Undi95/ReasoningEngine: Enhanced Reasoning Model

Undi95/ReasoningEngine is a 13 billion parameter language model that integrates components from two established models to specialize in reasoning tasks. It is primarily based on Stability AI's StableBeluga-13B, a fine-tuned LLaMA 2 model known for its strong general capabilities.

Key Enhancements

• Reasoning Adapter Integration: The model incorporates a 0.42 weight from the reasoning adapter of jondurbin/airoboros-lmoe-13b-2.1. This specific integration is designed to boost the model's ability to handle complex logical problems and inferential tasks.
• Parameter Count: With 13 billion parameters, it offers a balance between performance and computational efficiency for specialized reasoning.
• Context Length: The model supports a context length of 4096 tokens, allowing it to process and reason over moderately long inputs.

Good For

• Logical Problem Solving: Ideal for applications requiring the model to deduce, infer, and solve problems based on provided information.
• Analytical Tasks: Suitable for scenarios where enhanced analytical capabilities are beneficial, such as data interpretation or complex query resolution.
• Specialized Reasoning Applications: Developers looking for a model with a focused improvement in reasoning over its base components.