Using a sequence to sequence RNN model to translate sentences from English to French. The dataset is text files of ~140,000 fairly simple sentences in English and in French.
This is Project 4 of Udacity's Deep Learning Nanodegree Foundation. The aim of the project is to go through the process of implementing a sequence to sequence model in TensorFlow by building the encoder and decoder architectures, and finally be able to translate simple English sentences to French.
Hyperparameter | Number |
---|---|
Epochs | 30 |
Batch size | 256 |
RNN size | 100 |
LSTM layers | 2 |
Encoding embedding size | 100 |
Decoding embedding size | 100 |
Learning rate | 0.001 |
Keep probability | 0.9 |