1、put the given data into /path/to/yizhi/data
data WOZ train.csv val.csv test.csv dataloader models output tools
2、running train&test by sh traintest.sh
**report
采用要求的模型训练,去除stopwords使用bert分词,30轮可收敛,验证f1_macro:0.7886935960841901, f1_micro:0.9342634192480492, loss:0.0007
测试2024-03-26 19:51:57,219 INFO f1_macro:0.7855377539597839, f1_micro:0.9424256724870221, loss:0.0007 观察测试数据集 其中有'''thank 你好,我想吃美食街,帮我推荐一个人均消费在50-100元的餐馆,谢谢。''''对测试效果有一定影响