Class ITritonClient

(Latest Version)
class ITritonClient

Subclassed by morpheus::HttpTritonClient

Public Functions

virtual ~ITritonClient() = default

virtual triton::client::Error is_server_live(bool *live) = 0

Checks if Triton Server is live.

virtual triton::client::Error is_server_ready(bool *ready) = 0

Checks if Triton Server is ready.

virtual triton::client::Error is_model_ready(bool *ready, std::string &model_name) = 0

Checks if the given model is ready.

virtual triton::client::Error model_metadata(std::string *model_metadata, std::string &model_name) = 0

Gets metadata for the given model.

virtual triton::client::Error model_config(std::string *model_config, std::string &model_name) = 0

Gets the config for the given model.

virtual triton::client::Error async_infer(triton::client::InferenceServerHttpClient::OnCompleteFn callback, const triton::client::InferOptions &options, const std::vector<TritonInferInput> &inputs, const std::vector<TritonInferRequestedOutput> &outputs) = 0

Runs Triton Server inference given the model options, inputs, and outputs.

Previous Class InferenceMemoryNLP
Next Class KafkaOAuthCallback
© Copyright 2024, NVIDIA. Last updated on Apr 25, 2024.