mirror of
https://github.com/LostRuins/koboldcpp.git
synced 2026-09-11 07:19:10 +02:00
Concedo
d1bb126605
# Conflicts: # README.md # llama.cpp # otherarch/sdcpp/SDCPP_LICENSE # scripts/sync-ggml-am.sh # scripts/sync-ggml.sh
14 lines
391 B
Plaintext
14 lines
391 B
Plaintext
#include "common.cuh"
|
|
|
|
#define CUDA_DEQUANTIZE_BLOCK_SIZE 256
|
|
|
|
template<typename T>
|
|
using to_t_cuda_t = void (*)(const void * __restrict__ x, T * __restrict__ y, int64_t k, cudaStream_t stream);
|
|
|
|
typedef to_t_cuda_t<float> to_fp32_cuda_t;
|
|
typedef to_t_cuda_t<half> to_fp16_cuda_t;
|
|
|
|
to_fp16_cuda_t ggml_get_to_fp16_cuda(ggml_type type);
|
|
|
|
to_fp32_cuda_t ggml_get_to_fp32_cuda(ggml_type type);
|