nlpguy/Westgate: A Merged 7B Language Model

nlpguy/Westgate is a 7 billion parameter language model developed by nlpguy through the strategic merging of two pre-trained models: jsfs11/TurdusTrixBeagle-DARETIES-7B and senseable/garten2-7b. This model was created using the SLERP merge method, a technique designed to combine the strengths of its constituent models.

Key Capabilities & Performance

Evaluated on the Open LLM Leaderboard, Westgate demonstrates strong performance across a range of benchmarks, achieving an average score of 73.84. Notable results include:

• AI2 Reasoning Challenge (25-Shot): 71.42
• HellaSwag (10-Shot): 88.14
• MMLU (5-Shot): 65.11
• TruthfulQA (0-shot): 62.59
• Winogrande (5-shot): 85.71
• GSM8k (5-shot): 70.05

These scores indicate proficiency in common sense reasoning, language understanding, and mathematical problem-solving. The model's 4096-token context length supports processing moderately sized inputs.

Use Cases

Given its balanced performance across various reasoning and language understanding tasks, nlpguy/Westgate is well-suited for applications requiring a general-purpose 7B model, particularly where a blend of capabilities from its merged components is beneficial. It can be applied to tasks such as:

• General text generation and completion
• Question answering
• Reasoning-based tasks
• Summarization