Hugging Face's logo

hleAtKeeper
/
threat-categorization-classifier-setfit

Text Classification
setfit
Safetensors
sentence-transformers
bert
generated_from_setfit_trainer
Eval Results
Model card Files Community

SetFit with sentence-transformers/paraphrase-MiniLM-L6-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-MiniLM-L6-v2 as the Sentence Transformer embedding model. A LogisticRegression instance is used for classification.

The model has been trained using an efficient few-shot learning technique that involves:

  1. Fine-tuning a Sentence Transformer with contrastive learning.
  2. Training a classification head with features from the fine-tuned Sentence Transformer.

Model Details

Model Description

Model Sources

Model Labels

Label Examples
Log Modification
  • 'sudo rm -rf /var/log/*.log'
  • "sudo logger 'Custom log entry'"
  • 'sudo touch /var/log/custom.log'
RDP Session Activity
  • 'systemctl start xrdp'
  • 'xfreerdp /v:winserver.local /u:attacker /p:letmein123'
  • 'freerdp-shadow-cli /control'
Credential Usage
  • 'vault kv put secret/mysecret mykey=myvalue'
  • 'openssl genpkey -algorithm RSA -out private.pem'
  • 'git checkout -b new-branch'
Authentication Attempt
  • 'cat /var/log/auth.log'
  • 'sudo -u admin ls /home/admin'
  • 'chmod 600 ~/.ssh/id_rsa'
SSH Session Activity
  • 'ssh-add -l'
  • 'ssh-agent bash'
  • 'ssh-copy-id user@remotehost'
Container Management
  • 'docker push myimage'
  • 'docker start mycontainer'
  • 'docker build -t myapp .'
System Navigation
  • 'ls /sys'
  • 'df -h'
  • 'mount'
Potential Data Exfiltration
Package Management
  • 'apt remove apache2'
  • 'apt autoremove'
  • 'yum update'
System Configuration Change
  • 'sudo chmod 700 /home/user'
  • 'sudo systemctl stop nginx'
  • 'sudo groupadd newgroup'
Script Execution
  • 'python3 -m pip install requests'
  • 'sh script.sh'
  • 'chmod 755 script.py'
Network Configuration
  • 'curl ifconfig.me'
  • 'systemctl status NetworkManager'
  • 'nmtui'
Process Execution
  • 'strace -p 12345'
  • 'killall nginx'
  • 'systemctl disable apache2'
Unauthorized Access
  • 'sudo -s'
  • 'sudo visudo'
  • 'passwd --stdin user'
Sensitive File Access
  • 'cat /etc/ssh/sshd_config'
  • 'vim /etc/shadow'
  • 'gzip -c /etc/shadow > shadow.gz'
Kubernetes Interaction
  • 'kubectl get events'
  • 'kubectl get namespaces'
  • 'kubectl scale deployment nginx-deployment --replicas=5'
Privilege Escalation
  • "alias sudo='bash'"
  • 'chmod u+s /bin/bash'
  • 'ssh-keygen -f /tmp/key && cat /tmp/key.pub >> /root/.ssh/authorized_keys'
Unusual Command Pattern
  • 'history -c && unset HISTFILE'
  • 'while true; do :; done'
  • "echo '* * * * * root /bin/bash' >> /etc/crontab"
Database Access
  • "sqlite3 sensitive.db '.tables'"
  • "sqlite3 sensitive.db 'DELETE FROM logs;'"
  • 'pg_dump -U postgres -W sensitive_db > pg_backup.sql'

Evaluation

Metrics

Label Accuracy
all 0.7068

Uses

Direct Use for Inference

First install the SetFit library:

pip install setfit

Then you can load this model and run inference.

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("setfit_model_id")
# Run inference
preds = model("pstree -p")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 1 4.0921 22
Label Training Sample Count
Unauthorized Access 31
Privilege Escalation 26
Sensitive File Access 30
Potential Data Exfiltration 24
Unusual Command Pattern 28
System Navigation 31
Database Access 25
SSH Session Activity 26
RDP Session Activity 32
Kubernetes Interaction 28
Container Management 28
Process Execution 33
Network Configuration 27
Package Management 28
Script Execution 30
Authentication Attempt 27
Credential Usage 25
Log Modification 24
System Configuration Change 29

Training Hyperparameters

  • batch_size: (16, 16)
  • num_epochs: (1, 1)
  • max_steps: -1
  • sampling_strategy: oversampling
  • body_learning_rate: (2e-05, 1e-05)
  • head_learning_rate: 0.01
  • loss: CosineSimilarityLoss
  • distance_metric: cosine_distance
  • margin: 0.25
  • end_to_end: False
  • use_amp: False
  • warmup_proportion: 0.1
  • l2_weight: 0.01
  • seed: 42
  • eval_max_steps: -1
  • load_best_model_at_end: True

