mirror of
https://github.com/ggerganov/llama.cpp.git
synced 2024-12-25 22:08:46 +01:00
ggml : export symbols (#1155)
This commit is contained in:
Georgi Gerganov
GitHub
parent
0c5692345d
commit
8a0f8673ba
253
ggml.h
253
ggml.h
@ -169,14 +169,27 @@
|
||||
//
|
||||
//
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#ifdef GGML_SHARED
|
||||
# if defined(_WIN32) && !defined(__MINGW32__)
|
||||
# ifdef GGML_BUILD
|
||||
# define GGML_API __declspec(dllexport)
|
||||
# else
|
||||
# define GGML_API __declspec(dllimport)
|
||||
# endif
|
||||
# else
|
||||
# define GGML_API __attribute__ ((visibility ("default")))
|
||||
# endif
|
||||
#else
|
||||
# define GGML_API
|
||||
#endif
|
||||
|
||||
#include <stdint.h>
|
||||
#include <stddef.h>
|
||||
#include <stdbool.h>
|
||||
|
||||
#define GGML_FILE_MAGIC 0x67676d6c // "ggml"
|
||||
#define GGML_FILE_VERSION 1
|
||||
|
||||
#define GGML_MAX_DIMS 4
|
||||
#define GGML_MAX_NODES 4096
|
||||
#define GGML_MAX_PARAMS 16
|
||||
@ -184,6 +197,10 @@ extern "C" {
|
||||
#define GGML_MAX_OPT 4
|
||||
#define GGML_DEFAULT_N_THREADS 4
|
||||
|
||||
#ifdef __cplusplus
|
||||
extern "C" {
|
||||
#endif
|
||||
|
||||
#ifdef __ARM_NEON
|
||||
// we use the built-in 16-bit float type
|
||||
typedef __fp16 ggml_fp16_t;
|
||||
@ -192,14 +209,13 @@ typedef uint16_t ggml_fp16_t;
|
||||
#endif
|
||||
|
||||
// convert FP16 <-> FP32
|
||||
float ggml_fp16_to_fp32(ggml_fp16_t x);
|
||||
ggml_fp16_t ggml_fp32_to_fp16(float x);
|
||||
GGML_API float ggml_fp16_to_fp32(ggml_fp16_t x);
|
||||
GGML_API ggml_fp16_t ggml_fp32_to_fp16(float x);
|
||||
|
||||
struct ggml_object;
|
||||
struct ggml_context;
|
||||
|
||||
enum ggml_type {
|
||||
// explicitly numbered values are used in llama.cpp files
|
||||
GGML_TYPE_F32 = 0,
|
||||
GGML_TYPE_F16 = 1,
|
||||
GGML_TYPE_Q4_0 = 2,
|
||||
@ -341,60 +357,64 @@ struct ggml_init_params {
|
||||
bool no_alloc; // don't allocate memory for the tensor data
|
||||
};
|
||||
|
||||
void ggml_time_init(void); // call this once at the beginning of the program
|
||||
int64_t ggml_time_ms(void);
|
||||
int64_t ggml_time_us(void);
|
||||
int64_t ggml_cycles(void);
|
||||
int64_t ggml_cycles_per_ms(void);
|
||||
// misc
|
||||
|
||||
void ggml_print_object (const struct ggml_object * obj);
|
||||
void ggml_print_objects(const struct ggml_context * ctx);
|
||||
GGML_API void ggml_time_init(void); // call this once at the beginning of the program
|
||||
GGML_API int64_t ggml_time_ms(void);
|
||||
GGML_API int64_t ggml_time_us(void);
|
||||
GGML_API int64_t ggml_cycles(void);
|
||||
GGML_API int64_t ggml_cycles_per_ms(void);
|
||||
|
||||
int64_t ggml_nelements(const struct ggml_tensor * tensor);
|
||||
size_t ggml_nbytes (const struct ggml_tensor * tensor);
|
||||
GGML_API void ggml_print_object (const struct ggml_object * obj);
|
||||
GGML_API void ggml_print_objects(const struct ggml_context * ctx);
|
||||
|
||||
int ggml_blck_size (enum ggml_type type);
|
||||
size_t ggml_type_size (enum ggml_type type); // size in bytes for all elements in a block
|
||||
float ggml_type_sizef(enum ggml_type type); // ggml_type_size()/ggml_blck_size() as float
|
||||
GGML_API int64_t ggml_nelements(const struct ggml_tensor * tensor);
|
||||
GGML_API size_t ggml_nbytes (const struct ggml_tensor * tensor);
|
||||
|
||||
const char * ggml_type_name(enum ggml_type type);
|
||||
GGML_API int ggml_blck_size (enum ggml_type type);
|
||||
GGML_API size_t ggml_type_size (enum ggml_type type); // size in bytes for all elements in a block
|
||||
GGML_API float ggml_type_sizef(enum ggml_type type); // ggml_type_size()/ggml_blck_size() as float
|
||||
|
||||
size_t ggml_element_size(const struct ggml_tensor * tensor);
|
||||
GGML_API const char * ggml_type_name(enum ggml_type type);
