llama.cpp/ggml-alloc.h
Georgi Gerganov db3abcc114
sync : ggml (ggml-backend) (#3548)
* sync : ggml (ggml-backend)

ggml-ci

* zig : add ggml-backend to the build
2023-10-08 20:19:14 +03:00

34 lines
1.3 KiB
C

#pragma once
#include "ggml.h"
#ifdef __cplusplus
extern "C" {
#endif
struct ggml_backend_buffer;
GGML_API struct ggml_allocr * ggml_allocr_new(void * data, size_t size, size_t alignment);
GGML_API struct ggml_allocr * ggml_allocr_new_measure(size_t alignment);
GGML_API struct ggml_allocr * ggml_allocr_new_from_buffer(struct ggml_backend_buffer * buffer);
// tell the allocator to parse nodes following the order described in the list
// you should call this if your graph are optimized to execute out-of-order
GGML_API void ggml_allocr_set_parse_seq(struct ggml_allocr * alloc, const int * list, int n);
GGML_API void ggml_allocr_free (struct ggml_allocr * alloc);
GGML_API bool ggml_allocr_is_measure (struct ggml_allocr * alloc);
GGML_API void ggml_allocr_reset (struct ggml_allocr * alloc);
GGML_API void ggml_allocr_alloc (struct ggml_allocr * alloc, struct ggml_tensor * tensor);
GGML_API size_t ggml_allocr_alloc_graph(struct ggml_allocr * alloc, struct ggml_cgraph * graph);
GGML_API size_t ggml_allocr_max_size (struct ggml_allocr * alloc);
GGML_API size_t ggml_allocr_alloc_graph_n(
struct ggml_allocr * alloc,
struct ggml_cgraph ** graphs, int n_graphs,
struct ggml_tensor *** inputs, struct ggml_tensor *** outputs);
#ifdef __cplusplus
}
#endif