Training Results

Epoch Step Training Loss Validation Loss
0.0001 1 0.2873 -
0.0030 50 0.2416 -
0.0060 100 0.2458 -
0.0090 150 0.2196 -
0.0119 200 0.2001 -
0.0149 250 0.204 -
0.0179 300 0.1896 -
0.0209 350 0.1929 -
0.0239 400 0.1809 -
0.0269 450 0.1652 -
0.0299 500 0.1521 -
0.0328 550 0.143 -
0.0358 600 0.1367 -
0.0388 650 0.1367 -
0.0418 700 0.1288 -
0.0448 750 0.1213 -
0.0478 800 0.1137 -
0.0507 850 0.1213 -
0.0537 900 0.1042 -
0.0567 950 0.0942 -
0.0597 1000 0.0942 -
0.0627 1050 0.0948 -
0.0657 1100 0.0845 -
0.0687 1150 0.0817 -
0.0716 1200 0.0862 -
0.0746 1250 0.0798 -
0.0776 1300 0.069 -
0.0806 1350 0.0749 -
0.0836 1400 0.0773 -
0.0866 1450 0.0723 -
0.0896 1500 0.0733 -
0.0925 1550 0.0599 -
0.0955 1600 0.065 -
0.0985 1650 0.0644 -
0.1015 1700 0.0586 -
0.1045 1750 0.0575 -
0.1075 1800 0.0531 -
0.1104 1850 0.0515 -
0.1134 1900 0.05 -
0.1164 1950 0.046 -
0.1194 2000 0.0418 -
0.1224 2050 0.0418 -
0.1254 2100 0.0437 -
0.1284 2150 0.0434 -
0.1313 2200 0.0362 -
0.1343 2250 0.0366 -
0.1373 2300 0.0237 -
0.1403 2350 0.0396 -
0.1433 2400 0.0339 -
0.1463 2450 0.0341 -
0.1493 2500 0.0306 -
0.1522 2550 0.0261 -
0.1552 2600 0.0309 -
0.1582 2650 0.0223 -
0.1612 2700 0.0263 -
0.1642 2750 0.0184 -
0.1672 2800 0.0272 -
0.1701 2850 0.0201 -
0.1731 2900 0.0193 -
0.1761 2950 0.0218 -
0.1791 3000 0.0251 -
0.1821 3050 0.0185 -
0.1851 3100 0.0182 -
0.1881 3150 0.019 -
0.1910 3200 0.0155 -
0.1940 3250 0.0208 -
0.1970 3300 0.0231 -
0.2 3350 0.0202 -
0.2030 3400 0.0162 -
0.2060 3450 0.0188 -
0.2090 3500 0.016 -
0.2119 3550 0.0194 -
0.2149 3600 0.0187 -
0.2179 3650 0.0179 -
0.2209 3700 0.0166 -
0.2239 3750 0.0155 -
0.2269 3800 0.0175 -
0.2299 3850 0.018 -
0.2328 3900 0.0095 -
0.2358 3950 0.0145 -
0.2388 4000 0.0102 -
0.2418 4050 0.0131 -
0.2448 4100 0.0191 -
0.2478 4150 0.0136 -
0.2507 4200 0.0126 -
0.2537 4250 0.0138 -
0.2567 4300 0.0102 -
0.2597 4350 0.01 -
0.2627 4400 0.0108 -
0.2657 4450 0.0125 -
0.2687 4500 0.0129 -
0.2716 4550 0.0095 -
0.2746 4600 0.0089 -
0.2776 4650 0.012 -
0.2806 4700 0.0085 -
0.2836 4750 0.0102 -
0.2866 4800 0.0106 -
0.2896 4850 0.007 -
0.2925 4900 0.006 -
0.2955 4950 0.0088 -
0.2985 5000 0.0094 -
0.3015 5050 0.0074 -
0.3045 5100 0.0105 -
0.3075 5150 0.0062 -
0.3104 5200 0.0068 -
0.3134 5250 0.0134 -
0.3164 5300 0.0104 -
0.3194 5350 0.0061 -
0.3224 5400 0.0087 -
0.3254 5450 0.0093 -
0.3284 5500 0.0066 -
0.3313 5550 0.0068 -
0.3343 5600 0.0081 -
0.3373 5650 0.0056 -
0.3403 5700 0.0064 -
0.3433 5750 0.009 -
0.3463 5800 0.008 -
0.3493 5850 0.0075 -
0.3522 5900 0.0054 -
0.3552 5950 0.0072 -
0.3582 6000 0.0054 -
0.3612 6050 0.0058 -
0.3642 6100 0.0052 -
0.3672 6150 0.0063 -
0.3701 6200 0.0047 -
0.3731 6250 0.0044 -
0.3761 6300 0.008 -
0.3791 6350 0.0068 -
