This commit is contained in:
Concedo 2024-07-14 22:22:29 +08:00
parent 7fb2ae3193
commit 33a3beea32
2 changed files with 104406 additions and 104162 deletions

File diff suppressed because it is too large Load diff

View file

@ -1,178 +1,184 @@
#include <cstdint>
extern unsigned char matmul_f32_f16_aligned_fp32_data[10240];
const uint64_t matmul_f32_f16_aligned_fp32_len = 10240;
extern unsigned char matmul_f32_f32_fp32_data[10208];
const uint64_t matmul_f32_f32_fp32_len = 10208;
extern unsigned char matmul_f32_f16_fp32_data[10260];
const uint64_t matmul_f32_f16_fp32_len = 10260;
extern unsigned char matmul_f16_fp32_data[10276];
const uint64_t matmul_f16_fp32_len = 10276;
extern unsigned char matmul_f16_aligned_fp32_data[10288];
const uint64_t matmul_f16_aligned_fp32_len = 10288;
extern unsigned char matmul_f16_f32_fp32_data[10260];
const uint64_t matmul_f16_f32_fp32_len = 10260;
extern unsigned char matmul_q5_0_f32_fp32_data[10956];
const uint64_t matmul_q5_0_f32_fp32_len = 10956;
extern unsigned char matmul_q4_0_f32_aligned_fp32_data[10408];
const uint64_t matmul_q4_0_f32_aligned_fp32_len = 10408;
extern unsigned char matmul_f32_f16_fp32_data[10260];
const uint64_t matmul_f32_f16_fp32_len = 10260;
extern unsigned char matmul_q4_1_f32_fp32_data[10512];
const uint64_t matmul_q4_1_f32_fp32_len = 10512;
extern unsigned char matmul_f32_f32_fp32_data[10208];
const uint64_t matmul_f32_f32_fp32_len = 10208;
extern unsigned char matmul_f32_f32_aligned_fp32_data[10124];
const uint64_t matmul_f32_f32_aligned_fp32_len = 10124;
extern unsigned char matmul_f16_f32_aligned_fp32_data[10240];
const uint64_t matmul_f16_f32_aligned_fp32_len = 10240;
extern unsigned char matmul_q4_0_f32_fp32_data[10444];
const uint64_t matmul_q4_0_f32_fp32_len = 10444;
extern unsigned char matmul_q4_1_f32_aligned_fp32_data[10476];
const uint64_t matmul_q4_1_f32_aligned_fp32_len = 10476;
extern unsigned char matmul_q5_1_f32_fp32_data[10836];
const uint64_t matmul_q5_1_f32_fp32_len = 10836;
extern unsigned char matmul_q3_k_f32_aligned_fp32_data[12972];
const uint64_t matmul_q3_k_f32_aligned_fp32_len = 12972;
extern unsigned char matmul_q8_0_f32_aligned_fp32_data[10452];
const uint64_t matmul_q8_0_f32_aligned_fp32_len = 10452;
extern unsigned char matmul_q5_k_f32_fp32_data[12456];
const uint64_t matmul_q5_k_f32_fp32_len = 12456;
extern unsigned char matmul_q6_k_f32_fp32_data[11624];
const uint64_t matmul_q6_k_f32_fp32_len = 11624;
extern unsigned char matmul_id_f16_fp32_data[10748];
const uint64_t matmul_id_f16_fp32_len = 10748;
extern unsigned char matmul_q8_0_f32_fp32_data[10468];
const uint64_t matmul_q8_0_f32_fp32_len = 10468;
extern unsigned char matmul_f32_f16_aligned_fp32_data[10240];
const uint64_t matmul_f32_f16_aligned_fp32_len = 10240;
extern unsigned char matmul_q5_0_f32_aligned_fp32_data[10920];
const uint64_t matmul_q5_0_f32_aligned_fp32_len = 10920;
extern unsigned char matmul_q3_k_f32_fp32_data[13004];
const uint64_t matmul_q3_k_f32_fp32_len = 13004;
extern unsigned char matmul_q4_0_f32_fp32_data[10444];
const uint64_t matmul_q4_0_f32_fp32_len = 10444;
extern unsigned char matmul_q6_k_f32_aligned_fp32_data[11592];