|
||||
|
||||
bool ggml_is_quantized(enum ggml_type type);
|
||||
GGML_API size_t ggml_element_size(const struct ggml_tensor * tensor);
|
||||
|
||||
struct ggml_context * ggml_init(struct ggml_init_params params);
|
||||
void ggml_free(struct ggml_context * ctx);
|
||||
GGML_API bool ggml_is_quantized(enum ggml_type type);
|
||||
|
||||
size_t ggml_used_mem(const struct ggml_context * ctx);
|
||||
// main
|
||||
|
||||
size_t ggml_set_scratch(struct ggml_context * ctx, struct ggml_scratch scratch);
|
||||
GGML_API struct ggml_context * ggml_init(struct ggml_init_params params);
|
||||
GGML_API void ggml_free(struct ggml_context * ctx);
|
||||
|
||||
struct ggml_tensor * ggml_new_tensor(
|
||||
GGML_API size_t ggml_used_mem(const struct ggml_context * ctx);
|
||||
|
||||
GGML_API size_t ggml_set_scratch(struct ggml_context * ctx, struct ggml_scratch scratch);
|
||||
|
||||
GGML_API struct ggml_tensor * ggml_new_tensor(
|
||||
struct ggml_context * ctx,
|
||||
enum ggml_type type,
|
||||
int n_dims,
|
||||
const int64_t *ne);
|
||||
|
||||
struct ggml_tensor * ggml_new_tensor_1d(
|
||||
GGML_API struct ggml_tensor * ggml_new_tensor_1d(
|
||||
struct ggml_context * ctx,
|
||||
enum ggml_type type,
|
||||
int64_t ne0);
|
||||
|
||||
struct ggml_tensor * ggml_new_tensor_2d(
|
||||
GGML_API struct ggml_tensor * ggml_new_tensor_2d(
|
||||
struct ggml_context * ctx,
|
||||
enum ggml_type type,
|
||||
int64_t ne0,
|
||||
int64_t ne1);
|
||||
|
||||
struct ggml_tensor * ggml_new_tensor_3d(
|
||||
GGML_API struct ggml_tensor * ggml_new_tensor_3d(
|
||||
struct ggml_context * ctx,
|
||||
enum ggml_type type,
|
||||
int64_t ne0,
|
||||
int64_t ne1,
|
||||
int64_t ne2);
|
||||
|
||||
struct ggml_tensor * ggml_new_tensor_4d(
|
||||
GGML_API struct ggml_tensor * ggml_new_tensor_4d(
|
||||
struct ggml_context * ctx,
|
||||
enum ggml_type type,
|
||||
int64_t ne0,
|
||||
@ -402,128 +422,127 @@ struct ggml_tensor * ggml_new_tensor_4d(
|
||||
int64_t ne2,
|
||||
int64_t ne3);
|
||||
|
||||
struct ggml_tensor * ggml_new_i32(struct ggml_context * ctx, int32_t value);
|
||||
struct ggml_tensor * ggml_new_f32(struct ggml_context * ctx, float value);
|
||||
GGML_API struct ggml_tensor * ggml_new_i32(struct ggml_context * ctx, int32_t value);
|
||||
GGML_API struct ggml_tensor * ggml_new_f32(struct ggml_context * ctx, float value);
|
||||
|
||||
struct ggml_tensor * ggml_dup_tensor (struct ggml_context * ctx, const struct ggml_tensor * src);
|
||||
struct ggml_tensor * ggml_view_tensor(struct ggml_context * ctx, const struct ggml_tensor * src);
|
||||
GGML_API struct ggml_tensor * ggml_dup_tensor (struct ggml_context * ctx, const struct ggml_tensor * src);
|
||||
GGML_API struct ggml_tensor * ggml_view_tensor(struct ggml_context * ctx, const struct ggml_tensor * src);
|
||||
|
||||
struct ggml_tensor * ggml_set_zero(struct ggml_tensor * tensor);
|
||||
struct ggml_tensor * ggml_set_i32 (struct ggml_tensor * tensor, int32_t value);
|
||||
struct ggml_tensor * ggml_set_f32 (struct ggml_tensor * tensor, float value);
|
||||
GGML_API struct ggml_tensor * ggml_set_zero(struct ggml_tensor * tensor);
|
||||
GGML_API struct ggml_tensor * ggml_set_i32 (struct ggml_tensor * tensor, int32_t value);
|
||||
GGML_API struct ggml_tensor * ggml_set_f32 (struct ggml_tensor * tensor, float value);
|
||||
|
||||
int32_t ggml_get_i32_1d(const struct ggml_tensor * tensor, int i);
|
||||
void ggml_set_i32_1d(const struct ggml_tensor * tensor, int i, int32_t value);
|
||||
GGML_API int32_t ggml_get_i32_1d(const struct ggml_tensor * tensor, int i);
|
||||
GGML_API void ggml_set_i32_1d(const struct ggml_tensor * tensor, int i, int32_t value);
|
||||
|
||||
float ggml_get_f32_1d(const struct ggml_tensor * tensor, int i);
|
||||
