Command line toolbox for managing spoken language identification (LID) experiments.

Install TensorFlow 2.

Clone the repo and install lidbox as a Python package in setuptools develop mode. This makes it easier to experiment with the code.

git clone --depth 1 https://github.com/matiaslindgren/lidbox.git
pip install --editable ./lidbox

See this for a simple example to get started.

• more mutex locks when updating the caches
• simplify tf.data.Dataset pipelines, no spaghetti leaks
• efficient metric implementations, must run in tf graph