Code Monkey home page Code Monkey logo

gl-gin's Introduction

Hi there 👋 I'm Libo Qin.

  • 🌱 I’m a PhD. Candidate in HIT-SCIR.
  • 📫 How to reach me: Email or Homepage.
  • ⚡ Fun fact: Love music, sining, animation, sports and all good things in my life.

gl-gin's People

Contributors

awake020 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar

gl-gin's Issues

关于论文中metrics的问题

作者您好,
我这边按照readme做好了需求环境,
尝试了用您在repo中留的titan xp对应训练参数, 在tesla v100 上进行训练
并没有复现出论文的结果, 在 mixSNIPS clean 和 mixATIS clean上均无法接近论文的效果, 具体训练log我贴在附件内
mix_Atis.txt
mix_snips.txt

请问是有其他细节么,我想复现论文内sota的结果,最起码能接近

More information about t_test

Why we can't re-produce the score in the paper by using the recommended hyper-parameters? Besides, i am curious about the t-test, as far as i know, t-test need more than one-time exp. Although we acheived the exactly same performance by using the previded "best model", We can't pass t-test by using the recommended hyper-parameters. Could you please provide more best-model by differents hyper-parameters or provide more details about training.

想确认一下数据

作者您好,
请问repository里的data mixATIS_clean 和mixSNIPS_clean 和其分割文件就是paper里result对应使用的数据是么?

关于中文意图识别与槽位填充的问题,slot值在原文未出现该如何标注

您好,我看了你们的文章,并且也运行了代码,基本上复现了论文中的结果,非常感谢你们能够开源这项工作!
我目前希望利用这套代码处理中文的意图识别和槽位填充的问题,但还是有些地方不太明白想咨询一下:

我目前为了方便直接模拟英文的数据集,将中文按照字级别拆开,即['订', '机', '票']这样的结构,效果不是很理想。

所以想问一下,按照这套代码逻辑更适合下面哪一种分词呢:

  • 字级别分词
  • jieba分词
  • 套用Bert的词向量

还有一个问题困扰了我很久,就是如果slot在原文中没有对应的word,该如何去做标注呢?
比如:

  • text:自驾游去北京都经过那些地方啊?
  • slots:{'目的地': '北京', '询问类型': '导航'}
  • intent:旅游咨询

这里槽位值“导航”就没法在text里标注。。。

LSTM object has no attribute _flat_weights

环境:
CUDA Version: 11.6
NVCC Version: 11.6
torch Version: 1.7.0

问题:
你好,我google了一下该Error,可以降低torch版本到1.3,不过,这时报错:torch版本与现在的CUDA版本又不一致。请问,修改现有代码哪里可以适配现在的torch和CUDA版本? 谢谢~

code run error

hi,
I have encountered this problem, which has been unable to solve. Can you provide some ideas

File "D:\PythonCode\MachineLearning\slu\UGEN-main\utils\data_qa.py", line 164, in data_process
print(t)
UnicodeEncodeError: 'gbk' codec can't encode character '\xe2' in position 31: illegal multibyte sequence

Optimal hyper-parameters

Hi authors,

I couldn't find the optimal parameters to reproduce your results reported in the paper. In detail, I validate your provided model, the results are 84.3%, 72.8% and 38.4% for Slot F1, Intent accuracy and Overall accuracy, respectively. It is a large margin compared to your reported results in the paper (88.3%, 76.3% and 43.5%).
Could you provide more details and/or your pretrained model with the best result?

Thank you

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.