CS348Project/llama.cpp
5
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2025-10-31 08:51:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

Avoid using __fp16 on ARM with old nvcc (#10616)

Browse Source
This commit is contained in:
Frankie Robertson
2024-12-04 02:41:37 +02:00
committed by GitHub
parent da6aac91f1
commit cd2f37b304
1 changed files with 2 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
ggml/src/ggml-impl.h
View File

@@ -310,14 +310,14 @@ void ggml_aligned_free(void * ptr, size_t size);
// FP16 to FP32 conversion
#if defined(__ARM_NEON)
#ifdef _MSC_VER
#if defined(_MSC_VER) || (defined(__CUDACC__) && __CUDACC_VER_MAJOR__ <= 11)
typedef uint16_t ggml_fp16_internal_t;
#else
typedef __fp16 ggml_fp16_internal_t;
#endif
#endif
#if defined(__ARM_NEON) && !defined(_MSC_VER)
#if defined(__ARM_NEON) && !defined(_MSC_VER) && !(defined(__CUDACC__) && __CUDACC_VER_MAJOR__ <= 11)
#define GGML_COMPUTE_FP16_TO_FP32(x) ggml_compute_fp16_to_fp32(x)
#define GGML_COMPUTE_FP32_TO_FP16(x) ggml_compute_fp32_to_fp16(x)