We have added NFS (Network File Storage) as a new type of data source.
Configure endpoints for model deployments using the gRPC framework on HTTP/2, enabling high-performance inference services with lower latency and increased concurrent processing compared to REST API endpoints on HTTP/1.1.
Monitor REST API and gRPC model deployments with real-time insights into latency, request counts, failure rates, and unique user metrics to optimize and ensure reliable inference services.
For a detailed overview, watch our feature highlights video.