Model Overview

kuno-royale-v3-7b is an experimental 7 billion parameter language model developed by core-3. It is a product of a slerp merge combining two distinct models: SanjiWatsuki/Kunoichi-DPO-v2-7B and eren23/ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v3. This merging strategy aims to leverage the strengths of its constituent models.

Performance Benchmarks

Evaluated on the Open LLM Leaderboard, kuno-royale-v3-7b achieved an average score of 74.88. Key performance metrics include:

• AI2 Reasoning Challenge (25-Shot): 71.76
• HellaSwag (10-Shot): 88.23
• MMLU (5-Shot): 65.06
• TruthfulQA (0-shot): 71.13
• Winogrande (5-shot): 82.32
• GSM8k (5-shot): 70.81

These scores indicate its proficiency in various tasks, including reasoning, common sense, and general knowledge. More detailed evaluation results are available on the Open LLM Leaderboard.

Usage Considerations

As an experimental merge, kuno-royale-v3-7b is suitable for developers looking to explore models with combined characteristics from its base components. Its 7B parameter size and 4096-token context window make it a viable option for applications requiring a balance between performance and computational resources.