Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

causal serial 在LifeEST复现 #9

Open
jinyuhang opened this issue Nov 19, 2021 · 6 comments
Open

causal serial 在LifeEST复现 #9

jinyuhang opened this issue Nov 19, 2021 · 6 comments

Comments

@jinyuhang
Copy link

请问,有人复现了 causal serial 在 LifeEST上的效果没?我尝试了各种方法都不得行。

PS:我用pytroch重写了PeterRec_noncau_parallel_classifier.py这个文件,只不过是negative sample 设置为99

@fajieyuan
Copy link
Owner

Hi,源代码PeterRec_noncau_parallel_classifier.py 有问题吗?我记得当时都应该测试过

@jinyuhang
Copy link
Author

Hi,源代码PeterRec_noncau_parallel_classifier.py 有问题吗?我记得当时都应该测试过

我用pytorch参考PeterRec_noncau_parallel_classifier.py这个代码,写了causal_serial_classifier,在LifeEST数据上进行实验,复现不出论文Table2的实验结果。劳烦指正,我的复现如下:

(一)预训练NextitNet模型
代码:参考NextitNet_TF_Pretrain_topk.py
数据集:lifestatus_pretrain_desent.csv
参数: lr 0.001, batchsize 32, epoch 6, splite_rate 0.1, embedding-dim 256, negtive-smple 99, 使用sample softmaxt
'cardinality': 1 'dilations': [1, 4, 1, 4, 1, 4, 1, 4, ],

(二) 微调
数据集: lifestatus_finetune_desent.csv
数据划分: 训练集 0.7 验证集 0.03 测试集 0.27
medel_patch植入: 2 mp serial
训练方式:PeterRec_noncau_parallel_classifier.py的训练方式,不过negtive_samples 设置成99
epoch:80

(3) 编程语言:pytorch
(4) 复现结果:HR@5 0.538(论文中为0.610)

我弄了好久,都没弄出来,劳烦前辈指正一下,若是方便能不能微信沟通(本人微信, 875526037),非常感谢
在RecCold上复现出来了结果,结果略好与论文的。

@jinyuhang jinyuhang reopened this Nov 19, 2021
@fajieyuan
Copy link
Owner

论文代码可以跑一下

@jinyuhang
Copy link
Author

论文代码可以跑一下

就按照里面的默认的参数?

@fajieyuan
Copy link
Owner

论文里的参数哈

@jinyuhang
Copy link
Author

论文里的参数哈

好的😊

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants