OR-NMT

OR-NMT copied to clipboard

RuntimeError: masked_select(): self and result must have the same scalar type

3

This error occurs after saving the model. And my parameters are configured as follows: ``` python train.py ${data_dir} \$ --clip-norm 0.1 \$ --dropout 0.1 \$ --max-tokens ${max_tokens} \$ --seed ${seed}...

nomadlx

Can't reproduce the results on WMT En-De

2

Hi, all. I am trying to reproduce the **word-level oracle** results of this paper on WMT EN-DE dataset. I train the transformer model 150k steps with #Gpus = 4, #Freq...

Paulmzr

关于Oracle Word Selection

1

我按照论文思路，在lstm上使用了Oracle Word Selection+schedule sample，但是发现效果提升很微弱，我估计是我哪里设置的不对吧。 有一个问题，关于Oracle Word Selection 在论文公式11中，我直接选择了argmax oj-1作为最终的oracle word，没有经过softmax，我这样做的原因是argmax oj-1其实就是argmax Pj-1，也就是不需要经过softmax就可以得到oracle word，**为什么这里要加公式12，也就是softmax呢？**

goodluck110706112

Argparse version

1

Thank you for kindly sharing the source training code. However, in the RNNSearch model, the arguments in ``wargs.py`` are not written in ``argparse`` formats. Would you mind refactoring this part...

haorannlp

Sharing code for MIXER implementation from this paper

1

Hello All, will it be possible for you to share your code for MIXER and SS baseline implementation for the paper.

kushalarora