Skip to content

文心等预训练模型以及下游任务的可解释性 #849

@LangDaoAI

Description

@LangDaoAI

paddlenlp支撑的预训练模型以及相关下游任务, 迫切需要可解释性作为工业落地的另一个支撑, 比如基于预训练的多分类模型, paddle是否提供了分类的可解释性, 看到https://github.com/PaddlePaddle/InterpretDL 似乎是这样一个目标,但是与paddlenlp并没有集成, 请给与一些建议,谢谢!

Metadata

Metadata

Assignees

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