Open codezZing opened 3 weeks ago
Hi,
Thanks for your interest! I tested BAT mainly based on the codebase of GraphENS: https://github.com/JoonHyung-Park/GraphENS. Due to the coupled implementation design of many modules in there, the code looks very messy (with lots of if-else in the main function) if combined with baselines.
I don't have the bandwidth to refactor and clean the code for all the baselines, but I can provide you my working code scripts. Please send an email to liu326@illinois.edu, I will attach the code to my reply.
Thanks
Thank you very much for your reply. Could you please send me a copy of the code as an attachment?
周zZ @.***
------------------ 原始邮件 ------------------ 发件人: "Zhining @.>; 发送时间: 2024年11月22日(星期五) 中午11:29 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [ZhiningLiu1998/BAT] Inquiry about Experiment Results with BAT and Other IGL Baselines on Different GNN Backbones (Issue #1)
Hi, 你好
Thanks for your interest! I tested BAT mainly based on the codebase of GraphENS: https://github.com/JoonHyung-Park/GraphENS. Due to the coupled implementation design of many modules in there, the code looks very messy (with lots of if-else in the main function) if combined with baselines. 感谢您的关注!我主要基于 GraphENS 的代码库测试 BAT:https://github.com/JoonHyung-Park/GraphENS。由于其中许多模块的耦合实现设计,如果与基线结合使用,代码看起来非常混乱(主函数中有很多 if-else)。
I don't have the bandwidth to refactor and clean the code for all the baselines, but I can provide you my working code scripts. Please send an email to @., I will attach the code to my reply. 我没有足够的带宽来重构和清理所有基线的代码,但我可以为您提供我的工作代码脚本。请发送电子邮件给 @.,我会在回复中附上代码。
Thanks 谢谢
— Reply to this email directly, view it on GitHub, or unsubscribe. 直接回复此电子邮件,在 GitHub 上查看,或取消订阅。 You are receiving this because you authored the thread. 您收到此消息是因为您编写了该线程。 Message ID: @.***>
Hi,
I have been exploring the repository and noticed that the provided files only include experimental results for BAT combined with various backbones. However, I am particularly interested in seeing the results of BAT in combination with other IGL baselines, specifically across different GNN backbones.
Could you please provide guidance on how I can access or replicate these additional experimental results, or if they are available in any other form?
Thanks in advance for your help!