Vanilla Skip-gram model for word embedding.
All codes are tested under Python 3.5 and Tensorflow 1.1
To run the program, do the following steps:
- Change the directory to the folder
- Run the command
python train.py
The file processed_data.py is adapted from the example codes of Stanford CS 20SI 2017. Original codes were compatible for Python 2.7 and Tensorflow 0.8. We have been modified them and tested under Python 3.5 and Tensorflow 1.1 .