vbatts/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

CUBLAS_TF32_TENSOR_OP_MATH is not a macro

Browse source
This commit is contained in:
slaren 2023-12-23 17:57:25 +01:00
parent b7da1ba00e
commit d8b06c2148
1 changed files with 4 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

7
ggml-cuda.cu
View file

@ -91,13 +91,14 @@
#include <cuda.h>
#include <cublas_v2.h>
#include <cuda_fp16.h>
// CUDA 10.2 does not have these macro definitions.
#ifndef CUBLAS_TF32_TENSOR_OP_MATH
#if CUDART_VERSION < 11020
#define CUBLAS_TF32_TENSOR_OP_MATH CUBLAS_TENSOR_OP_MATH
#define CUBLAS_COMPUTE_16F CUDA_R_16F
#define CUBLAS_COMPUTE_32F CUDA_R_32F
#define cublasComputeType_t cudaDataType_t
#endif
#endif // CUDART_VERSION < 11020
#endif // defined(GGML_USE_HIPBLAS)
#include "ggml-cuda.h"