Llama-Spark: A Conversational AI Model by Arcee.ai

Llama-Spark is an 8 billion parameter conversational AI model developed by Arcee.ai, designed to be a leading performer in its size class. It is built on the Llama-3.1-8B base model and further enhanced by fine-tuning with Arcee.ai's proprietary Tome Dataset, then merged with Llama-3.1-8B-Instruct. This combination results in a highly capable model for dialogue-based interactions.

Key Capabilities

• Conversational Excellence: Optimized for engaging in natural, informative, and coherent conversations.
• Strong Foundation: Leverages the robust architecture of Llama-3.1-8B.
• Continuous Improvement: Arcee.ai is committed to updating and improving Spark as new base models become available to maintain its competitive edge.

Intended Uses

Llama-Spark is specifically designed for conversational AI applications, making it suitable for:

• Chatbots
• Virtual assistants
• Dialogue systems

Performance Metrics

Evaluations on the Open LLM Leaderboard show an average score of 24.90, with notable results in IFEval (0-Shot) at 79.11 and MMLU-PRO (5-shot) at 30.23. Detailed results are available on the Open LLM Leaderboard.