0.3821 6400 0.0054 -
0.3851 6450 0.0065 -
0.3881 6500 0.0094 -
0.3910 6550 0.0055 -
0.3940 6600 0.0069 -
0.3970 6650 0.0027 -
0.4 6700 0.0045 -
0.4030 6750 0.0041 -
0.4060 6800 0.0051 -
0.4090 6850 0.0048 -
0.4119 6900 0.0065 -
0.4149 6950 0.0064 -
0.4179 7000 0.0049 -
0.4209 7050 0.0068 -
0.4239 7100 0.0069 -
0.4269 7150 0.0072 -
0.4299 7200 0.0024 -
0.4328 7250 0.0047 -
0.4358 7300 0.0063 -
0.4388 7350 0.0034 -
0.4418 7400 0.0038 -
0.4448 7450 0.0049 -
0.4478 7500 0.0047 -
0.4507 7550 0.0047 -
0.4537 7600 0.0081 -
0.4567 7650 0.0059 -
0.4597 7700 0.0071 -
0.4627 7750 0.006 -
0.4657 7800 0.005 -
0.4687 7850 0.0051 -
0.4716 7900 0.0043 -
0.4746 7950 0.0062 -
0.4776 8000 0.0042 -
0.4806 8050 0.0036 -
0.4836 8100 0.0049 -
0.4866 8150 0.0024 -
0.4896 8200 0.0052 -
0.4925 8250 0.005 -
0.4955 8300 0.0041 -
0.4985 8350 0.0043 -
0.5015 8400 0.0061 -
0.5045 8450 0.0055 -
0.5075 8500 0.005 -
0.5104 8550 0.0054 -
0.5134 8600 0.004 -
0.5164 8650 0.0039 -
0.5194 8700 0.0059 -
0.5224 8750 0.0063 -
0.5254 8800 0.0048 -
0.5284 8850 0.0031 -
0.5313 8900 0.0039 -
0.5343 8950 0.0049 -
0.5373 9000 0.0046 -
0.5403 9050 0.0035 -
0.5433 9100 0.0045 -
0.5463 9150 0.0022 -
0.5493 9200 0.0041 -
0.5522 9250 0.0037 -
0.5552 9300 0.0046 -
0.5582 9350 0.004 -
0.5612 9400 0.0079 -
0.5642 9450 0.0071 -
0.5672 9500 0.0049 -
0.5701 9550 0.0036 -
0.5731 9600 0.005 -
0.5761 9650 0.005 -
0.5791 9700 0.005 -
0.5821 9750 0.0036 -
0.5851 9800 0.0029 -
0.5881 9850 0.0044 -
0.5910 9900 0.0031 -
0.5940 9950 0.0047 -
0.5970 10000 0.0061 -
0.6 10050 0.0016 -
0.6030 10100 0.0049 -
0.6060 10150 0.0044 -
0.6090 10200 0.0032 -
0.6119 10250 0.0036 -
0.6149 10300 0.0037 -
0.6179 10350 0.0036 -
0.6209 10400 0.0033 -
0.6239 10450 0.0057 -
0.6269 10500 0.0038 -
0.6299 10550 0.0039 -
0.6328 10600 0.0024 -
0.6358 10650 0.0057 -
0.6388 10700 0.0044 -
0.6418 10750 0.0036 -
0.6448 10800 0.0064 -
0.6478 10850 0.0054 -
0.6507 10900 0.0023 -
0.6537 10950 0.0051 -
0.6567 11000 0.0038 -
0.6597 11050 0.0048 -
0.6627 11100 0.004 -
0.6657 11150 0.004 -
0.6687 11200 0.0027 -
0.6716 11250 0.0046 -
0.6746 11300 0.0044 -
0.6776 11350 0.0066 -
0.6806 11400 0.0031 -
0.6836 11450 0.0019 -
0.6866 11500 0.0022 -
0.6896 11550 0.0028 -
0.6925 11600 0.0043 -
0.6955 11650 0.0041 -
0.6985 11700 0.0064 -
0.7015 11750 0.0033 -
0.7045 11800 0.0019 -
0.7075 11850 0.0045 -
0.7104 11900 0.0052 -
0.7134 11950 0.0034 -
0.7164 12000 0.0036 -
0.7194 12050 0.0029 -
0.7224 12100 0.0042 -
0.7254 12150 0.0044 -
0.7284 12200 0.0032 -
0.7313 12250 0.0038 -
0.7343 12300 0.002 -
0.7373 12350 0.0027 -
0.7403 12400 0.006 -
0.7433 12450 0.0017 -
0.7463 12500 0.0025 -
0.7493 12550 0.004 -
0.7522 12600 0.0048 -
0.7552 12650 0.0026 -
0.7582 12700 0.0028 -
