Skip to content

GRPO训练使用ref_adapters时如何合并LoRA #7377

@RandomCoins

Description

@RandomCoins

我用sft后的lora开始训练GRPO,使用swift export进行merge时,是直接merge训练后的lora,还是先merge sft的lora,再merge grpo的lora

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions