shibing624 / pytextclassifier

pytextclassifier is a toolkit for text classification. 文本分类,LR,Xgboost,TextCNN,FastText,TextRNN,BERT等分类模型实现,开箱即用。
https://shibing624.github.io/pytextclassifier/
Apache License 2.0
499 stars 74 forks source link

bert 分类报错 #7

Closed songsh closed 2 years ago

songsh commented 2 years ago

bert 中文分类时 ,No module named 'simpletransformers' fasttext 中文分类时,准确率不高,

shibing624 commented 2 years ago

pip install simpletransformers发自我的iPhone------------------ 原始邮件 ------------------发件人: chendeming @.>发送时间: 2022年3月28日 16:39收件人: shibing624/pytextclassifier @.>抄送: Subscribed @.***>主题: Re: [shibing624/pytextclassifier] bert 分类报错 (Issue #7) bert 中文分类时 ,No module named 'simpletransformers' fasttext 中文分类时,准确率不高,

—Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you are subscribed to this thread.Message ID: @.> [ { @.": "http://schema.org", @.": "EmailMessage", "potentialAction": { @.": "ViewAction", "target": "https://github.com/shibing624/pytextclassifier/issues/7", "url": "https://github.com/shibing624/pytextclassifier/issues/7", "name": "View Issue" }, "description": "View this Issue on GitHub", "publisher": { @.***": "Organization", "name": "GitHub", "url": "https://github.com" } } ]

songsh commented 2 years ago

感谢回复,用fasttext 分类时,运行两次,预测结果还不一样,是啥问题。

shibing624 commented 2 years ago

更新了example下的demo,默认只加载the best模型,而不是最后一个epoch的模型。同一个模型预测两次相同文本,结果是一样的。