DeepMoji and torchMoji
The torchMoji implementation is a PyTorch port of the original DeepMoji model, making them ecosystem siblings where torchMoji provides an alternative framework-specific implementation of the same underlying architecture.
About DeepMoji
bfelbo/DeepMoji
State-of-the-art deep learning model for analyzing sentiment, emotion, sarcasm etc.
Built on Keras with Theano/TensorFlow backends, the model leverages 1.2 billion tweets to learn emoji-emotion associations, then applies transfer learning for fine-tuning on custom datasets. Extracts 2304-dimensional emotional feature vectors and predicts emoji distributions for new text, enabling both direct emoji classification and downstream task adaptation. A PyTorch alternative (torchMoji) is available via HuggingFace for modern workflows.
About torchMoji
huggingface/torchMoji
😇A pyTorch implementation of the DeepMoji model: state-of-the-art deep learning model for analyzing sentiment, emotion, sarcasm etc
Built on 1.2 billion tweets with emoji supervision, the model learns emotional representations transferable to downstream tasks via fine-tuning. The architecture extracts 2304-dimensional emotional feature vectors from text and predicts emoji distributions, enabling both direct emoji classification and transfer learning for sentiment/sarcasm detection on custom datasets. Integrates with scikit-learn and the emoji library for preprocessing and analysis workflows.
Scores updated daily from GitHub, PyPI, and npm data. How scores work