skip-thoughts

An implementation of Skip-Thought Vectors in PyTorch.

Blog

Here's a blog explaining the subtleties of Skip-Thoughts  

Instructions

Training

• Download BookCorpus or any other data-set and concatenate all sentences into one file and put it in ./data/ directory

• Modify the following line in Train.ipynb notebook accordingly: d = DataLoader("./data/dummy_corpus.txt")

• There is no early stopping.

• The Train notebook runs at the rate of 1 epoch / 2 days on an Nvidia 1080 Ti.

• Your model is saved when ./saved_models when the average training loss in the last 20 iterations dips below the previous best.

Evaluation

Only implemented on classification tasks

• Download the movie review dataset and put rt-polarity.neg and rt-polarity.pos in the ./tasks/mr_data directory.

• You may also test on other classification tasks by downloading the datasets and providing their path and tasks type in Evaluate.ipynb