CodaLab

A collaborative platform for reproducible research.

Running 100 experiments in parallel on different versions of your code/data? Don't remember how you got that result from 6 months ago? CodaLab allows you to run your jobs on a cluster, document and share your experiments, all while keeping track of full provenance, so you can be a more efficient researcher.

To get started, read more about the mission, sign up for an account, and go through the tutorial to try it out!

Upload

Upload code (in any programming language) and data (in any format) to CodaLab. If you're using the command-line interface (CLI), type, for example:
$ cl upload train.py
This produces an immutable bundle:
train.py

Run

Run any bash shell command, which can depend on previous bundles. In the CLI or the web terminal, type, for example:
$ cl run :train.py 'python train.py --eta 0.2'
CodaLab records the execution environment using docker to ensure reproducibility, and produces a new bundle:
train.py
run17
This bundle is added to the global dependency graph over all research assets (see right).

Present

Create worksheets to manage your runs, document your experimental results, and publish executable papers. Use CodaLab markdown to create custom tables and graphs.