finalfrontier is a Rust library and set of utilities for learning and using word embeddings. finalfrontier currently has the following features:
- Models:
- skip-gram (Mikolov et al., 2013)
- structured skip-gram (Ling et al., 2015)
- Noise contrastive estimation (Gutmann and Hyvärinen, 2012)
- Subword representations (Bojanowski et al., 2016)
- Hogwild SGD (Recht et al., 2011)
This is an early release of finalfrontier, we are planning to add more features
in the future. The trained embeddings are stored in finalfusion format, which
can be read and used with the rust2vec
crate and the finalfusion Python
module.
- Installation
- Quickstart
- Manual pages:
- ff-train(1) — train word embeddings with subword representations
- rust2vec rust2vec crate
- Python module