Struct LoadBalancer

• Defined in File deploy.h

Struct Documentation

struct LoadBalancer

The default LoadBalancer for torch::deploy which handles allocating and freeing subinterpreters.

Public Functions

inline explicit LoadBalancer(size_t n)

Creates a Loadbalancer which handles n interpreters.

inline void setResourceLimit(size_t n)

Changes the amount of subinterpreters which is handled by the load balancer.

int acquire()

Allocates an subinterpreter, and return its ID which is used to free it.

void free(int where)

Frees the subinterpreter with ID where.

This ID is returned by LoadBalancer::acquire()