- Dataset: Indosum dataset: https://github.com/kata-ai/indosum
- Pre-Trained Word Embedding: Pre-trained word2vec: https://github.com/deryrahman/word2vec-bahasa-indonesia
- Custom Attention Layer: Attention Layer: https://github.com/thushv89/attention_keras
We create our model using seq2seq with addition of attention layer and pre-trained word embedding