const uint64_t matmul_q6_k_f32_aligned_fp32_len = 11592;
extern unsigned char matmul_f16_fp32_data[10276];
const uint64_t matmul_f16_fp32_len = 10276;
extern unsigned char matmul_id_f32_f16_aligned_fp32_data[10928];
const uint64_t matmul_id_f32_f16_aligned_fp32_len = 10928;
extern unsigned char matmul_q4_0_f32_aligned_fp32_data[10408];
const uint64_t matmul_q4_0_f32_aligned_fp32_len = 10408;
extern unsigned char matmul_id_f32_f32_fp32_data[10680];
const uint64_t matmul_id_f32_f32_fp32_len = 10680;
extern unsigned char matmul_q4_1_f32_aligned_fp32_data[10476];
const uint64_t matmul_q4_1_f32_aligned_fp32_len = 10476;
extern unsigned char matmul_q2_k_f32_aligned_fp32_data[11068];
const uint64_t matmul_q2_k_f32_aligned_fp32_len = 11068;
extern unsigned char matmul_q5_0_f32_fp32_data[10956];
const uint64_t matmul_q5_0_f32_fp32_len = 10956;
extern unsigned char matmul_f16_f32_fp32_data[10308];
const uint64_t matmul_f16_f32_fp32_len = 10308;
extern unsigned char matmul_q5_k_f32_aligned_fp32_data[12424];
const uint64_t matmul_q5_k_f32_aligned_fp32_len = 12424;
extern unsigned char matmul_q8_0_f32_fp32_data[10488];
const uint64_t matmul_q8_0_f32_fp32_len = 10488;
extern unsigned char matmul_q5_k_f32_fp32_data[12476];
const uint64_t matmul_q5_k_f32_fp32_len = 12476;
extern unsigned char matmul_id_f16_fp32_data[10748];
const uint64_t matmul_id_f16_fp32_len = 10748;
extern unsigned char matmul_q4_k_f32_fp32_data[11952];
const uint64_t matmul_q4_k_f32_fp32_len = 11952;
extern unsigned char matmul_q8_0_f32_aligned_fp32_data[10452];
const uint64_t matmul_q8_0_f32_aligned_fp32_len = 10452;
extern unsigned char matmul_q5_1_f32_fp32_data[10836];
const uint64_t matmul_q5_1_f32_fp32_len = 10836;
extern unsigned char matmul_id_f32_f16_aligned_fp32_data[10928];
const uint64_t matmul_id_f32_f16_aligned_fp32_len = 10928;
extern unsigned char matmul_q4_k_f32_aligned_fp32_data[11900];
const uint64_t matmul_q4_k_f32_aligned_fp32_len = 11900;
extern unsigned char matmul_q6_k_f32_aligned_fp32_data[11592];
const uint64_t matmul_q6_k_f32_aligned_fp32_len = 11592;
extern unsigned char matmul_q6_k_f32_fp32_data[11644];
const uint64_t matmul_q6_k_f32_fp32_len = 11644;
extern unsigned char matmul_q2_k_f32_fp32_data[11120];
const uint64_t matmul_q2_k_f32_fp32_len = 11120;
extern unsigned char matmul_q2_k_f32_aligned_fp32_data[11068];
const uint64_t matmul_q2_k_f32_aligned_fp32_len = 11068;
extern unsigned char matmul_id_f32_f16_fp32_data[10732];
const uint64_t matmul_id_f32_f16_fp32_len = 10732;
extern unsigned char matmul_id_f16_aligned_fp32_data[10976];
const uint64_t matmul_id_f16_aligned_fp32_len = 10976;
extern unsigned char matmul_q3_k_f32_aligned_fp32_data[12972];
const uint64_t matmul_q3_k_f32_aligned_fp32_len = 12972;
extern unsigned char matmul_id_f16_f32_aligned_fp32_data[10928];
const uint64_t matmul_id_f16_f32_aligned_fp32_len = 10928;
extern unsigned char matmul_id_f16_f32_fp32_data[10780];
const uint64_t matmul_id_f16_f32_fp32_len = 10780;
extern unsigned char matmul_id_f32_f32_fp32_data[10680];
const uint64_t matmul_id_f32_f32_fp32_len = 10680;
extern unsigned char matmul_id_f32_f32_aligned_fp32_data[10812];
const uint64_t matmul_id_f32_f32_aligned_fp32_len = 10812;
