hankcs
hankcs
我当时用的是CPU是Intel(R) Xeon(R) CPU E3-1220 v5 @ 3.00GHz，内存好像是8G，跑了一两个星期。你可从控制台输出的每个epoch的时间乘以总epoch的时间去估算。
请参考：https://github.com/hankcs/multi-criteria-cws/blob/master/model.py#L30
不清楚，看上去像是内存不够，试试更大内存的机器。我的实验环境是8个G。
Dynet版本号不匹配，必须是2.0.1：https://github.com/clab/dynet/releases/tag/2.0.1
Dynet版本号不匹配，必须是2.0.1：https://github.com/clab/dynet/releases/tag/2.0.1 而你安装的是dyNET (2.0.2)
感谢反馈，抱歉我提供了错误的版本号，正确的版本号应该是https://github.com/clab/dynet/releases/tag/2.0.1 ，已经反复验证过了。 当时由于从源码编译安装的Dynet版本号只显示dyNET (0.0.0)，而论文试验是8月份开始的，所以按照Dynet的发布日志猜测是v2.0。安装后果然可以启动，但每个epoch会出现找不到learning_rate的问题。现在从git commit hash（87df34103625102493f8c660684146a636e2482c）看，应该属于2.0到2.0.1之间的一个版本。通过反复验证，发现2.0.1可以正常运行。 麻烦按照：https://github.com/hankcs/multi-criteria-cws/issues/1#issuecomment-351279371 重新安装2.0.1，谢谢。
Hi orange text team. May I recommend a third option? I'm maintaining an open source multilingual NLP package called [HanLP](https://github.com/hankcs/HanLP/tree/master) backed by the start-of-the-art deep learning techniques and also efficient...
Sure, glad to help. Let's decide the version first since new package means new depdendencies. What kind of dependencies would you like to introduce? - HanLP 2.x is using TensorFlow...
Great, HanLPerceptron is a good choice. Let's see what needs to be done.
Sounds good. I'll work together with the author of `HanLPerceptron ` to have the wheels built and tested first.