Model Overview

Gille/StrangeMerges_49-7B-dare_ties is a 7 billion parameter language model developed by Gille. It is a product of merging three distinct 7B models: Gille/StrangeMerges_32-7B-slerp, AurelPx/Percival_01-7b-slerp, and louisbrulenaudet/Pearl-7B-slerp. This merge was performed using the dare_ties method via LazyMergekit, indicating an advanced approach to combining model strengths.

Key Capabilities & Performance

This model exhibits strong performance across a variety of benchmarks, as evaluated on the Open LLM Leaderboard. It achieves an average score of 75.50, with notable results in:

• AI2 Reasoning Challenge (25-Shot): 72.35
• HellaSwag (10-Shot): 88.30
• MMLU (5-Shot): 64.31
• TruthfulQA (0-shot): 74.70
• Winogrande (5-shot): 83.74
• GSM8k (5-shot): 69.60

These scores suggest a well-rounded model capable of handling tasks requiring common sense reasoning, factual knowledge, and logical problem-solving.

Good For

• General-purpose text generation and understanding.
• Applications requiring robust performance in multiple-choice reasoning and question answering.
• Use cases where a 7B parameter model with strong benchmark performance is desired.