integrations/torchserve/metadata.yaml (5 lines of code) (raw):
id: torchserve
short_name: TorchServe
display_name: TorchServe
description: |
TorchServe is a performant, flexible and easy-to-use tool for serving PyTorch models in production. The integration collects metrics related to throughput, latency, resource usage, and errors.