0.7612 12750 0.0036 -
0.7642 12800 0.0027 -
0.7672 12850 0.0029 -
0.7701 12900 0.0025 -
0.7731 12950 0.0045 -
0.7761 13000 0.0039 -
0.7791 13050 0.0027 -
0.7821 13100 0.0033 -
0.7851 13150 0.0038 -
0.7881 13200 0.003 -
0.7910 13250 0.0045 -
0.7940 13300 0.0062 -
0.7970 13350 0.0039 -
0.8 13400 0.005 -
0.8030 13450 0.0042 -
0.8060 13500 0.0031 -
0.8090 13550 0.0032 -
0.8119 13600 0.0024 -
0.8149 13650 0.005 -
0.8179 13700 0.0037 -
0.8209 13750 0.0035 -
0.8239 13800 0.0024 -
0.8269 13850 0.0024 -
0.8299 13900 0.0033 -
0.8328 13950 0.0028 -
0.8358 14000 0.0019 -
0.8388 14050 0.0036 -
0.8418 14100 0.0028 -
0.8448 14150 0.0034 -
0.8478 14200 0.0035 -
0.8507 14250 0.0032 -
0.8537 14300 0.0019 -
0.8567 14350 0.0022 -
0.8597 14400 0.0044 -
0.8627 14450 0.0028 -
0.8657 14500 0.0024 -
0.8687 14550 0.0028 -
0.8716 14600 0.005 -
0.8746 14650 0.0063 -
0.8776 14700 0.003 -
0.8806 14750 0.002 -
0.8836 14800 0.0038 -
0.8866 14850 0.0029 -
0.8896 14900 0.003 -
0.8925 14950 0.0038 -
0.8955 15000 0.0028 -
0.8985 15050 0.0046 -
0.9015 15100 0.0043 -
0.9045 15150 0.0034 -
0.9075 15200 0.0023 -
0.9104 15250 0.0041 -
0.9134 15300 0.003 -
0.9164 15350 0.0039 -
0.9194 15400 0.0031 -
0.9224 15450 0.0035 -
0.9254 15500 0.0051 -
0.9284 15550 0.0028 -
0.9313 15600 0.0061 -
0.9343 15650 0.0031 -
0.9373 15700 0.0027 -
0.9403 15750 0.0034 -
0.9433 15800 0.0035 -
0.9463 15850 0.0016 -
0.9493 15900 0.0023 -
0.9522 15950 0.0025 -
0.9552 16000 0.0028 -
0.9582 16050 0.0019 -
0.9612 16100 0.0026 -
0.9642 16150 0.003 -
0.9672 16200 0.0032 -
0.9701 16250 0.0037 -
0.9731 16300 0.004 -
0.9761 16350 0.0033 -
0.9791 16400 0.0038 -
0.9821 16450 0.0032 -
0.9851 16500 0.0039 -
0.9881 16550 0.0029 -
0.9910 16600 0.0034 -
0.9940 16650 0.0034 -
0.9970 16700 0.0023 -
1.0 16750 0.0034 0.1525

Framework Versions

  • Python: 3.13.2
  • SetFit: 1.1.2
  • Sentence Transformers: 4.0.2
  • Transformers: 4.51.0
  • PyTorch: 2.6.0
  • Datasets: 3.5.0
  • Tokenizers: 0.21.1

Citation

BibTeX 

@article{https://doi.org/10.48550/arxiv.2209.11055,
    doi = {10.48550/ARXIV.2209.11055},
    url = {https://arxiv.org/abs/2209.11055},
    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
    title = {Efficient Few-Shot Learning Without Prompts},
    publisher = {arXiv},
    year = {2022},
    copyright = {Creative Commons Attribution 4.0 International}
}
Downloads last month
155
Safetensors
Model size
22.7M params
Tensor type
F32
·
Inference Providers NEW
Text Classification
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for hleAtKeeper/threat-categorization-classifier-setfit

Base model

sentence-transformers/paraphrase-MiniLM-L6-v2
Finetuned
(11)
this model

Evaluation results

View on Papers With Code