium Precision | Medium Recall | Medium F1 | High Precision | High Recall | High F1 | Critical Precision | Critical Recall | Critical F1 |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:--------:|:-------------:|:----------:|:------:|:----------------:|:-------------:|:---------:|:--------------:|:-----------:|:-------:|:------------------:|:---------------:|:-----------:|
| 2.9393        | 1.0   | 16760 | 2.5264          | 0.7363   | 0.6358   | 0.5562        | 0.3500     | 0.4297 | 0.7994           | 0.7985        | 0.7990    | 0.6842         | 0.7790      | 0.7286  | 0.7063             | 0.5008          | 0.5861      |
| 2.3025        | 2.0   | 33520 | 2.2776          | 0.7710   | 0.6785   | 0.6646        | 0.3369     | 0.4471 | 0.7978           | 0.8607        | 0.8280    | 0.7771         | 0.7286      | 0.7520  | 0.6673             | 0.7073          | 0.6867      |
| 1.7004        | 3.0   | 50280 | 2.1303          | 0.7911   | 0.7164   | 0.5839        | 0.4895     | 0.5325 | 0.8184           | 0.8623        | 0.8398    | 0.7968         | 0.7644      | 0.7803  | 0.7224             | 0.7041          | 0.7131      |
| 1.3977        | 4.0   | 67040 | 2.0143          | 0.8087   | 0.7335   | 0.6898        | 0.4418     | 0.5386 | 0.8412           | 0.8622        | 0.8516    | 0.7999         | 0.8029      | 0.8014  | 0.7357             | 0.7489          | 0.7422      |
| 1.5018        | 5.0   | 83800 | 1.9981          | 0.8188   | 0.7502   | 0.6490        | 0.5113     | 0.572  | 0.8488           | 0.8681        | 0.8584    | 0.8148         | 0.8129      | 0.8138  | 0.7592             | 0.7543          | 0.7567      |


### Framework versions

- Transformers 5.10.2
- Pytorch 2.12.0+cu130
- Datasets 4.8.5
- Tokenizers 0.22.2