deepgram/kur
A deep learning framework that enables designing, training, and evaluating neural network models through declarative YAML configuration files without writing code.
Not currently ranked — collecting fresh signals.
star history
Kur is a descriptive deep learning framework that allows users to define neural network models using simple YAML configuration files rather than writing code. It abstracts away the complexity of backends by supporting Theano, TensorFlow, and PyTorch under a unified interface. The tool provides out-of-the-box multi-GPU training capabilities and includes templating via Jinja2 for rapid experimentation with model variants.
Frequently asked
- What is deepgram/kur?
- A deep learning framework that enables designing, training, and evaluating neural network models through declarative YAML configuration files without writing code.
- Is kur open source?
- Yes — deepgram/kur is open source, released under the Apache-2.0 license.
- What language is kur written in?
- deepgram/kur is primarily written in Python.
- How popular is kur?
- deepgram/kur has 824 stars on GitHub.
- Where can I find kur?
- deepgram/kur is on GitHub at https://github.com/deepgram/kur.