DSKSD/DeepNLP-models-Pytorch

Covers word embeddings (Skip-gram, GloVe), sequence labeling (NER), structured prediction (dependency parsing), and sequence-to-sequence models with attention for machine translation. Implementations span from foundational embedding techniques to advanced architectures like recursive neural networks for sentiment analysis and dynamic memory networks for question answering. Each model includes Jupyter notebooks with accompanying research papers and datasets, designed for practitioners working through Stanford's CS224N curriculum.

2,949 stars. No commits in the last 6 months.