misc 0 0 0

TahaGorji/BlackHat-KaliLinux-CyberSecurity-dataset-Instruct

published by @TahaGorji

Alignment dataset tracked in the /datasets sub-catalog.

Lifetime downloads
431
Last 30 days
225
Likes
3
Size
449 KB
Created on HF
2026-07-02
Age
3 months ago

Description

KaliLinux Dataset

A cleaned instruction-following dataset focused on penetration testing, offensive security tools, and red team techniques.

Dataset Info

Field

Value

File

cleaned_dataset.csv

Total rows

434

File size

~445 KB

Format

CSV (instruction, input, output)

Language

Language

Count

Persian (فارسی)

173

English

261

Output Stats

Words

Average

105

Minimum

15

Maximum

331… See the full description on the dataset page: https://huggingface.co/datasets/TahaGorji/BlackHat-KaliLinux-CyberSecurity-dataset-Instruct.

Tags

language:enlanguage:falicense:apache-2.0region:ussecuritykalilinux

README current version from Hugging Face


license: apache-2.0
language:

  • en
  • fa
    tags:
  • security
  • kalilinux

KaliLinux Dataset

A cleaned instruction-following dataset focused on penetration testing, offensive security tools, and red team techniques.


Dataset Info

Field Value
File cleaned_dataset.csv
Total rows 434
File size ~445 KB
Format CSV (instruction, input, output)

Language

Language Count
Persian (فارسی) 173
English 261

Output Stats

Words
Average 105
Minimum 15
Maximum 331

Topics

  • Port scanning (nmap, masscan)
  • Web exploitation (SQLMap, Burp Suite, Gobuster)
  • Password attacks (Hydra, Hashcat, John)
  • Post-exploitation & privilege escalation
  • Persistence & backdoors
  • Network attacks (MitM, DNS tunneling, WiFi)
  • Payload generation (msfvenom, Cobalt Strike)
  • OSINT & reconnaissance
  • Active Directory attacks (DCSync, Golden Ticket)
  • Log & forensic evasion

Cleaning Applied

  • Merged 3 source files (dataset.csv, d2.csv, g.csv)
  • Removed duplicate instruction,input,output header rows
  • Removed rows with output shorter than 15 words
  • Removed 379 duplicate rows
  • Removed Chinese characters from all fields
  • Fixed malformed CSV escaping (\" → standard "")

دیتاست

یک دیتاست تمیزشده برای fine-tuning مدل‌های زبانی در حوزه امنیت تهاجمی و تست نفوذ.

فیلد مقدار
فایل cleaned_dataset.csv
تعداد ردیف ۴۳۴
حجم ~۴۴۵ کیلوبایت
ساختار CSV سه‌ستونی

زبان: ۱۷۳ ردیف فارسی، ۲۶۱ ردیف انگلیسی

میانگین طول خروجی: ۱۰۵ کلمه (حداقل ۱۵، حداکثر ۳۳۱)

README history 5 revisions

Every night we snapshot the README of every dataset in the catalog. When the SHA changes we archive the new version and diff it against the last. This is the evolving thought record of the alignment-data field: what the author decided to say about the corpus, and how that framing shifted over time.

  1. 2026-07-02Upload cleaned_dataset.csv0a962681.8 KB
    Loading...
  2. 2026-07-02Delete dataset.csva2e8b861.8 KB
    Loading...
  3. 2026-07-02Update README.mdbb841201.8 KB
    Loading...
  4. 2026-07-02Upload dataset.csv6e600cc28 B
    Loading...
  5. 2026-07-02initial commitb8c497028 B
    Loading...

Used by abliterated models none yet

No abliterated models in our catalog have declared this dataset in their YAML metadata yet. This can mean the dataset is used but not declared, is used outside abliteration workflows, or is new. The link will populate automatically as models are indexed.

Community discussions 1 thread · 1 comments

Threads posted by the community on Hugging Face: bug reports, questions about usage, pull requests on the README, license clarifications. Click any thread to expand and read the full conversation inline. Newest first.

  1. 2026-09-18PRUpdate README.mdclosed1 💬#1
    Loading...

Related datasets same stage · misc

Read further

Metadata

License
apache-2.0

Repository

Download the dataset

0

Copy any of these snippets into your notebook or terminal. All three fetch directly from Hugging Face using your own credentials.

Recommended: datasets library (Python)
from datasets import load_dataset
ds = load_dataset("TahaGorji/BlackHat-KaliLinux-CyberSecurity-dataset-Instruct")
Raw snapshot (Python)
from huggingface_hub import snapshot_download
snapshot_download(repo_id="TahaGorji/BlackHat-KaliLinux-CyberSecurity-dataset-Instruct", repo_type="dataset")
git clone (requires git-lfs)
git clone https://huggingface.co/datasets/TahaGorji/BlackHat-KaliLinux-CyberSecurity-dataset-Instruct
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration