A fine-tuned variant of Meta's Llama 3.1 8B Instruct, adapted by pearl-ai for purposes not fully detailed in available documentation. It carries the core traits of the base model — reasonable instruction-following and a large 128K context window — while potentially incorporating pearl-ai's specific adjustments. Details about what distinguishes this variant from the base model are limited.