wsdm_cup_2020_solution
wsdm_cup_2020_solution copied to clipboard
myeclipse
First place solution of WSDM CUP 2020, pairwise-bert, lightgbm
Hello, your dataset is not provided in the project. Could you please contact with me via [email protected]. I'd be grateful if you would provide me your dataset.
你好作者，请问pairwise bert里run和run_online函数有什么区别呢？我看你在主函数里运行的是run_online，请问它俩之间的区别在哪里，以及该运行哪个呢？
你好，我在pairwise bert代码里看到，训练数据集是直接读取的train_pairwise.csv这个文件，请问这个数据集是如何构造的呢，谢谢！
bert pointwise、pairwise 构建正负样本的时候，train是否加入了recall的样本,负样本采样的比例大概有多少
The model training is ok. But when eval or predict, bug occurs! It does not matter the data. I have change train data and dev data, and the bug also...
