-
Notifications
You must be signed in to change notification settings - Fork 588
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
[🐛BUG] LightGCN在ml-100k数据集上性能不佳 #2031
Comments
请你的lightgcn训练样本数和验证样本数正常吗,我跑lightgcn的时候,训练集的数量远小于验证机和测试集的数量 |
你是否设置了相关参数来划分训练集、验证集和测试集? |
model configembedding_size: 32 dataset configfield_separator: "\t" #指定数据集field的分隔符 training settingsepochs: 500 #训练的最大轮数 val_interval: 这是我的参数,请大佬指教一下 |
您是这样的吗 |
可以把 stopping_step 增大,或者关掉试试 |
我使用如下的yaml文件:
tmodel settings
embedding_size: 64 # (int) The embedding size of users and items.
n_layers: 2 # (int) The number of layers in LightGCN.
reg_weight: 1e-05 # (float) The L2 regularization weight.
training settings
stopping_step: 10 #控制训练收敛的步骤数,在该步骤数内若选取的评测标准没有什么变化,就可以提前停止了
evalution settings
split_ratio: [0.8,0.1,0.1] #切分比例
metrics: ["Recall", "MRR","NDCG","Hit","Precision","MAP", "GAUC","ItemCoverage","AveragePopularity","GiniIndex","ShannonEntropy","TailPercentage"] #评测标准
topk: [10] #评测标准使用topk,设置成10评测标准就是["Recall@10", "MRR@10", "NDCG@10", "Hit@10", "Precision@10"]
valid_metric: Precision@10 #选取哪个评测标准作为作为提前停止训练的标准
eval_batch_size: 4096 #评测的batch_size
运行得出的结果
Precision@10:0.1716
这项指标远低于预期
请问是我参数文件使用错误?还是伯乐使用的数据集或者指标公式不同?
The text was updated successfully, but these errors were encountered: