a Pytorch implementation of neural machine translation system from english (chinese) to chinese (english) based on 30m parallel data.
this is a neural machine translation system from english to chinese and vice versa.
The training set contians 30m english parallel corpora from chinese to english.
As we use Google's sentencepiece encoding methods for both en and ch, the shared vocabulary enables us to learn a single model to translate either english or chinese.
The model is a 8 layer 512 transformer enhanced with LSTM.
I use a webservice to make the model output more instant visible.
Usage:
1)pip3 install -r requirements.txt
2)download the model from https://drive.google.com/file/d/1JMf3daDdMvi2GXH39EatcSmZws8cxFZW/view?usp=sharing
3)python3 interface.py
and then open your browser with http://localhost:4000/
I also build a public website: https://bingning.wang/translation/ (due to the mvps memory limits, the model is a 512-5 layer transformer).
some snapshots of our model.
此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。