ver4a/llama.cpp
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

CUDA: remove DMMV, consolidate F16 mult mat vec (#10318)

Browse source
This commit is contained in:
Johannes Gäßler 2024-11-17 09:09:55 +01:00 • committed by GitHub
parent 467576b6cc
commit c3ea58aca4
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
10 changed files with 246 additions and 1000 deletions
Download patch file Download diff file Expand all files Collapse all files

20
ggml/src/ggml-cuda/dmmv.cuh
View file

@ -1,20 +0,0 @@
#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(
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 char * src1_ddq_i, float * dst_dd_i, const int64_t row_low, const int64_t row_high, const int64_t src1_ncols,
const int64_t src1_padded_row_size, cudaStream_t stream);
bool ggml_cuda_dmmv_type_supported(ggml_type src0_type);