LATEST PROJECTS
PROJECTSCompute Cluster
CDS Compute Cluster is an in-house cluster developed for the Cornell Data Science project team for compute and project resources. The cluster consists of five computers, with one head node and four compute nodes. Currently, the cluster hosts two LLMs using vLLM and Llama.cpp for use by CDS members as well as a health dashboard and the CDS website, accessible on the internet. It also provides a local pipeline for MathSearch, a previous CDS project that originally utilized AWS.