Run Serverless GPUs on Private Cloud

Run serverless GPUs on private cloud: deploy custom or open-source ML models with HTTP-driven autoscaling, scale to zero, and 20-second warm-node cold starts.

Overview

Tensorfuse helps you run serverless GPUs on private cloud. You can deploy any custom or open-source ML model and will auto-scale wrt to the number of http request your app receives. Automatically scale down to 0 and upto required number of GPU nodes.