Correct reproducibility issues as cited by MLSys 2020 reproducibility reviews

We are adding these changes to improve reproducibility of our artifact.

Changes:

• Added machine + kernel specifications to the README
• Decreased parallelism on small-core and small-RAM machines (now minimum core count per job is 12 versus 4)
• Added script by reviewer that reproduces all results

Artifact re-release for MLSys 2020 paper (DOI issuance)

This is an update of the MLSys 2020 paper artifact setup instructions, re-released for Zenodo DOI issuance.

Artifact re-release for MLSys 2020 paper (DOI issuance)

This release represents the artifact corresponding to the results in our MLSys 2020 paper. Instructions to reproduce experiments are contained in the README.

We are re-relasing the code to generate a DOI on Zenodo.

Artifact release for MLSys 2020 paper

This release represents the artifact corresponding to the results in our MLSys 2020 paper. Instructions to reproduce experiments are contained in the README.

Working Checkmate release using high-level Keras API

This release encompasses the first release of Checkmate that will load a graph from Keras and output a schedule that can be run by user-code.