extern unsigned char matmul_id_f16_f32_fp32_data[10732];
const uint64_t matmul_id_f16_f32_fp32_len = 10732;
extern unsigned char matmul_id_q8_0_f32_aligned_fp32_data[11140];
const uint64_t matmul_id_q8_0_f32_aligned_fp32_len = 11140;
extern unsigned char matmul_id_q4_0_f32_aligned_fp32_data[11096];
const uint64_t matmul_id_q4_0_f32_aligned_fp32_len = 11096;
extern unsigned char matmul_id_q4_1_f32_aligned_fp32_data[11164];
const uint64_t matmul_id_q4_1_f32_aligned_fp32_len = 11164;
extern unsigned char matmul_id_q4_1_f32_fp32_data[10984];
const uint64_t matmul_id_q4_1_f32_fp32_len = 10984;
extern unsigned char matmul_id_q5_0_f32_fp32_data[11412];
const uint64_t matmul_id_q5_0_f32_fp32_len = 11412;
extern unsigned char matmul_f32_f16_aligned_data[11360];
const uint64_t matmul_f32_f16_aligned_len = 11360;
extern unsigned char matmul_id_f16_aligned_fp32_data[10976];
const uint64_t matmul_id_f16_aligned_fp32_len = 10976;
extern unsigned char matmul_id_q5_1_f32_fp32_data[11308];
const uint64_t matmul_id_q5_1_f32_fp32_len = 11308;
extern unsigned char matmul_f16_aligned_data[11200];
const uint64_t matmul_f16_aligned_len = 11200;
extern unsigned char matmul_id_f16_f32_aligned_fp32_data[10928];
const uint64_t matmul_id_f16_f32_aligned_fp32_len = 10928;
extern unsigned char matmul_id_q8_0_f32_fp32_data[10940];
const uint64_t matmul_id_q8_0_f32_fp32_len = 10940;
extern unsigned char matmul_id_q4_0_f32_aligned_fp32_data[11096];
const uint64_t matmul_id_q4_0_f32_aligned_fp32_len = 11096;
extern unsigned char matmul_id_q4_0_f32_fp32_data[10916];
const uint64_t matmul_id_q4_0_f32_fp32_len = 10916;
extern unsigned char matmul_id_q2_k_f32_aligned_fp32_data[11756];
const uint64_t matmul_id_q2_k_f32_aligned_fp32_len = 11756;
extern unsigned char matmul_id_q4_k_f32_fp32_data[12404];
const uint64_t matmul_id_q4_k_f32_fp32_len = 12404;
extern unsigned char matmul_id_q3_k_f32_aligned_fp32_data[13660];
const uint64_t matmul_id_q3_k_f32_aligned_fp32_len = 13660;
extern unsigned char matmul_id_q4_k_f32_aligned_fp32_data[12588];
const uint64_t matmul_id_q4_k_f32_aligned_fp32_len = 12588;
extern unsigned char matmul_id_q3_k_f32_fp32_data[13476];
const uint64_t matmul_id_q3_k_f32_fp32_len = 13476;
extern unsigned char matmul_id_q5_k_f32_fp32_data[12928];
const uint64_t matmul_id_q5_k_f32_fp32_len = 12928;
extern unsigned char matmul_id_q2_k_f32_fp32_data[11572];
const uint64_t matmul_id_q2_k_f32_fp32_len = 11572;
extern unsigned char matmul_id_q6_k_f32_fp32_data[12096];
const uint64_t matmul_id_q6_k_f32_fp32_len = 12096;
extern unsigned char matmul_id_q6_k_f32_aligned_fp32_data[12280];
const uint64_t matmul_id_q6_k_f32_aligned_fp32_len = 12280;
extern unsigned char matmul_id_q5_k_f32_aligned_fp32_data[13112];
const uint64_t matmul_id_q5_k_f32_aligned_fp32_len = 13112;
extern unsigned char matmul_id_q4_1_f32_fp32_data[10984];
const uint64_t matmul_id_q4_1_f32_fp32_len = 10984;
extern unsigned char matmul_id_q5_0_f32_aligned_fp32_data[11592];
const uint64_t matmul_id_q5_0_f32_aligned_fp32_len = 11592;
extern unsigned char matmul_f32_f16_data[10332];
const uint64_t matmul_f32_f16_len = 10332;
extern unsigned char matmul_id_q5_0_f32_fp32_data[11412];
