Identifying security vulnerabilities in source code
Limitations:
May not identify all vulnerabilities if multiple are present, Prone to false positives, Results should be verified by human experts, Affected by code complexity and context
Considerations:
Should be integrated within a broader security review process.
Supported Languages
Go (High), Python (High), C (High), C++ (High), Fortran (High), Ruby (High), Java (High), Kotlin (High), C# (High), PHP (High), Swift (High), JavaScript (High), TypeScript (High)
Training Details
Data Sources:
Proprietary dataset for vulnerability detection
Methodology:
Fine-tuned for vulnerability detection; trained using Parameter-Efficient Fine-Tuning (PEFT).
Hardware Used:
A100 GPUs
Input Output
Input Format:
Programming code snippets
Accepted Modalities:
text
Output Format:
Textual analysis of vulnerabilities and quality issues
Performance Tips:
Best performance with appropriate input code snippet length.
Note: green Score (e.g. "73.2") means that the model is better than rootxhacker/CodeAstra-7B.
Rank the CodeAstra 7B Capabilities
๐ Have you tried this model? Rate its performance. This feedback would greatly assist ML community in identifying the most suitable model for their needs. Your contribution really does make a difference! ๐
Instruction Following and Task Automation
Factuality and Completeness of Knowledge
Censorship and Alignment
Data Analysis and Insight Generation
Text Generation
Text Summarization and Feature Extraction
Code Generation
Multi-Language Support and Translation
What open-source LLMs or SLMs are you in search of? 52721 in total.