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

iq4_nl: another fix after merging with master

Browse source
This commit is contained in:
Iwan Kawrakow 2024-02-19 11:40:35 +02:00
parent 1d90021241
commit e7b999c3dc
1 changed files with 1 additions and 0 deletions
Download patch file Download diff file Expand all files Collapse all files

1
ggml-cuda.cu
View file

@ -6852,6 +6852,7 @@ static void dequantize_row_iq1_s_cuda(const void * vx, dst_t * y, const int k, c
dequantize_block_iq1_s<<<nb, 32, 0, stream>>>(vx, y); dequantize_block_iq1_s<<<nb, 32, 0, stream>>>(vx, y);
} }
template<typename dst_t>
static void dequantize_row_iq4_nl_cuda(const void * vx, dst_t * y, const int k, cudaStream_t stream) { static void dequantize_row_iq4_nl_cuda(const void * vx, dst_t * y, const int k, cudaStream_t stream) {
const int nb = (k + QK_K - 1) / QK_K; const int nb = (k + QK_K - 1) / QK_K;
dequantize_block_iq4_nl<<<nb, 32, 0, stream>>>(vx, y); dequantize_block_iq4_nl<<<nb, 32, 0, stream>>>(vx, y);