multipath / metanmt Goto Github PK

I would like to run meta_nmt5.py as I wish to do some work on meta-learning for low resource languages.
I am unable to run the code as I could not find any standards showing what a .tok file should contain and how it is formatted.
I would extremely appreciate it if you could write up a short guide on how to prepare a dataset for use with this code or if you could share your pre-processed tokenized data (the accompanying word embeddings) that you used with this code.

您好，我讀了您在EMNLP上利用元學習方法對NMT做改進，我目前是一名研一的學生，想諮詢一下您這份代碼就是論文中所使用的方法嗎，希望得到您的回復，謝謝

Hi,
I'm investigating the topic of NMT in the low-resource language and would like to reproduce your work. I failed in using multilingual word vectors trained by MUSE which fits bad in the network this code defined. I'm not sure about the exact training parameters of these word vectors and their format in files. As they are the initialization of ULR, could you please give some hints about the steps in detailed?
Thanks

How to run this code correctly，I'm very interested in your research. I hope to reproduce your experiment，Can you give us a brief introduction of the experiment process？Thank you。