https://github.com/captaindra/translator_ch-en
CH-EN transformer
https://github.com/captaindra/translator_ch-en
attention-model pyhton translator-app ui
Last synced: 4 months ago
JSON representation
CH-EN transformer
- Host: GitHub
- URL: https://github.com/captaindra/translator_ch-en
- Owner: CaptainDra
- Created: 2020-07-04T02:45:14.000Z (over 5 years ago)
- Default Branch: master
- Last Pushed: 2020-11-14T03:25:24.000Z (almost 5 years ago)
- Last Synced: 2025-01-03T04:16:12.234Z (10 months ago)
- Topics: attention-model, pyhton, translator-app, ui
- Language: Python
- Homepage:
- Size: 2.13 MB
- Stars: 0
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Translator_CH-EN
This is a Chinese-English Translator demo, refferred to the thesis:[Attention Is All You Need](https://arxiv.org/abs/1706.03762).
I tried to build rebuild a model as the thesis mentioned, and make a simple user interface for the application, the UI like this:
The system run as following system flow chart:
For chinese Translator, we need to add word_segmentation function to unified sentence structure, as following picture:
Then, we need preprocess as other kind of Translator(Count the number of occurrences of each word):
After that, we can train the model:
The total result:
We could get some translation as following(similar meaning in different result, but ignore draft):
Good result for some sentences that have appeared or stereotyped expression:
However, the application had some problems with the punctuation, time and number:
This problem can be solved if you add a function to find all punctuations, times and numbers in preprocess function. If you want to have better result, please add that yourself.