search

raghavchalapathy / Bidirectional-LSTM-CRF-for-Clinical-Concept-Extraction

Bidirectional LSTM-CRF for Clinical Concept Extraction using i2b2-2010 data
89 stars 39 forks source link

Bug in calculating cost? #1

Open zjy-ucas opened 7 years ago

zjy-ucas commented 7 years ago

Hi I employed the blstm-crf model for Chinese named entity recognition,it worked but results was no better than the one without crf, and I got real_path_score larger than all_path score,is that possible?

zjy-ucas commented 7 years ago

I have fixed the bug,thanks for your code!

Zhangzirui commented 7 years ago

@zjy-ucas 能问一个问题吗，就是将这个程序用作中文命名实体识别，那个char-dim是如何处理的。如果将源程序中char-dim和char_lstm_dim置0，那么程序就会报错

Zhangzirui commented 7 years ago

@zjy-ucas 您愿意把您对中文命名实体是程序上传到github上吗，我想看看中文是怎么处理的

zjy-ucas commented 7 years ago

@Zhangzirui 我参考了他的crf层，中文和英文的处理稍有不同，有一些bug还要修改，改好了会发，可以留个邮箱先给你

Zhangzirui commented 7 years ago

谢谢啦，太开心能收到您的回复了。

------------------ 原始邮件 ------------------ 发件人: "zjy-ucas";notifications@github.com; 发送时间: 2016年11月13日(星期天) 下午3:17 收件人: "raghavchalapathy/Bidirectional-LSTM-CRF-for-Clinical-Concept-Extraction"Bidirectional-LSTM-CRF-for-Clinical-Concept-Extraction@noreply.github.com; 抄送: "夜"411489774@qq.com; "Mention"mention@noreply.github.com; 主题: Re:[raghavchalapathy/Bidirectional-LSTM-CRF-for-Clinical-Concept-Extraction] Bugin calculating cost? (#1)

@Zhangzirui 我参考了他的crf层，中文和英文的处理稍有不同，有一些bug还要修改，改好了会发，可以留个邮箱先给你

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

SophiaCY commented 7 years ago

您好~不知道您现在进展如何？ 我也想将其用在中文命名实体识别上，不过程序没跑通…… 我训练模型后只能生成mappings.pkl和parameters.pkl，想问一下char_layer、word_layer、final_layer、tanh_layer和transitions都是怎么生成的呀？ 另外我也想看看您处理中文的程序，非常感谢！