metal : reduce command encoding overhead (llama/9698)

This commit is contained in:
Georgi Gerganov 2024-10-02 15:12:16 +03:00
parent ff2cb0811f
commit 162a455402
2 changed files with 1999 additions and 1893 deletions

View File

@ -25,9 +25,6 @@
#include <stddef.h>
#include <stdbool.h>
// max memory buffers that can be mapped to the device
#define GGML_METAL_MAX_BUFFERS 64
struct ggml_tensor;
struct ggml_cgraph;
@ -48,8 +45,6 @@ GGML_API bool ggml_backend_is_metal(ggml_backend_t backend);
GGML_API GGML_CALL ggml_backend_buffer_t ggml_backend_metal_buffer_from_ptr(void * data, size_t size, size_t max_size);
GGML_API void ggml_backend_metal_set_n_cb(ggml_backend_t backend, int n_cb);
GGML_API void ggml_backend_metal_set_abort_callback(ggml_backend_t backend, ggml_abort_callback abort_callback, void * user_data);
GGML_API GGML_CALL ggml_backend_buffer_type_t ggml_backend_metal_buffer_type(void);

File diff suppressed because it is too large Load Diff