mirror of
https://github.com/ggerganov/whisper.cpp.git
synced 2025-02-22 09:40:52 +00:00
sync : llama.cpp (skip)
ggml-ci
This commit is contained in:
parent
61b05815e0
commit
97878e53fd
@ -231,12 +231,6 @@ typedef float dfloat; // dequantize float
|
|||||||
typedef float2 dfloat2;
|
typedef float2 dfloat2;
|
||||||
#endif //GGML_CUDA_F16
|
#endif //GGML_CUDA_F16
|
||||||
|
|
||||||
// dmmv = dequantize_mul_mat_vec
|
|
||||||
// TODO: remove this?
|
|
||||||
#ifndef GGML_CUDA_DMMV_X
|
|
||||||
#define GGML_CUDA_DMMV_X 32
|
|
||||||
#endif
|
|
||||||
|
|
||||||
[[noreturn]]
|
[[noreturn]]
|
||||||
static __device__ void no_device_code(
|
static __device__ void no_device_code(
|
||||||
const char * file_name, const int line, const char * function_name, const int arch, const char * arch_list) {
|
const char * file_name, const int line, const char * function_name, const int arch, const char * arch_list) {
|
||||||
|
@ -2,10 +2,6 @@
|
|||||||
#include "dequantize.cuh"
|
#include "dequantize.cuh"
|
||||||
#include "convert.cuh"
|
#include "convert.cuh"
|
||||||
|
|
||||||
#ifndef GGML_CUDA_MMV_Y
|
|
||||||
#define GGML_CUDA_MMV_Y 1
|
|
||||||
#endif
|
|
||||||
|
|
||||||
#ifndef K_QUANTS_PER_ITERATION
|
#ifndef K_QUANTS_PER_ITERATION
|
||||||
#define K_QUANTS_PER_ITERATION 2
|
#define K_QUANTS_PER_ITERATION 2
|
||||||
#else
|
#else
|
||||||
|
@ -1,5 +1,16 @@
|
|||||||
#include "common.cuh"
|
#include "common.cuh"
|
||||||
|
|
||||||
|
// dmmv = dequantize_mul_mat_vec
|
||||||
|
|
||||||
|
// TODO: remove this?
|
||||||
|
#ifndef GGML_CUDA_DMMV_X
|
||||||
|
#define GGML_CUDA_DMMV_X 32
|
||||||
|
#endif
|
||||||
|
|
||||||
|
#ifndef GGML_CUDA_MMV_Y
|
||||||
|
#define GGML_CUDA_MMV_Y 1
|
||||||
|
#endif
|
||||||
|
|
||||||
void ggml_cuda_op_dequantize_mul_mat_vec(
|
void ggml_cuda_op_dequantize_mul_mat_vec(
|
||||||
ggml_backend_cuda_context & ctx,
|
ggml_backend_cuda_context & ctx,
|
||||||
const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst, const char * src0_dd_i, const float * src1_ddf_i,
|
const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst, const char * src0_dd_i, const float * src1_ddf_i,
|
||||||
|
Loading…
x
Reference in New Issue
Block a user