A Beautiful Visualization Dashboard For Machine Learning
Project description
For detailed codumentation, seeml-dash-tutorial
ML-dash replaces visdom and tensorboard. It allows you to see real-time updates, review 1000+ of experiments quickly, and dive in-depth into individual experiments with minimum mental effort.
Parallel Coordinates
Aggregating Over Multiple Runs (with different seeds)
Preview Videos, ``matplotlib`` figures, and images.
Usage
To make sure you install the newest version of ml_dash:
conda install pycurl
pip install ml-logger ml-dash --upgrade --no-cache
Just doing this would not work. The landscape of python modules is a lot messier than that of javascript. The most up-to-date graphene requires the following versioned dependencies:
yes | pip install graphene==2.1.3
yes | pip install graphql-core==2.1
yes | pip install graphql-relay==0.4.5
yes | pip install graphql-server-core==1.1.1
There are two servers:
a server that serves the static web-application files ml_dash.app
This is just a static server that serves the web application client.
To run this:
python -m ml_dash.app
the visualization backend ml_dash.server
This server usually lives on your logging server. It offers a graphQL API backend for the dashboard client.
python -m ml_dash.server --logdir=my/folder
Note: the server accepts requests from ``localhost`` only by default for safety reasons. To overwrite this, see the documentation here: ml-dash-tutorial
Implementation Notes
See https://github.com/episodeyang/ml_logger/tree/master/ml-dash-server/notes/README.md
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.