Nexusflow/Athene-V2-Chat
Nexusflow/Athene-V2-Chat is a 72.7 billion parameter instruction-tuned causal language model developed by Nexusflow, fine-tuned from Qwen 2.5 72B-Instruct. With a 131,072 token context length, it is optimized for chat, math, and coding tasks, demonstrating performance on par with GPT-4o across various benchmarks. This model excels in instruction following and multi-turn conversations, making it suitable for complex interactive applications.
Loading preview...
Athene-V2-Chat-72B Overview
Athene-V2-Chat-72B is a 72.7 billion parameter open-weights large language model developed by Nexusflow, fine-tuned from Qwen 2.5 72B-Instruct. It is designed for chat-based applications and demonstrates strong performance across various benchmarks, rivaling proprietary models like GPT-4o.
Key Capabilities
- Chat Performance: Achieves performance on par with GPT-4o-0513 in instruction following, longer queries, and multi-turn conversations.
- Mathematical Reasoning: Excels in hard and mathematical categories, outperforming GPT-4o-0513 on Chatbot Arena.
- Coding Proficiency: Shows comparable performance to GPT-4o-0513 in coding tasks.
- Context Length: Features a substantial 131,072 token context window, supporting extensive interactions.
- RLHF Training: Benefits from Reinforcement Learning from Human Feedback (RLHF) to enhance conversational quality and alignment.
Good For
- General Chat Applications: Ideal for building highly capable conversational agents.
- Complex Problem Solving: Suitable for tasks requiring strong mathematical and logical reasoning.
- Code Generation and Analysis: Effective for programming-related queries and code assistance.
- Interactive Systems: Its robust instruction following and multi-turn capabilities make it well-suited for dynamic user interactions.
Top 3 parameter combinations used by Featherless users for this model. Click a tab to see each config.