ml-lab / reco-rl

Codes of AAAI 2020 paper "What Makes A Good Story? Designing Composite Rewards for Visual Storytelling"

What Makes A Good Story? Designing Composite Rewards for Visual Storytelling

Implemented by Junjie Hu

Contact: [email protected]

If you use the codes in this repo, please cite our AAAI2020 paper.

@inproceedings{hu20aaai,
    title = {What Makes A Good Story? Designing Composite Rewards for Visual Storytelling},
    author = {Junjie Hu and Yu Cheng and Zhe Gan and Jingjing Liu and Jianfeng Gao and Graham Neubig},
    booktitle = {Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI)},
    address = {New York, USA},
    month = {February},
    url = {https://arxiv.org/abs/1909.05316},
    year = {2020}
}

Installation

Please use the following Anaconda environment.

• python=3.6
• pytorch=1.0.1
• pytorch_pretrained_bert=0.6.2
• spacy, nltk, numpy, scipy, h5py, json, pickle

conda env create --file conda-env.txt

Downloads

The preprocessed data and pre-trained models can be found here. Extract data.zip under the ReCo-RL/data directory. Extract reco-rl-model.zip under the ReCo-RL/outputs/rl/ directory. Extract bert-base-uncased.zip under ReCo-RL/bert-weight/ directory.

• data.zip: train/dev/test data including image features, VIST captions and entities preprocessed by spacy.
• reco-rl-model.zip: model file (model.bin) and vocab file (vocab.bin).
• bert-base-uncased.zip: BERT's next sentence predictor model and its vocab file.

Demo

• Decode the test set using pre-trained model. We recommend to use our pre-trained model for further comparison in your paper. After decoding, you would expect to see the decoding automatic scores as follows (improved scores over those reported in the paper).

Score	SPICE	BLEU-4	ROUGE-L	CIDEr	METEOR
ReCo-RL	11.2	13.5	29.5	11.8	35.8

bash scripts/test.sh [GPU id]

• Train a MLE model from scratch

bash scripts/train_mle.sh [GPU id]

• Train the model initialized by the MLE-trained model

bash scripts/train_rl.sh [GPU id]