Signs • Key Features • Examples • Install • Support • Docs • Issues • License • Download
If you want the simplest possible way to generate embeddings for deep learning models without sacrificing the power of state-of-the-art vector models, Signs is for you.
Signs is a set of tools for text preparation, vectorization and processing and radically simplifies raw text to Keras embeddings workflow. Signs unifies Gensim and SpaCy vectorization backends for Keras users and provides an easy-to-use vectorization solution to manage otherwise complex workflows. Signs provides a meaningful replacement for dozens of lines of redundant code that are currently required to transform raw text into a a Keras Embeds layer.
- unifies Gensim and SpaCy vectorization backends
- supports using common vector models: GloVe, Fasttext, and word2vec
- removes NLP learning curve
- adds no more than a few lines of code to your worflow
- From text to Keras embedding layer in a single command
- Train, save, and load custom vectors
- Evaluate results after training a Keras prediction model
- Powerful text preprocessing features
- Allows completely automated text preprocessing
get source for the below example.
Several example notebooks with common workflows can be found here.
Stable version:
Daily development version:
If you want ask a "how can I use Signs to..." question, the right place is StackOverflow.
If you found a bug or want to suggest a feature, check the issues or create a new issue.