Overview

Kukedlc/NeuralMaxime-7B-slerp is a 7 billion parameter language model developed by Kukedlc. It is a product of a slerp merge using LazyMergekit, combining the strengths of two base models: mlabonne/AlphaMonarch-7B and mlabonne/NeuralMonarch-7B.

Key Capabilities

• General Reasoning: Achieves a strong average score of 76.17 on the Open LLM Leaderboard, indicating robust performance across various benchmarks.
• Benchmark Performance:
  • AI2 Reasoning Challenge (25-Shot): 73.38
  • HellaSwag (10-Shot): 89.18
  • MMLU (5-Shot): 64.44
  • TruthfulQA (0-shot): 77.79
  • Winogrande (5-shot): 84.45
  • GSM8k (5-shot): 67.78

Good For

• Applications requiring solid general-purpose language understanding and generation.
• Tasks benefiting from a model with balanced performance across diverse reasoning and knowledge-based benchmarks.
• Developers looking for a 7B model with competitive scores on the Open LLM Leaderboard, as detailed here.