Llama3天瘦成Mamba！推理速度提升1.5倍
Mamba 团队的研究成果集中于将大型 Transformer 模型 Llama “蒸馏”为 Mamba，通过设计新型推理解码算法，显著提高了推理速度。研究目标旨在利用 Llama 的丰富知识，同时降低从零开始训练大型模型的高成本。研究团队结合了渐进式蒸馏、监督微调和定向偏好优化等方法，将 Zephyr-7B 和 Llama-38B 成功转换为线性 RNN 模型，性能与原始模型相当。关键创新在于为 Mamba 设计的全新算法，结合硬件特性实现基于 Mamba 的推测解码，实现速度与性能的双重提升。研究中，通过逐步替换注意力层、应用监督微调和用户偏好优化，以及引入推测解码算法，研究人员在保持模型性能的同时，实现了高达1.5倍的推理加速。整个过程在8卡80G A100上运行3到4天，展现了高效的训练效率。这项研究为未来模型的推理速度和性能提升提供了新思路。
%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
