Skip to content

solargrammars/name_color

Repository files navigation

name-color

Transform a name or description into a color representation

This is the code used in this article where the paper Character Sequence Models for Colorful Words is analyzed.

This code was tested on Python 3.6 / Pytorch 1.0

How to use it

  1. Modify settings.py according to your system.

  2. run_download_data.py will download tuples (color, description) from ColourLovers using their API.

  3. run_process_data.py will clean and split the data.

  4. run_train.py will train and save the best model.

  5. run_inference.py and run_incremental_inference.py are used to obtain the reuslts on the test set.

Acknowledgement

Most of the content on this repository is based on the great Pytorch tutorials by epochx

About

Pytorch implementation of the paper Character Sequence Models for Colorful Words by Kawakami et al. EMNLP '16

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published