void ggml_set_f32_1d(const struct ggml_tensor * tensor, int i, float value);
|
||||
GGML_API float ggml_get_f32_1d(const struct ggml_tensor * tensor, int i);
|
||||
GGML_API void ggml_set_f32_1d(const struct ggml_tensor * tensor, int i, float value);
|
||||
|
||||
void * ggml_get_data (const struct ggml_tensor * tensor);
|
||||
float * ggml_get_data_f32(const struct ggml_tensor * tensor);
|
||||
GGML_API void * ggml_get_data (const struct ggml_tensor * tensor);
|
||||
GGML_API float * ggml_get_data_f32(const struct ggml_tensor * tensor);
|
||||
|
||||
//
|
||||
// operations on tensors with backpropagation
|
||||
//
|
||||
|
||||
struct ggml_tensor * ggml_dup(
|
||||
GGML_API struct ggml_tensor * ggml_dup(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_add(
|
||||
GGML_API struct ggml_tensor * ggml_add(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
|
||||
struct ggml_tensor * ggml_add_inplace(
|
||||
GGML_API struct ggml_tensor * ggml_add_inplace(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_sub(
|
||||
GGML_API struct ggml_tensor * ggml_sub(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_mul(
|
||||
GGML_API struct ggml_tensor * ggml_mul(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_div(
|
||||
GGML_API struct ggml_tensor * ggml_div(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_sqr(
|
||||
GGML_API struct ggml_tensor * ggml_sqr(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_sqrt(
|
||||
GGML_API struct ggml_tensor * ggml_sqrt(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// return scalar
|
||||
// TODO: compute sum along rows
|
||||
struct ggml_tensor * ggml_sum(
|
||||
GGML_API struct ggml_tensor * ggml_sum(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// mean along rows
|
||||
struct ggml_tensor * ggml_mean(
|
||||
GGML_API struct ggml_tensor * ggml_mean(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// if a is the same shape as b, and a is not parameter, return a
|
||||
// otherwise, return a new tensor: repeat(a) to fit in b
|
||||
struct ggml_tensor * ggml_repeat(
|
||||
GGML_API struct ggml_tensor * ggml_repeat(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_abs(
|
||||
GGML_API struct ggml_tensor * ggml_abs(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_sgn(
|
||||
GGML_API struct ggml_tensor * ggml_sgn(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_neg(
|
||||
GGML_API struct ggml_tensor * ggml_neg(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_step(
|
||||
GGML_API struct ggml_tensor * ggml_step(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_relu(
|
||||
GGML_API struct ggml_tensor * ggml_relu(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// TODO: double-check this computation is correct
|
||||
struct ggml_tensor * ggml_gelu(
|
||||
GGML_API struct ggml_tensor * ggml_gelu(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_silu(
|
||||
GGML_API struct ggml_tensor * ggml_silu(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// normalize along rows
|
||||
// TODO: eps is hardcoded to 1e-5 for now
|
||||
struct ggml_tensor * ggml_norm(
|
||||
GGML_API struct ggml_tensor * ggml_norm(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_rms_norm(
|
||||
GGML_API struct ggml_tensor * ggml_rms_norm(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// A: m rows, n columns
|
||||
// B: p rows, n columns (i.e. we transpose it internally)
|
||||
// result is m columns, p rows
|
||||
struct ggml_tensor * ggml_mul_mat(
|
||||
GGML_API struct ggml_tensor * ggml_mul_mat(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
@ -533,32 +552,32 @@ struct ggml_tensor * ggml_mul_mat(
|
||||
//
|
||||
|
||||
// in-place, returns view(a)
|
||||
struct ggml_tensor * ggml_scale(
|
||||
