CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-11-18 11:46:58 +00:00
Code Issues Packages Projects Releases Wiki Activity

Merge branch 'master' into mpi

Browse Source
This commit is contained in:
Evan Miller
2023-07-06 19:04:50 -04:00
parent 042c5b278f dfd9fce6d6
commit 32deabfdc8
33 changed files with 5794 additions and 2048 deletions
Download Patch File Download Diff File Expand all files Collapse all files

145
ggml.h
View File

@@ -201,6 +201,8 @@
#define GGML_MAX_NAME 48
#define GGML_DEFAULT_N_THREADS 4
#define GGML_UNUSED(x) (void)(x)
#define GGML_ASSERT(x) \
do { \
if (!(x)) { \
@@ -209,6 +211,30 @@
} \
} while (0)
// used to copy the number of elements and stride in bytes of tensors into local variables.
// main purpose is to reduce code duplication and improve readability.
//
// example:
//
// GGML_TENSOR_LOCALS(int64_t, ne1, src1, ne);
// GGML_TENSOR_LOCALS(size_t, nb1, src1, nb);
//
#define GGML_TENSOR_LOCALS_1(type, prefix, pointer, array) \
const type prefix##0 = (pointer)->array[0]; \
GGML_UNUSED(prefix##0);
#define GGML_TENSOR_LOCALS_2(type, prefix, pointer, array) \
GGML_TENSOR_LOCALS_1 (type, prefix, pointer, array) \
const type prefix##1 = (pointer)->array[1]; \
GGML_UNUSED(prefix##1);
#define GGML_TENSOR_LOCALS_3(type, prefix, pointer, array) \
GGML_TENSOR_LOCALS_2 (type, prefix, pointer, array) \
const type prefix##2 = (pointer)->array[2]; \
GGML_UNUSED(prefix##2);
#define GGML_TENSOR_LOCALS(type, prefix, pointer, array) \
GGML_TENSOR_LOCALS_3 (type, prefix, pointer, array) \
const type prefix##3 = (pointer)->array[3]; \
GGML_UNUSED(prefix##3);
#ifdef __cplusplus
extern "C" {
#endif
@@ -224,8 +250,8 @@ extern "C" {
GGML_API float ggml_fp16_to_fp32(ggml_fp16_t x);
GGML_API ggml_fp16_t ggml_fp32_to_fp16(float x);
GGML_API void ggml_fp16_to_fp32_row(const ggml_fp16_t * x, float * y, size_t n);
GGML_API void ggml_fp32_to_fp16_row(const float * x, ggml_fp16_t * y, size_t n);
GGML_API void ggml_fp16_to_fp32_row(const ggml_fp16_t * x, float * y, int n);
GGML_API void ggml_fp32_to_fp16_row(const float * x, ggml_fp16_t * y, int n);
struct ggml_object;
struct ggml_context;
@@ -295,12 +321,15 @@ extern "C" {
GGML_OP_SUM,
GGML_OP_SUM_ROWS,
GGML_OP_MEAN,
GGML_OP_ARGMAX,
GGML_OP_REPEAT,
GGML_OP_REPEAT_BACK,
GGML_OP_ABS,
GGML_OP_SGN,
GGML_OP_NEG,
GGML_OP_STEP,
GGML_OP_TANH,
GGML_OP_ELU,
GGML_OP_RELU,
GGML_OP_GELU,
GGML_OP_GELU_QUICK,
@@ -332,9 +361,8 @@ extern "C" {
GGML_OP_ROPE_BACK,
GGML_OP_ALIBI,
GGML_OP_CLAMP,
GGML_OP_CONV_1D_S1_PH,
GGML_OP_CONV_1D_S2_PH,
GGML_OP_CONV_2D_SK_P0,
GGML_OP_CONV_1D,
GGML_OP_CONV_2D,
GGML_OP_FLASH_ATTN,
GGML_OP_FLASH_FF,
@@ -703,6 +731,11 @@ extern "C" {
struct ggml_context * ctx,
struct ggml_tensor * a);
// argmax along rows
GGML_API struct ggml_tensor * ggml_argmax(
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
GGML_API struct ggml_tensor * ggml_repeat(
@@ -747,6 +780,22 @@ extern "C" {
struct ggml_context * ctx,
struct ggml_tensor * a);
GGML_API struct ggml_tensor * ggml_tanh(
struct ggml_context * ctx,
struct ggml_tensor * a);
GGML_API struct ggml_tensor * ggml_tanh_inplace(
struct ggml_context * ctx,
struct ggml_tensor * a);
GGML_API struct ggml_tensor * ggml_elu(
struct ggml_context * ctx,
struct ggml_tensor * a);
GGML_API struct ggml_tensor * ggml_elu_inplace(
