environment : Google colab
Maybe every package is most recent version
dataset : nsmc(naver sentiment movie corpus)
dependency
- KoNLPy Mecab (Tokenizer)
- torch
- torchtext
Done
- LSTM
- Bi-LSTM with learning rate schduler
- add pretrained embedding layer in LSTM
- LSTM with attention layer
- CNN
- CNN with attention layer
Todo
- CNN Embedding + LSTM
- BERT
- ...