const uint64_t matmul_id_q5_0_f32_fp32_len = 11412;
extern unsigned char matmul_q5_1_f32_aligned_fp32_data[10800];
const uint64_t matmul_q5_1_f32_aligned_fp32_len = 10800;
extern unsigned char matmul_id_q5_1_f32_aligned_fp32_data[11488];
const uint64_t matmul_id_q5_1_f32_aligned_fp32_len = 11488;
extern unsigned char matmul_f16_f32_data[10332];
const uint64_t matmul_f16_f32_len = 10332;
extern unsigned char matmul_id_q4_1_f32_aligned_fp32_data[11164];
const uint64_t matmul_id_q4_1_f32_aligned_fp32_len = 11164;
extern unsigned char matmul_f32_f32_data[10324];
const uint64_t matmul_f32_f32_len = 10324;
extern unsigned char matmul_id_q8_0_f32_fp32_data[10960];
const uint64_t matmul_id_q8_0_f32_fp32_len = 10960;
extern unsigned char matmul_id_q8_0_f32_aligned_fp32_data[11140];
const uint64_t matmul_id_q8_0_f32_aligned_fp32_len = 11140;
extern unsigned char matmul_f32_f16_data[10332];
const uint64_t matmul_f32_f16_len = 10332;
extern unsigned char matmul_f32_f16_aligned_data[11360];
const uint64_t matmul_f32_f16_aligned_len = 11360;
extern unsigned char matmul_id_q3_k_f32_aligned_fp32_data[13660];
const uint64_t matmul_id_q3_k_f32_aligned_fp32_len = 13660;
extern unsigned char matmul_id_q2_k_f32_aligned_fp32_data[11756];
const uint64_t matmul_id_q2_k_f32_aligned_fp32_len = 11756;
extern unsigned char matmul_id_q4_k_f32_fp32_data[12424];
const uint64_t matmul_id_q4_k_f32_fp32_len = 12424;
extern unsigned char matmul_f16_f32_data[10864];
const uint64_t matmul_f16_f32_len = 10864;
extern unsigned char matmul_id_q5_k_f32_fp32_data[12948];
const uint64_t matmul_id_q5_k_f32_fp32_len = 12948;
extern unsigned char matmul_id_q5_k_f32_aligned_fp32_data[13112];
const uint64_t matmul_id_q5_k_f32_aligned_fp32_len = 13112;
extern unsigned char matmul_id_q6_k_f32_fp32_data[12116];
const uint64_t matmul_id_q6_k_f32_fp32_len = 12116;
extern unsigned char matmul_id_q6_k_f32_aligned_fp32_data[12280];
const uint64_t matmul_id_q6_k_f32_aligned_fp32_len = 12280;
extern unsigned char matmul_q4_1_f32_aligned_data[11224];
const uint64_t matmul_q4_1_f32_aligned_len = 11224;
extern unsigned char matmul_f16_data[10316];
const uint64_t matmul_f16_len = 10316;
extern unsigned char matmul_q4_0_f32_aligned_data[11156];
const uint64_t matmul_q4_0_f32_aligned_len = 11156;
extern unsigned char matmul_q8_0_f32_data[10608];
const uint64_t matmul_q8_0_f32_len = 10608;
extern unsigned char matmul_q2_k_f32_data[11240];
const uint64_t matmul_q2_k_f32_len = 11240;
extern unsigned char matmul_q5_0_f32_aligned_data[11668];
const uint64_t matmul_q5_0_f32_aligned_len = 11668;
extern unsigned char matmul_f16_f32_aligned_data[11360];
const uint64_t matmul_f16_f32_aligned_len = 11360;
@ -183,14 +189,20 @@ const uint64_t matmul_q4_0_f32_len = 10564;
extern unsigned char matmul_q4_1_f32_data[10632];
const uint64_t matmul_q4_1_f32_len = 10632;
extern unsigned char matmul_q6_k_f32_aligned_data[12324];
const uint64_t matmul_q6_k_f32_aligned_len = 12324;
extern unsigned char matmul_q5_0_f32_data[11076];
const uint64_t matmul_q5_0_f32_len = 11076;
extern unsigned char matmul_q4_1_f32_aligned_data[11224];
const uint64_t matmul_q4_1_f32_aligned_len = 11224;
