/sequence-models

Repository contains examples of sequence deep learning models used in small projects

Primary LanguageJupyter Notebook

image Image Source

Sequence Models

Repository contains examples of sequence deep learning models used in small projects

Introduction

When I first found out about sequence models, I was amazed with how easily they can be applied to a wide range of problems: text classification, text generation, music generation, machine translation and others. I got several ideas of how to have fun with sequence models, that's why I created this repository for storing Kaggle kernels and scripts with implementations.

1. Kernel Title Generation for Kaggle

I got an idea to use Meta Kaggle dataset to train a model for generation of new kernel titles. This could help to capture some trends for Kaggle kernels and give an inspiration. In this kernel:

  • I loaded and preprocessed Kaggle data on kernels.
  • Implemented and trained a sequence model with LSTM for generation of new Kaggle titles.

Link to the notebook.

References and Further Reading

  1. The Unreasonable Effectiveness of Recurrent Neural Networks
  2. Long Short-Term Memory: From Zero to Hero with PyTorch
  3. LSTMs for Time Series in PyTorch
  4. Sampling from an RNN