Fast, Scalable, and Standardized AI Inference Deployment for Multiple Frameworks, Diverse Models on CPUs and GPUs with Open-source NVIDIA Triton_1647043886829001tG1D