Run inference on deployed models
/predict
– Standard synchronous inference./async_predict
– Asynchronous inference for long-running tasks.predict
endpoints as well as a special sync
endpoint. By using the sync
endpoint you are able to call different routes in your custom server.
https://model-{model_id}.../sync/health
-> /health
https://model-{model_id}.../sync/items
-> /items
https://model-{model_id}.../sync/items/123
-> /items/123
truss predict