opencsg/csg-wukong-2b-ultrafeedback-chinese-binarized-lowest

来源：🤗apache-2.02.2B 参数4.3 GBsafetensors✓ 2 个校验和663 天前更新

We apply DPO to opencsg/csg-wukong-2b-smoltalk-chinese, using ultrafeedback-chinese-binarized-lowest as the DPO dataset.
Train 2 epochs, learning rate=5e-6, batchsize=128
This model's evaluation score is higher than that one trained on ultrafeedback-chinese-binarized

为此模型做种

有文件？提交种子。

或

用一条命令打包文件

已有完整文件并有权分享？请替换示例文件夹路径。脚本会校验文件、制作种子并提交磁力链接和校验和。请保持种子客户端做种。从设置页面获取社区密钥，运行前可先阅读脚本。

curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo opencsg/csg-wukong-2b-ultrafeedback-chinese-binarized-lowest ./model-folder