vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-11 01:24:36 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Merge 'origin/master' into hipblas

Browse source
This commit is contained in:
Henri Vasserman 2023-05-01 22:45:29 +03:00
commit fcbc262eb9
No known key found for this signature in database
GPG key ID: 2995FC0F58B1A986
29 changed files with 1285 additions and 472 deletions
Download patch file Download diff file Expand all files Collapse all files

46
ggml-cuda.h
View file

@ -2,6 +2,7 @@
#include "hipblas/hipblas.h"
#include "hip/hip_runtime.h"
#define CUBLAS_COMPUTE_32F HIPBLAS_R_32F
#define CUBLAS_COMPUTE_32F_FAST_16F HIPBLAS_R_32F
#define CUBLAS_GEMM_DEFAULT HIPBLAS_GEMM_DEFAULT
#define CUBLAS_OP_N HIPBLAS_OP_N
#define CUBLAS_OP_T HIPBLAS_OP_T
@ -49,49 +50,16 @@
extern "C" {
#endif
#define CUDA_CHECK(err) \
do { \
cudaError_t err_ = (err); \
if (err_ != cudaSuccess) { \
fprintf(stderr, "CUDA error %d at %s:%d: %s\n", err_, __FILE__, __LINE__, \
cudaGetErrorString(err_)); \
exit(1); \
} \
} while (0)
#define CUBLAS_CHECK(err) \
do { \
cublasStatus_t err_ = (err); \
if (err_ != CUBLAS_STATUS_SUCCESS) { \
fprintf(stderr, "cuBLAS error %d at %s:%d\n", err_, __FILE__, __LINE__); \
exit(1); \
} \
} while (0)
extern cublasHandle_t g_cublasH;
extern cudaStream_t g_cudaStream;
extern cudaStream_t g_cudaStream2;
extern cudaEvent_t g_cudaEvent;
void ggml_init_cublas(void);
bool ggml_cuda_can_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
size_t ggml_cuda_mul_mat_get_wsize(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst);
void ggml_cuda_mul_mat(const struct ggml_tensor * src0, const struct ggml_tensor * src1, struct ggml_tensor * dst, void * wdata, size_t wsize);
// TODO: export these with GGML_API
void * ggml_cuda_host_malloc(size_t size);
void ggml_cuda_host_free(void * ptr);
void * ggml_cuda_pool_malloc(size_t size, size_t * actual_size);
void ggml_cuda_pool_free(void * ptr, size_t size);
void dequantize_row_q4_0_cuda(const void * vx, float * y, int k, cudaStream_t stream);
void dequantize_row_q4_1_cuda(const void * vx, float * y, int k, cudaStream_t stream);
void dequantize_row_q4_2_cuda(const void * vx, float * y, int k, cudaStream_t stream);
void dequantize_row_q5_0_cuda(const void * vx, float * y, int k, cudaStream_t stream);
void dequantize_row_q5_1_cuda(const void * vx, float * y, int k, cudaStream_t stream);
void dequantize_row_q8_0_cuda(const void * vx, float * y, int k, cudaStream_t stream);
cudaError_t ggml_cuda_h2d_tensor_2d(void * dst, const struct ggml_tensor * src, uint64_t i3, uint64_t i2, cudaStream_t stream);
typedef void (*dequantize_row_q_cuda_t)(const void * x, float * y, int k, cudaStream_t stream);
dequantize_row_q_cuda_t ggml_get_dequantize_row_q_cuda(enum ggml_type type);
#ifdef __cplusplus
}
#endif