|
Download README.md from Gamestatue/adaption-cyber-sec-code-instruct: direct link, hf CLI and curl.
- Browser
- Download file 2.2 kB
-
https://huggingface.co/datasets/Gamestatue/adaption-cyber-sec-code-instruct/resolve/main/README.md
- Command line
-
hf download hf://datasets/Gamestatue/adaption-cyber-sec-code-instruct/README.md
-
curl -L -o README.md https://huggingface.co/datasets/Gamestatue/adaption-cyber-sec-code-instruct/resolve/main/README.md
2.2 kB
| annotations_creators: [] | |
| language: | |
| - en | |
| - kn | |
| - th | |
| language_creators: [] | |
| license: [] | |
| multilinguality: | |
| - multilingual | |
| pretty_name: 'adaption-cyber_sec_code_instruct' | |
| size_categories: | |
| - 1K<n<10K | |
| source_datasets: [] | |
| tags: | |
| - adaption | |
| - instruction-tuning | |
| - code | |
| - governance | |
| - technology | |
| task_categories: [] | |
| task_ids: [] | |
|  | |
| This dataset is a remastered version prepared using [Adaption's](https://adaptionlabs.ai/app/auth) Adaptive Data platform. | |
| # adaption-cyber_sec_code_instruct | |
| This dataset consists of instruction-response pairs focused on cybersecurity incident analysis, threat detection, and secure software engineering. Tasks span multiple programming languages, including Python, JavaScript, C, SQL, and Rust, covering vulnerability assessment, remediation, and log investigation. Prompts range from straightforward coding tasks to complex incident scenarios with missing context, evaluating technical reasoning and cross-language handling. | |
| ### Dataset size | |
| There are 7,959 data points in this dataset. This is an instruction tuning dataset. | |
| ### Quality of Remastered Dataset | |
| The final quality is A, with a relative quality improvement of 2.2%. | |
| ### Domain | |
| - Code (92%) | |
| - Governance (6%) | |
| - Technology (2%) | |
| ### Language | |
| - English (12%) | |
| - Kannada (10%) | |
| - Thai (8%) | |
| ### Tone | |
| - Technical (90%) | |
| - Educational (4%) | |
| - Urgent (2%) | |
| ### Evaluation Results | |
| - **Quality Gains:** | |
| <img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/15a15c26-97b2-49be-982a-aed35961c357.png" alt="QualityGains" style="max-width: 50%; display: block; margin-left: auto; margin-right: auto;" /> | |
| - **Grade Improvement:** | |
| <img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/0056f360-6f75-4cde-b0e2-d2b7d92f53b6.png" alt="Grade" style="max-width: 50%; display: block; margin-left: auto; margin-right: auto;" /> | |
| - **Percentile Chart:** | |
| <img src="https://proteus-prod-public.s3.us-east-1.amazonaws.com/temp/1386de99-ea56-4aeb-86f7-8be194ab2627.png" alt="Percentile Chart" style="max-width: 50%; display: block; margin-left: auto; margin-right: auto;" /> | |