Open jianhuigogo opened 7 months ago
I also encountered this problem. Have you solved it?
你现在跑成功了吗?
From: Jin Ke @.> Sent: 30 June 2024 17:03 To: mufusu21/REMIT @.> Cc: jianhui.Jiang @.>; Author @.> Subject: Re: [mufusu21/REMIT] the result is 2.2 i don't know what went wrong? (Issue #2)
也是相同的问题,仔细看了看数据处理也没啥问题,原来mate网络效果2.5左右,感觉通过RL训练方式优化mate网络提升到1.2难度还挺大。
― Reply to this email directly, view it on GitHubhttps://github.com/mufusu21/REMIT/issues/2#issuecomment-2198489068, or unsubscribehttps://github.com/notifications/unsubscribe-auth/A2ZKV5XM2KVQ4PXU4BYUDPTZJ7CUTAVCNFSM6AAAAABF343BXSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCOJYGQ4DSMBWHA. You are receiving this because you authored the thread.Message ID: @.***>
+1
I notice that the pkl file contain user_index and user embedding. I wonder different data process may cause different user id to user index's map function?That means the data in author's paper may be different from ours, the embedding pkl can not represent the same user?