tensorflow::serving::ServerCore::Options Struct Reference

Options for configuring a ServerCore object. More...

#include <server_core.h>

Public Attributes
ModelServerConfig	model_server_config
absl::optional< string >	model_config_list_root_dir
std::unique_ptr< AspiredVersionPolicy >	aspired_version_policy
AspiredVersionsManager::CustomSortActionsFn	custom_sort_actions
int32	num_load_threads = 0
int32	num_initial_load_threads = 4.0 * port::NumSchedulableCPUs()
int32	num_unload_threads = 0
uint64_t	total_model_memory_limit_bytes
int32	max_num_load_retries = 5
int64_t	load_retry_interval_micros = 1LL * 60 * 1000 * 1000
int32	file_system_poll_wait_seconds = 30
bool	flush_filesystem_caches = false
PlatformConfigMap	platform_config_map
ServableStateMonitorCreator	servable_state_monitor_creator
CustomModelConfigLoader	custom_model_config_loader
bool	allow_version_labels = true
bool	fail_if_no_model_versions_found = false
bool	enable_reload_servables_with_error = false
bool	servable_versions_always_present = false
std::unique_ptr< ServerRequestLogger >	server_request_logger
ServerRequestLoggerUpdater	server_request_logger_updater
PreLoadHook	pre_load_hook
bool	allow_version_labels_for_unavailable_models = false
bool	force_allow_any_version_labels_for_unavailable_models = false
internal::PredictResponseTensorSerializationOption	predict_response_tensor_serialization_option
std::string	storage_path_prefix
bool	enable_cors_support = false
bool	with_current_context = false
absl::Duration	servable_state_waiter_timeout = absl::InfiniteDuration()
std::function< bool(absl::Status)>	should_retry_model_load

Detailed Description

Options for configuring a ServerCore object.

Definition at line 96 of file server_core.h.

Member Data Documentation

predict_response_tensor_serialization_option

internal::PredictResponseTensorSerializationOption tensorflow::serving::ServerCore::Options::predict_response_tensor_serialization_option

Initial value:

=
            internal::PredictResponseTensorSerializationOption::kAsProtoField

Definition at line 205 of file server_core.h.

total_model_memory_limit_bytes

uint64_t tensorflow::serving::ServerCore::Options::total_model_memory_limit_bytes

Initial value:

=
        std::numeric_limits<uint64_t>::max()

Definition at line 123 of file server_core.h.

---

The documentation for this struct was generated from the following file:

• tensorflow_serving/model_servers/server_core.h