Nexusraven 13B - Details

Nexusraven 13B is a large language model developed by Nexusflow, a company specializing in advanced AI solutions. With 13b parameters, it is designed for versatile function calling tasks, offering robust performance across diverse applications. The model operates under the Nexusflowai License Terms (Nexusflowai-LT), ensuring compliance with specific usage guidelines. Its architecture emphasizes efficiency and adaptability, making it suitable for complex computational workflows.

Nexusraven 13B is a fine-tuned version of codellama/CodeLlama-13b-Instruct-hf designed for function calling with high success rates in cybersecurity tools. It offers lower cost and faster inference compared to GPT-4 while maintaining open-source accessibility and commercial viability. The model excels in generalizing to unseen tools and performing effectively in zero-shot settings, making it a versatile choice for dynamic and complex task environments.

Nexusraven 13B has 13b parameters, placing it in the mid-scale category of open-source LLMs, which balances performance and resource efficiency for moderate complexity tasks. Its 100k context length falls into the long context range, enabling effective handling of extended texts but requiring more computational resources. This combination allows the model to manage intricate workflows while maintaining scalability. The 13b parameter size ensures versatility without excessive demands, while the 100k context length supports detailed, extended interactions.

• Parameter Size: 13b
• Context Length: 100k

Nexusraven 13B is a versatile model that could support possible uses in function calling for cybersecurity tools such as CVE/CPE search and VirusTotal integration, enabling automated threat analysis or data retrieval. It might also facilitate code generation and execution via Python functions, offering potential for dynamic scripting or tool development. Commercial applications could leverage its low-cost, high-efficiency capabilities for tasks requiring scalable language model support. These possible uses would need thorough investigation to ensure alignment with specific requirements and constraints.

• Possible use: Function calling for cybersecurity tools like CVE/CPE search and VirusTotal
• Possible use: Code generation and execution via Python functions
• Possible use: Commercial applications requiring low-cost, high-efficiency language model capabilities

Nexusraven 13B could support possible applications in cybersecurity workflows, such as automated threat analysis or data retrieval through function calls for tools like CVE/CPE search and VirusTotal. It might also enable possible uses in code generation and execution via Python functions, offering potential for dynamic scripting or tool development. Commercial applications could leverage its low-cost, high-efficiency capabilities for tasks requiring scalable language model support. Additionally, it might facilitate possible uses in general-purpose task automation, where its 13b parameter size and 100k context length could handle complex, extended interactions. Each of these possible applications would require thorough evaluation and testing to ensure suitability for specific use cases.

• Possible application: Function calling for cybersecurity tools like CVE/CPE search and VirusTotal
• Possible application: Code generation and execution via Python functions
• Possible application: Commercial applications requiring low-cost, high-efficiency language model capabilities
• Possible application: General-purpose task automation with extended context handling

Nexusraven 13B’s medium q4 version requires a GPU with at least 20GB VRAM (e.g., RTX 3090) and 16GB–32GB VRAM for optimal performance, along with 32GB+ system memory and adequate cooling. This configuration balances precision and efficiency for tasks like function calling and code generation. The 13b parameter size and 100k context length demand sufficient computational resources to handle extended interactions.

fp16, q2, q3, q4, q5, q6, q8

Nexusraven 13B is a 13b-parameter model optimized for function calling tasks, particularly in cybersecurity tools and code generation, with a 100k context length for extended interactions. It operates under the Nexusflowai License Terms, offering open-source accessibility and commercial viability while balancing performance and efficiency.

Huggingface Model Page
Ollama Model Page