struct ggml_context * ctx,
struct ggml_tensor * a);
GGML_API struct ggml_tensor * ggml_relu(
struct ggml_context * ctx,
struct ggml_tensor * a);
@@ -1097,58 +1146,33 @@ extern "C" {
float min,
float max);
// TODO: implement general-purpose convolutions
// GGML_API struct ggml_tensor * ggml_conv_1d(
// struct ggml_context * ctx,
// struct ggml_tensor * a,
// struct ggml_tensor * b,
// int s0
// int p0,
// int d0);
//
// GGML_API struct ggml_tensor * ggml_conv_2d(
// struct ggml_context * ctx,
// struct ggml_tensor * a,
// struct ggml_tensor * b,
// int s0,
// int s1,
// int p0,
// int p1,
// int d0,
// int d1);
// padding = half
// TODO: we don't support extra parameters for now
// that's why we are hard-coding the stride, padding, and dilation
// not great ..
// example:
// a: 3 80 768 1
// b: 3000 80 1 1
// res: 3000 768 1 1
// used in whisper
GGML_API struct ggml_tensor * ggml_conv_1d_s1_ph(
GGML_API struct ggml_tensor * ggml_conv_1d(
struct ggml_context * ctx,
struct ggml_tensor * a,
struct ggml_tensor * b);
struct ggml_tensor * b,
int s0, // stride
int p0, // padding
int d0); // dilation
// used in whisper
GGML_API struct ggml_tensor * ggml_conv_1d_s2_ph(
GGML_API struct ggml_tensor * ggml_conv_2d(
struct ggml_context * ctx,
struct ggml_tensor * a,
struct ggml_tensor * b);
struct ggml_tensor * b,
int s0,
int s1,
int p0,
int p1,
int d0,
int d1);
// kernel size is a->ne[0] x a->ne[1]
// stride is equal to kernel size
// padding is zero
// example:
// a: 16 16 3 768
// b: 1024 1024 3 1
// res: 64 64 768 1
// used in sam
GGML_API struct ggml_tensor * ggml_conv_2d_sk_p0(
// conv_1d with padding = half
// alias for ggml_conv_1d(a, b, s, a->ne[0]/2, d)
GGML_API struct ggml_tensor* ggml_conv_1d_ph(
struct ggml_context * ctx,
struct ggml_tensor * a,
struct ggml_tensor * b);
struct ggml_tensor * b,
int s,
int d);
GGML_API struct ggml_tensor * ggml_flash_attn(
struct ggml_context * ctx,
@@ -1504,25 +1528,24 @@ extern "C" {
//
#ifdef __cplusplus
// restrict not standard in C++
// restrict not standard in C++
#define GGML_RESTRICT
#else
#define GGML_RESTRICT restrict
#endif
typedef void (*dequantize_row_q_t)(const void * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
typedef void (*quantize_row_q_t) (const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
typedef void (*vec_dot_q_t) (const int n, float * GGML_RESTRICT s, const void * GGML_RESTRICT x, const void * GGML_RESTRICT y);
typedef void (*ggml_to_float_t) (const void * GGML_RESTRICT x, float * GGML_RESTRICT y, int k);
typedef void (*ggml_from_float_t)(const float * GGML_RESTRICT x, void * GGML_RESTRICT y, int k);
typedef void (*ggml_vec_dot_t) (const int n, float * GGML_RESTRICT s, const void * GGML_RESTRICT x, const void * GGML_RESTRICT y);
typedef struct {
dequantize_row_q_t dequantize_row_q;
quantize_row_q_t quantize_row_q;
quantize_row_q_t quantize_row_q_reference;
quantize_row_q_t quantize_row_q_dot;
vec_dot_q_t vec_dot_q;
enum ggml_type vec_dot_type;
} quantize_fns_t;
ggml_to_float_t to_float;
ggml_from_float_t from_float;
ggml_from_float_t from_float_reference;
ggml_vec_dot_t vec_dot;
enum ggml_type vec_dot_type;
} ggml_type_traits_t;
quantize_fns_t ggml_internal_get_quantize_fn(size_t i);
ggml_type_traits_t ggml_internal_get_type_traits(enum ggml_type i);
#ifdef __cplusplus
}