prithivMLmods/Cygnus-II-14B

TEXT GENERATIONConcurrency Cost:1Model Size:14.8BQuant:FP8Ctx Length:32kPublished:Mar 2, 2025License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

prithivMLmods/Cygnus-II-14B is a 14.8 billion parameter language model based on the Qwen 2.5 architecture, developed by prithivMLmods. It is optimized for general-purpose reasoning and answering, excelling in contextual understanding, logical deduction, and multi-step problem-solving. Fine-tuned with a long chain-of-thought reasoning model, it supports a 32768 token context length and offers multilingual proficiency across 29 languages. This model is designed for applications requiring enhanced reasoning, structured responses, and conversational intelligence.

Loading preview...

Overview

Cygnus-II-14B is a 14.8 billion parameter model built on the Qwen 2.5 14B architecture, developed by prithivMLmods. It is specifically designed to enhance reasoning capabilities, excelling in contextual understanding, logical deduction, and multi-step problem-solving. The model has been fine-tuned using a long chain-of-thought reasoning approach and specialized datasets to improve comprehension, structured responses, and conversational intelligence.

Key Capabilities

  • Enhanced General Knowledge: Provides broad knowledge across various domains for accurate and coherent responses.
  • Improved Instruction Following: Advanced understanding and execution of complex instructions, generating structured outputs.
  • Versatile Adaptability: Resilient to diverse prompts, handling a wide range of topics and conversation styles.
  • Long-Context Support: Supports up to 128K tokens for input context and can generate up to 8K tokens in a single output.
  • Multilingual Proficiency: Supports over 29 languages, including English, Chinese, French, Spanish, and more.

Performance Highlights

Evaluations on the Open LLM Leaderboard show an Average score of 40.53%.

  • IFEval (0-Shot): 61.84%
  • BBH (3-Shot): 52.14%
  • MMLU-PRO (5-shot): 48.78%

Good For

  • General-Purpose Reasoning: Assisting with logical reasoning, diverse question answering, and general knowledge problems.
  • Educational and Informational Assistance: Providing explanations, summaries, and research-based responses.
  • Conversational AI and Chatbots: Building intelligent agents requiring contextual understanding and dynamic responses.
  • Multilingual Applications: Supporting global communication, translations, and multilingual content generation.
  • Structured Data Processing: Analyzing and generating structured outputs like tables and JSON.
  • Long-Form Content Generation: Creating extended responses such as articles, reports, and guides while maintaining coherence.