WizardLM-13B-V1.2
| Property | Value |
|---|---|
| Base Model | Llama-2 13B |
| License | Llama 2 License |
| MT-Bench Score | 7.06 |
| AlpacaEval Score | 89.17% |
| Paper | WizardLM Paper |
What is WizardLM-13B-V1.2?
WizardLM-13B-V1.2 is an advanced language model built on the Llama-2 13B architecture, specifically designed to handle complex instructions and engage in detailed multi-turn conversations. This version represents a significant improvement over its predecessors, demonstrating enhanced capabilities in following intricate prompts and generating high-quality responses.
Implementation Details
The model implements the Vicuna-style prompt format and is optimized for multi-turn conversations. It utilizes a specific prompt structure that begins with a system description defining the interaction between a curious user and an AI assistant.
- Built on Llama-2 13B architecture
- Supports multi-turn conversations
- Implements Vicuna-style prompting
- Requires specific formatting for optimal performance
Core Capabilities
- Achieves 7.06 on MT-Bench Leaderboard
- Scores 89.17% on AlpacaEval Leaderboard
- Reaches 101.4% on WizardLM Eval
- 36.6 pass@1 on HumanEval
- Provides detailed and polite responses to complex queries
Frequently Asked Questions
Q: What makes this model unique?
WizardLM-13B-V1.2 stands out for its exceptional ability to follow complex instructions and its improved performance metrics compared to previous versions. It achieves state-of-the-art results on multiple benchmarks while maintaining high-quality, coherent responses.
Q: What are the recommended use cases?
The model excels in scenarios requiring detailed responses, complex problem-solving, and multi-turn conversations. It's particularly well-suited for applications needing sophisticated instruction following and natural language understanding.
