kongqi404 / timegrad

17 stars 4 forks source link

报错应该是gluonts版本问题 #1

Closed endlessfireable closed 1 year ago

endlessfireable commented 1 year ago

感谢作者的注释,看到你的报错应该是版本问题,将gluonts版本改成gluonts=10.0.0.0即可,不知道作者是否还在进行timegrad的源码研究,不知道是否能共同交流

kongqi404 commented 1 year ago

对的,但是我本来是准备迁移到latest版本的gluonts或者用纯pytorch重写下timegrad,现在只完成了概率输出的重写,diffusion的重写还没进行

endlessfireable commented 1 year ago

想请问下纯pytorch的部分概率重写是可以开源的吗,gluonts封装的我晕了,我想魔改一下这个模型做时序相关的其他工作,不知道您的代码可以开源吗

---原始邮件--- 发件人: @.> 发送时间: 2023年3月16日 18:44:57 收件人: @.>; 抄送: @.**@.>; 主题: Re: [kongqi404/timegrad] 报错应该是gluonts版本问题 (Issue #1)

对的,但是我本来是准备迁移到latest版本的gluonts或者用纯pytorch重写下timegrad,现在只完成了概率输出的重写,diffusion的重写还没进行

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

kongqi404 commented 1 year ago

可以,等我整理下就可以开源,实现很简单

endlessfireable @.***> 于 2023年3月17日周五 19:05写道:

想请问下纯pytorch的部分概率重写是可以开源的吗,gluonts封装的我晕了,我想魔改一下这个模型做时序相关的其他工作,不知道您的代码可以开源吗

---原始邮件--- 发件人: @.> 发送时间: 2023年3月16日 18:44:57 收件人: @.>; 抄送: @.**@.>; 主题: Re: [kongqi404/timegrad] 报错应该是gluonts版本问题 (Issue #1)

对的,但是我本来是准备迁移到latest版本的gluonts或者用纯pytorch重写下timegrad,现在只完成了概率输出的重写,diffusion的重写还没进行

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

— Reply to this email directly, view it on GitHub https://github.com/kongqi404/timegrad/issues/1#issuecomment-1473666731, or unsubscribe https://github.com/notifications/unsubscribe-auth/AHXXZIR7Y55YD4Z77YP4HX3W4RAOHANCNFSM6AAAAAAV44KQ2E . You are receiving this because you commented.Message ID: @.***>

kongqi404 commented 1 year ago

想请问下纯pytorch的部分概率重写是可以开源的吗,gluonts封装的我晕了,我想魔改一下这个模型做时序相关的其他工作,不知道您的代码可以开源吗 ---原始邮件--- 发件人: @.> 发送时间: 2023年3月16日 18:44:57 收件人: @.>; 抄送: @.**@.>; 主题: Re: [kongqi404/timegrad] 报错应该是gluonts版本问题 (Issue #1) 对的,但是我本来是准备迁移到latest版本的gluonts或者用纯pytorch重写下timegrad,现在只完成了概率输出的重写,diffusion的重写还没进行 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

here is prob forecasting code:https://gist.github.com/kongqi404/336a8031901a3cebc05585b78fa650ef