Simple, minimalist, spartan, lightweight, single-class deep learning experiment tracker. Also, library agnostic.
-
no installation. Just put the .py file in your project.
-
no docs, tutorial, or supports (I got a sample code though...)
-
does not automate runs
-
does not support distributed computing and what not
-
does not even inlcude the codes to analyze between runs
A knock-off of Weights & Biases because I cannot get it work.
Check the .ipynb for sample usage.