Skip to content

基座模型的通用能力有下降吗? #12

Description

@xiexie123

nice work!

有2个问题非常期待作者的分享:

  1. 因为paper是围绕memory这部分有很多实验。但是好像没有看到M+ 和 基座模型(Llama-3.1-8B)在通识能力上的对比。即,引入了memory这部分,加上3个stage的后训练后得到的M+ 模型的通识能力如何。就是作为用户,我既希望模型很强又希望模型拥有“记忆”。

  2. 目前训练得到的M+模型的"memory和retrieve"的能力是“通用”的吗?即 用户可以直接使用。还是跟训练数据有baise呢,比如医学知识或者特别小众领域的知识长文档, M+的memory指标都能稳定在线?

再次感谢作者的开源

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions