分布式训练之后的多个模型pt文件如何转化为可推理的bin模型文件？

[acadaiaca]

分布式训练之后的模型格式是： zero_pp_rank_0_mp_rank_00_model_states.pt zero_pp_rank_0_mp_rank_00_optim_states.pt zero_pp_rank_1_mp_rank_00_model_states.pt zero_pp_rank_1_mp_rank_00_optim_states.pt

请问如何转化成适用于llama_inference推理的模型格式？https://github.com/fengyh3/llama_inference

[ydli-ai]

保存模型的路径里有一个转换脚本

---

发件人: acadaiaca @.> 发送时间: Sunday, April 23, 2023 7:10:14 PM 收件人: ydli-ai/Chinese-ChatLLaMA @.> 抄送: Subscribed @.***> 主题: [ydli-ai/Chinese-ChatLLaMA] 分布式训练之后的多个模型pt文件如何转化为可推理的bin模型文件？ (Issue #41)

分布式训练之后的模型格式是： zero_pp_rank_0_mp_rank_00_model_states.pt zero_pp_rank_0_mp_rank_00_optim_states.pt zero_pp_rank_1_mp_rank_00_model_states.pt zero_pp_rank_1_mp_rank_00_optim_states.pt

请问如何转化成适用于https://github.com/fengyh3/llama_inference推理的模型格式？https://github.com/fengyh3/llama_inference%E6%8E%A8%E7%90%86%E7%9A%84%E6%A8%A1%E5%9E%8B%E6%A0%BC%E5%BC%8F%EF%BC%9F

― Reply to this email directly, view it on GitHubhttps://github.com/ydli-ai/Chinese-ChatLLaMA/issues/41, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AE3SPV7LEWFAAPHH5Z2VLJDXCUEZNANCNFSM6AAAAAAXIOZFLQ. You are receiving this because you are subscribed to this thread.Message ID: @.***>