extern unsigned char matmul_f32_f32_data[10968];
const uint64_t matmul_f32_f32_len = 10968;
extern unsigned char matmul_id_q3_k_f32_fp32_data[13496];
const uint64_t matmul_id_q3_k_f32_fp32_len = 13496;
extern unsigned char matmul_id_q2_k_f32_fp32_data[11592];
const uint64_t matmul_id_q2_k_f32_fp32_len = 11592;
extern unsigned char matmul_f32_f32_aligned_data[11432];
const uint64_t matmul_f32_f32_aligned_len = 11432;
extern unsigned char matmul_q5_1_f32_data[10956];
const uint64_t matmul_q5_1_f32_len = 10956;
@ -198,149 +210,131 @@ const uint64_t matmul_q5_1_f32_len = 10956;
extern unsigned char matmul_q5_1_f32_aligned_data[11548];
const uint64_t matmul_q5_1_f32_aligned_len = 11548;
extern unsigned char matmul_id_f16_aligned_data[11888];
const uint64_t matmul_id_f16_aligned_len = 11888;
extern unsigned char matmul_id_f32_f32_data[10796];
const uint64_t matmul_id_f32_f32_len = 10796;
extern unsigned char matmul_id_f16_f32_data[10804];
const uint64_t matmul_id_f16_f32_len = 10804;
extern unsigned char matmul_q8_0_f32_data[10588];
const uint64_t matmul_q8_0_f32_len = 10588;
extern unsigned char matmul_q5_k_f32_data[12576];
const uint64_t matmul_q5_k_f32_len = 12576;
extern unsigned char matmul_q3_k_f32_data[13124];
const uint64_t matmul_q3_k_f32_len = 13124;
extern unsigned char matmul_q4_k_f32_data[12052];
const uint64_t matmul_q4_k_f32_len = 12052;
extern unsigned char matmul_q5_k_f32_aligned_data[13172];
const uint64_t matmul_q5_k_f32_aligned_len = 13172;
extern unsigned char matmul_q6_k_f32_data[11744];
const uint64_t matmul_q6_k_f32_len = 11744;
extern unsigned char matmul_q3_k_f32_aligned_data[13704];
const uint64_t matmul_q3_k_f32_aligned_len = 13704;
extern unsigned char matmul_id_f32_f32_aligned_data[12120];
const uint64_t matmul_id_f32_f32_aligned_len = 12120;
extern unsigned char matmul_id_f32_f16_aligned_data[12048];
const uint64_t matmul_id_f32_f16_aligned_len = 12048;
extern unsigned char matmul_id_f16_data[10788];
const uint64_t matmul_id_f16_len = 10788;
extern unsigned char matmul_q4_k_f32_aligned_data[12648];
const uint64_t matmul_q4_k_f32_aligned_len = 12648;
extern unsigned char matmul_f32_f32_aligned_data[11432];
const uint64_t matmul_f32_f32_aligned_len = 11432;
extern unsigned char matmul_q2_k_f32_fp32_data[11100];
const uint64_t matmul_q2_k_f32_fp32_len = 11100;
extern unsigned char matmul_q2_k_f32_data[11220];
const uint64_t matmul_q2_k_f32_len = 11220;
extern unsigned char matmul_id_q8_0_f32_aligned_data[11888];
const uint64_t matmul_id_q8_0_f32_aligned_len = 11888;
extern unsigned char matmul_q2_k_f32_aligned_data[11800];
const uint64_t matmul_q2_k_f32_aligned_len = 11800;
extern unsigned char matmul_id_q4_0_f32_data[11036];
const uint64_t matmul_id_q4_0_f32_len = 11036;
extern unsigned char matmul_q5_0_f32_aligned_data[11668];
const uint64_t matmul_q5_0_f32_aligned_len = 11668;
extern unsigned char matmul_id_q5_1_f32_data[11428];
const uint64_t matmul_id_q5_1_f32_len = 11428;
extern unsigned char matmul_id_q4_0_f32_aligned_data[11844];
const uint64_t matmul_id_q4_0_f32_aligned_len = 11844;
extern unsigned char matmul_f16_aligned_data[11200];
const uint64_t matmul_f16_aligned_len = 11200;