GGML_API struct ggml_tensor * ggml_scale(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
// a -> b, return view(b)
|
||||
struct ggml_tensor * ggml_cpy(
|
||||
GGML_API struct ggml_tensor * ggml_cpy(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
// make contiguous
|
||||
struct ggml_tensor * ggml_cont(
|
||||
GGML_API struct ggml_tensor * ggml_cont(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
// return view(a), b specifies the new shape
|
||||
// TODO: when we start computing gradient, make a copy instead of view
|
||||
struct ggml_tensor * ggml_reshape(
|
||||
GGML_API struct ggml_tensor * ggml_reshape(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
// return view(a)
|
||||
// TODO: when we start computing gradient, make a copy instead of view
|
||||
struct ggml_tensor * ggml_reshape_2d(
|
||||
GGML_API struct ggml_tensor * ggml_reshape_2d(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int64_t ne0,
|
||||
@ -566,7 +585,7 @@ struct ggml_tensor * ggml_reshape_2d(
|
||||
|
||||
// return view(a)
|
||||
// TODO: when we start computing gradient, make a copy instead of view
|
||||
struct ggml_tensor * ggml_reshape_3d(
|
||||
GGML_API struct ggml_tensor * ggml_reshape_3d(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int64_t ne0,
|
||||
@ -574,13 +593,13 @@ struct ggml_tensor * ggml_reshape_3d(
|
||||
int64_t ne2);
|
||||
|
||||
// offset in bytes
|
||||
struct ggml_tensor * ggml_view_1d(
|
||||
GGML_API struct ggml_tensor * ggml_view_1d(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int64_t ne0,
|
||||
size_t offset);
|
||||
|
||||
struct ggml_tensor * ggml_view_2d(
|
||||
GGML_API struct ggml_tensor * ggml_view_2d(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int64_t ne0,
|
||||
@ -588,7 +607,7 @@ struct ggml_tensor * ggml_view_2d(
|
||||
size_t nb1, // row stride in bytes
|
||||
size_t offset);
|
||||
|
||||
struct ggml_tensor * ggml_view_3d(
|
||||
GGML_API struct ggml_tensor * ggml_view_3d(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int64_t ne0,
|
||||
@ -598,7 +617,7 @@ struct ggml_tensor * ggml_view_3d(
|
||||
size_t nb2, // slice stride in bytes
|
||||
size_t offset);
|
||||
|
||||
struct ggml_tensor * ggml_permute(
|
||||
GGML_API struct ggml_tensor * ggml_permute(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int axis0,
|
||||
@ -607,24 +626,24 @@ struct ggml_tensor * ggml_permute(
|
||||
int axis3);
|
||||
|
||||
// alias for ggml_permute(ctx, a, 1, 0, 2, 3)
|
||||
struct ggml_tensor * ggml_transpose(
|
||||
GGML_API struct ggml_tensor * ggml_transpose(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
struct ggml_tensor * ggml_get_rows(
|
||||
GGML_API struct ggml_tensor * ggml_get_rows(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
// set elements above the diagonal to -INF
|
||||
// in-place, returns view(a)
|
||||
struct ggml_tensor * ggml_diag_mask_inf(
|
||||
GGML_API struct ggml_tensor * ggml_diag_mask_inf(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int n_past);
|
||||
|
||||
// in-place, returns view(a)
|
||||
struct ggml_tensor * ggml_soft_max(
|
||||
GGML_API struct ggml_tensor * ggml_soft_max(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a);
|
||||
|
||||
@ -633,7 +652,7 @@ struct ggml_tensor * ggml_soft_max(
|
||||
// if mode & 1 == 1, skip n_past elements
|
||||
// if mode & 2 == 1, GPT-NeoX style
|
||||
// TODO: avoid creating a new tensor every time
|
||||
struct ggml_tensor * ggml_rope(
|
||||
GGML_API struct ggml_tensor * ggml_rope(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
int n_past,
|
||||
@ -644,24 +663,24 @@ struct ggml_tensor * ggml_rope(
|
||||
// TODO: we don't support extra parameters for now
|
||||
// that's why we are hard-coding the stride, padding, and dilation
|
||||
// not great ..
