This repository is a fork of the nvidia-gpu-device-plugin in Google Kubernetes Engine. This fork allows overcommitment of GPU resources, meant so multiple pods can share 1 GPU by reporting more actual available GPUs to Kubernetes.
This can be set by setting -overcommit-multiply-by=2 or a higher number, this specifies how many times 1 GPU will be reported.
More details on the nvidia-gpu-device-plugin are here.