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

ggml : fix Q5_0 qh -> uint32_t

Browse source
This commit is contained in:
Georgi Gerganov 2023-04-26 10:43:26 +03:00
parent 5bebc0a6e2
commit 2576c16f00
No known key found for this signature in database
GPG key ID: 449E073F9DC10735
2 changed files with 4 additions and 4 deletions
Download patch file Download diff file Expand all files Collapse all files

4
ggml-cuda.cu
View file

@ -41,10 +41,10 @@ static_assert(sizeof(block_q4_3) == 2 * sizeof(ggml_fp16_t) + QK4_3 / 2, "wrong
typedef struct {
__half d; // delta
__half m; // min
int32_t qh; // 5-th bit of quants
uint32_t qh; // 5-th bit of quants
uint8_t qs[QK5_0 / 2]; // nibbles / quants
} block_q5_0;
static_assert(sizeof(block_q5_0) == 2 * sizeof(ggml_fp16_t) + sizeof(int32_t) + QK5_0 / 2, "wrong q5_0 block size/padding");
static_assert(sizeof(block_q5_0) == 2 * sizeof(ggml_fp16_t) + sizeof(uint32_t) + QK5_0 / 2, "wrong q5_0 block size/padding");
#define QK8_0 32
typedef struct {

4
ggml.c
View file

@ -677,10 +677,10 @@ static_assert(sizeof(block_q4_3) == 2 * sizeof(ggml_fp16_t) + QK4_3 / 2, "wrong
typedef struct {
ggml_fp16_t d; // delta
ggml_fp16_t m; // min
int32_t qh; // 5-th bit of quants
uint32_t qh; // 5-th bit of quants
uint8_t qs[QK5_0 / 2]; // nibbles / quants
} block_q5_0;
static_assert(sizeof(block_q5_0) == 2 * sizeof(ggml_fp16_t) + sizeof(int32_t) + QK5_0 / 2, "wrong q5_0 block size/padding");
static_assert(sizeof(block_q5_0) == 2 * sizeof(ggml_fp16_t) + sizeof(uint32_t) + QK5_0 / 2, "wrong q5_0 block size/padding");
#define QK8_0 32
typedef struct {