|
||||
struct ggml_tensor * ggml_conv_1d_1s(
|
||||
GGML_API struct ggml_tensor * ggml_conv_1d_1s(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_conv_1d_2s(
|
||||
GGML_API struct ggml_tensor * ggml_conv_1d_2s(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b);
|
||||
|
||||
struct ggml_tensor * ggml_flash_attn(
|
||||
GGML_API struct ggml_tensor * ggml_flash_attn(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * q,
|
||||
struct ggml_tensor * k,
|
||||
struct ggml_tensor * v,
|
||||
bool masked);
|
||||
|
||||
struct ggml_tensor * ggml_flash_ff(
|
||||
GGML_API struct ggml_tensor * ggml_flash_ff(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b0,
|
||||
@ -670,15 +689,15 @@ struct ggml_tensor * ggml_flash_ff(
|
||||
struct ggml_tensor * c1);
|
||||
|
||||
// Mapping operations
|
||||
typedef void (*ggml_unary_op_f32_t)(const int, float *, const float *);
|
||||
typedef void (*ggml_binary_op_f32_t)(const int, float *, const float *, const float *);
|
||||
GGML_API typedef void (*ggml_unary_op_f32_t)(const int, float *, const float *);
|
||||
GGML_API typedef void (*ggml_binary_op_f32_t)(const int, float *, const float *, const float *);
|
||||
|
||||
struct ggml_tensor * ggml_map_unary_f32(
|
||||
GGML_API struct ggml_tensor * ggml_map_unary_f32(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
const ggml_unary_op_f32_t fun);
|
||||
|
||||
struct ggml_tensor * ggml_map_binary_f32(
|
||||
GGML_API struct ggml_tensor * ggml_map_binary_f32(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * a,
|
||||
struct ggml_tensor * b,
|
||||
@ -688,23 +707,23 @@ struct ggml_tensor * ggml_map_binary_f32(
|
||||
// automatic differentiation
|
||||
//
|
||||
|
||||
void ggml_set_param(
|
||||
GGML_API void ggml_set_param(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_tensor * tensor);
|
||||
|
||||
void ggml_build_forward_expand(struct ggml_cgraph * cgraph, struct ggml_tensor * tensor);
|
||||
GGML_API void ggml_build_forward_expand(struct ggml_cgraph * cgraph, struct ggml_tensor * tensor);
|
||||
|
||||
struct ggml_cgraph ggml_build_forward (struct ggml_tensor * tensor);
|
||||
struct ggml_cgraph ggml_build_backward(struct ggml_context * ctx, struct ggml_cgraph * gf, bool keep);
|
||||
GGML_API struct ggml_cgraph ggml_build_forward (struct ggml_tensor * tensor);
|
||||
GGML_API struct ggml_cgraph ggml_build_backward(struct ggml_context * ctx, struct ggml_cgraph * gf, bool keep);
|
||||
|
||||
void ggml_graph_compute(struct ggml_context * ctx, struct ggml_cgraph * cgraph);
|
||||
void ggml_graph_reset (struct ggml_cgraph * cgraph);
|
||||
GGML_API void ggml_graph_compute(struct ggml_context * ctx, struct ggml_cgraph * cgraph);
|
||||
GGML_API void ggml_graph_reset (struct ggml_cgraph * cgraph);
|
||||
|
||||
// print info and performance information for the graph
|
||||
void ggml_graph_print(const struct ggml_cgraph * cgraph);
|
||||
GGML_API void ggml_graph_print(const struct ggml_cgraph * cgraph);
|
||||
|
||||
// dump the graph into a file using the dot format
|
||||
void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph * gf, const char * filename);
|
