[cpu] Implement all gather/reduce for arm64 cpu (#12527)
This commit is contained in:
@@ -4,7 +4,6 @@
|
||||
|
||||
#ifndef __SHM_COLLECTIVES__
|
||||
#define __SHM_COLLECTIVES__
|
||||
#define VECTOR_LENGTH_IN_BYTES 32
|
||||
void shm_initialize(int size, int rank, const char* addr_string, const char* port_string);
|
||||
void all_reduce_outer_loop(torch::Tensor& data, size_t numel, int data_size);
|
||||
torch::Tensor& all_gather(torch::Tensor& result, torch::Tensor& data, int dim, size_t numel, int data_size);
|
||||
|
||||
Reference in New Issue
Block a user