A sample PyTorch/NLP project on the dataset GoEmotions.
-
This is a sample project to demonstrate basic coding skills in PyTorch and NLP.
-
The model is a simple LSTM-based multi-label emotion classifier.
-
The focus is on clean structure and readable code, rather than model performance.
To train, evaluate, and test the model with default settings:
python main.pyTo customize model hyperparameters:
python main.py --embedding_size 128 --hidden_size 256 --batch_size 32 --epochs 5 --shuffle