||||
GGML_API void ggml_graph_dump_dot(const struct ggml_cgraph * gb, const struct ggml_cgraph * gf, const char * filename);
|
||||
|
||||
//
|
||||
// optimization
|
||||
@ -797,10 +816,10 @@ struct ggml_opt_params {
|
||||
} lbfgs;
|
||||
};
|
||||
|
||||
struct ggml_opt_params ggml_opt_default_params(enum ggml_opt_type type);
|
||||
GGML_API struct ggml_opt_params ggml_opt_default_params(enum ggml_opt_type type);
|
||||
|
||||
// optimize the function defined by the tensor f
|
||||
enum ggml_opt_result ggml_opt(
|
||||
GGML_API enum ggml_opt_result ggml_opt(
|
||||
struct ggml_context * ctx,
|
||||
struct ggml_opt_params params,
|
||||
struct ggml_tensor * f);
|
||||
@ -809,32 +828,32 @@ enum ggml_opt_result ggml_opt(
|
||||
// quantization
|
||||
//
|
||||
|
||||
size_t ggml_quantize_q4_0(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
size_t ggml_quantize_q4_1(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
size_t ggml_quantize_q4_2(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
size_t ggml_quantize_q4_3(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
GGML_API size_t ggml_quantize_q4_0(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
GGML_API size_t ggml_quantize_q4_1(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
GGML_API size_t ggml_quantize_q4_2(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
GGML_API size_t ggml_quantize_q4_3(const float * src, void * dst, int n, int k, int64_t * hist);
|
||||
|
||||
size_t ggml_quantize_chunk(enum ggml_type type, const float * src, void * dst, int start, int n, int64_t * hist);
|
||||
GGML_API size_t ggml_quantize_chunk(enum ggml_type type, const float * src, void * dst, int start, int n, int64_t * hist);
|
||||
|
||||
//
|
||||
// system info
|
||||
//
|
||||
|
||||
int ggml_cpu_has_avx(void);
|
||||
int ggml_cpu_has_avx2(void);
|
||||
int ggml_cpu_has_avx512(void);
|
||||
int ggml_cpu_has_avx512_vbmi(void);
|
||||
int ggml_cpu_has_avx512_vnni(void);
|
||||
int ggml_cpu_has_fma(void);
|
||||
int ggml_cpu_has_neon(void);
|
||||
int ggml_cpu_has_arm_fma(void);
|
||||
int ggml_cpu_has_f16c(void);
|
||||
int ggml_cpu_has_fp16_va(void);
|
||||
int ggml_cpu_has_wasm_simd(void);
|
||||
int ggml_cpu_has_blas(void);
|
||||
int ggml_cpu_has_cublas(void);
|
||||
int ggml_cpu_has_sse3(void);
|
||||
int ggml_cpu_has_vsx(void);
|
||||
GGML_API int ggml_cpu_has_avx (void);
|
||||
GGML_API int ggml_cpu_has_avx2 (void);
|
||||
GGML_API int ggml_cpu_has_avx512 (void);
|
||||
GGML_API int ggml_cpu_has_avx512_vbmi(void);
|
||||
GGML_API int ggml_cpu_has_avx512_vnni(void);
|
||||
GGML_API int ggml_cpu_has_fma (void);
|
||||
GGML_API int ggml_cpu_has_neon (void);
|
||||
GGML_API int ggml_cpu_has_arm_fma (void);
|
||||
GGML_API int ggml_cpu_has_f16c (void);
|
||||
GGML_API int ggml_cpu_has_fp16_va (void);
|
||||
GGML_API int ggml_cpu_has_wasm_simd (void);
|
||||
GGML_API int ggml_cpu_has_blas (void);
|
||||
GGML_API int ggml_cpu_has_cublas (void);
|
||||
GGML_API int ggml_cpu_has_sse3 (void);
|
||||
GGML_API int ggml_cpu_has_vsx (void);
|
||||
|
||||
|
||||
//
|
||||
|
||||
Loading…
Reference in New Issue
Block a user