|
---
|
|
license: mit
|
|
|
|
language:
|
|
- de
|
|
- en
|
|
|
|
tags:
|
|
- translation
|
|
- pytorch
|
|
|
|
datasets:
|
|
- multi30k
|
|
|
|
metrics:
|
|
- bleu
|
|
|
|
model-index:
|
|
- name: multi30k
|
|
results:
|
|
- task:
|
|
type: translation
|
|
dataset:
|
|
type: multi30k
|
|
name: multi30k-de-en
|
|
metrics:
|
|
- type: bleu
|
|
value: 33.468
|
|
name: Test BLEU
|
|
args: n_gram=4
|
|
---
|
|
|
|
# Seq2seq + Attention
|
|
|
|
Pytorch implementation of [Neural Machine Translation by Jointly Learning to Align and Translate](https://arxiv.org/abs/1409.0473). Trained on the [Multi30k-de-en](http://www.statmt.org/wmt16/multimodal-task.html#task1) dataset with sentencepiece as the tokenizer.
|
|
|
|
Here's the attention heatmap of a random sample from the test set:
|
|
|
|
![attention-heatmap](attention-heatmap.png)
|
|
|