一、前言
就在前几天开源社区又发布了qwen1.5版本,它是qwen2模型的测试版本。在本篇学习中,将集成vllm实现模型推理加速,现在,我们赶紧跟上技术发展的脚步,去体验一下新版本模型的推理质量。
二、术语
2.1. vLLM
vLLM是一个开源的大模型推理加速框架,通过PagedAttention高效地管理attention中缓存的张量,实现了比HuggingFace Transformers高14-24倍的吞吐量。
2.2. qwen1.5
Qwen1.5是Qwen2的测试版,这是一个基于转换器的纯解码器语言模型,在大量数据上进行了预训练。
In comparison with the previous released Qwen, the improvements include:
- 6 model sizes, including 0.5B, 1.8B, 4B, 7B, 14B, and 72B;
- Significant performance improvement in human preference for chat models;
- Multilingual support of both base and chat models;
- Stable support of 32K context length for models of all sizes
- No need of
trust_remote_code
.