adamo1139/toxic-dpo-natural-v5
收藏Hugging Face2024-05-03 更新2024-06-12 收录
下载链接：
https://hf-mirror.com/datasets/adamo1139/toxic-dpo-natural-v5
下载链接
链接失效反馈官方服务：
资源简介：
---
license: other
license_name: other
license_link: LICENSE
---
I mixed in toxid-dpo-natural-v4 and rawrr v2-1 stage 2 with chosen field from original no_robots and got myself toxic-dpo-natural-v5. Goal is to avoid overfitting via DPO to a specific type of instruct, and instead just DPO the model to be more open to answering and also answer like a human being. ~~We'll see whether this works.~~\
I trained Yi 34B with this dataset and ORPO, it does work very nicely so far!
提供机构：
adamo1139
原始信息汇总
数据集概述
数据集名称
- 名称: toxic-dpo-natural-v5
数据集来源
- 混合来源: 结合了toxid-dpo-natural-v4、rawrr v2-1 stage 2以及原始数据集no_robots中选定的字段。
数据集目的
- 目的: 避免模型过度拟合于特定指令类型，而是通过DPO（数据处理优化）使模型更开放地回答问题，并模拟人类回答方式。
数据集应用
- 应用: 用于训练Yi 34B模型，结合ORPO方法，目前效果良好。
许可信息
- 许可类型: other
- 许可文件: LICENSE
