sequence-modeling-with-BRNNs-LSTMs-1D-convolutions-etc

this repository contains code for modeling temporal sequences using LSTMs, Bidirectional RNNs, pretrained Embeddings (GloVe) and 1D CNNs, codes use the standard IMDB dataset

GloVe

the code uses the glove6B 50d embeddings that can be downloaded from https://nlp.stanford.edu/projects/glove/

Analysis

1D convolutions stacked with LSTMs proved to outperform other methods by a factor of 3 to 5% on modeling temporal sequences,still overfitting yet achieved a 98 % train accuracy, 87 % validation and 86 % test accuracy