diff options
author | Iwan Kawrakow <iwan.kawrakow@gmail.com> | 2024-08-01 16:08:32 +0300 |
---|---|---|
committer | Kawrakow <48489457+ikawrakow@users.noreply.github.com> | 2024-08-09 16:00:31 +0200 |
commit | cfb0410067be051dd8ea76c08280a3b04d5a5188 (patch) | |
tree | 64ffc06e18f75a8cfc4d04b98875f68cd7fa99d4 /ggml/src/ggml-common.h | |
parent | a9f302ebe2373321c12b01d8760904901aa064a4 (diff) |
iq6_k: WIP (nothing works)
Diffstat (limited to 'ggml/src/ggml-common.h')
-rw-r--r-- | ggml/src/ggml-common.h | 12 |
1 files changed, 12 insertions, 0 deletions
diff --git a/ggml/src/ggml-common.h b/ggml/src/ggml-common.h index 5847d903..2fbac06a 100644 --- a/ggml/src/ggml-common.h +++ b/ggml/src/ggml-common.h @@ -142,6 +142,9 @@ typedef sycl::half2 ggml_half2; #define QI5_XS (QK_K / (4*QR5_XS)) #define QR5_XS 2 +#define QI6_XS (QK_K / (4*QR6_XS)) +#define QR6_XS 2 + #define QI3_S (QK_K / (4*QR3_S)) #define QR3_S 4 @@ -493,6 +496,15 @@ typedef struct { } block_iq5_k; static_assert(sizeof(block_iq5_k) == sizeof(ggml_half) + sizeof(uint16_t) + QK_K/2 + QK_K/8 + 3*QK_K/64, "wrong iq5_k block size/padding"); +typedef struct { + ggml_half d; + uint16_t extra; + int8_t scales[QK_K/16]; + uint8_t qs[QK_K/2]; + uint8_t qh[QK_K/4]; +} block_iq6_k; +static_assert(sizeof(block_iq6_k) == sizeof(ggml_half) + sizeof(uint16_t) + QK_K/2 + QK_K/4 + QK_K/16, "wrong iq6_k block size/padding"); + #endif // GGML_COMMON_DECL #endif // GGML_COMMON_DECL |