Llama 3 8B Lunaris

An 8B parameter model based on Llama 3, created through merging multiple models to balance creative and logical capabilities.

Strengths

Handles both creative tasks and reasoning-based workloads through its multi-model merge approach, making it versatile across different types of problems.

Best for

General-purpose tasks that mix creative writing, roleplay, and logical reasoning without requiring extremely long context.

Limitations

The 8K token context window limits it for processing long documents or maintaining extended conversations; smaller models typically lag on complex reasoning benchmarks.