extern unsigned char matmul_id_f32_f16_data[10804];
const uint64_t matmul_id_f32_f16_len = 10804;
extern unsigned char matmul_id_q3_k_f32_data[13596];
const uint64_t matmul_id_q3_k_f32_len = 13596;
extern unsigned char matmul_id_q4_k_f32_aligned_fp32_data[12588];
const uint64_t matmul_id_q4_k_f32_aligned_fp32_len = 12588;
extern unsigned char matmul_id_q3_k_f32_aligned_data[14392];
const uint64_t matmul_id_q3_k_f32_aligned_len = 14392;
extern unsigned char matmul_id_f16_aligned_data[11888];
const uint64_t matmul_id_f16_aligned_len = 11888;
extern unsigned char matmul_id_q2_k_f32_data[11692];
const uint64_t matmul_id_q2_k_f32_len = 11692;
extern unsigned char matmul_id_f16_f32_data[11336];
const uint64_t matmul_id_f16_f32_len = 11336;
extern unsigned char matmul_q5_k_f32_data[12596];
const uint64_t matmul_q5_k_f32_len = 12596;
extern unsigned char matmul_q4_k_f32_aligned_data[12648];
const uint64_t matmul_q4_k_f32_aligned_len = 12648;
extern unsigned char matmul_id_q4_0_f32_aligned_data[11844];
const uint64_t matmul_id_q4_0_f32_aligned_len = 11844;
extern unsigned char matmul_q6_k_f32_data[11764];
const uint64_t matmul_q6_k_f32_len = 11764;
extern unsigned char matmul_q6_k_f32_aligned_data[12324];
const uint64_t matmul_q6_k_f32_aligned_len = 12324;
extern unsigned char matmul_id_q5_0_f32_data[11532];
const uint64_t matmul_id_q5_0_f32_len = 11532;
extern unsigned char matmul_q8_0_f32_aligned_data[11200];
const uint64_t matmul_q8_0_f32_aligned_len = 11200;
extern unsigned char matmul_id_q5_k_f32_aligned_data[13860];
const uint64_t matmul_id_q5_k_f32_aligned_len = 13860;
extern unsigned char matmul_id_f16_data[10788];
const uint64_t matmul_id_f16_len = 10788;
extern unsigned char matmul_id_q5_1_f32_aligned_data[12236];
const uint64_t matmul_id_q5_1_f32_aligned_len = 12236;
extern unsigned char matmul_q2_k_f32_aligned_data[11800];
const uint64_t matmul_q2_k_f32_aligned_len = 11800;
extern unsigned char matmul_q3_k_f32_data[13144];
const uint64_t matmul_q3_k_f32_len = 13144;
extern unsigned char matmul_q3_k_f32_aligned_data[13704];
const uint64_t matmul_q3_k_f32_aligned_len = 13704;
extern unsigned char matmul_id_q4_k_f32_data[12544];
const uint64_t matmul_id_q4_k_f32_len = 12544;
extern unsigned char matmul_id_q5_0_f32_aligned_data[12340];
const uint64_t matmul_id_q5_0_f32_aligned_len = 12340;
extern unsigned char matmul_id_q4_1_f32_aligned_data[11912];
const uint64_t matmul_id_q4_1_f32_aligned_len = 11912;
extern unsigned char matmul_id_q5_1_f32_data[11428];
const uint64_t matmul_id_q5_1_f32_len = 11428;
extern unsigned char matmul_q4_k_f32_data[12072];
const uint64_t matmul_q4_k_f32_len = 12072;
extern unsigned char matmul_id_q4_1_f32_data[11104];
const uint64_t matmul_id_q4_1_f32_len = 11104;
extern unsigned char matmul_q5_k_f32_aligned_data[13172];
const uint64_t matmul_q5_k_f32_aligned_len = 13172;
extern unsigned char matmul_id_q8_0_f32_data[11080];
const uint64_t matmul_id_q8_0_f32_len = 11080;
extern unsigned char matmul_id_f32_f16_aligned_data[12048];
const uint64_t matmul_id_f32_f16_aligned_len = 12048;
extern unsigned char matmul_id_q2_k_f32_data[11712];
const uint64_t matmul_id_q2_k_f32_len = 11712;
extern unsigned char matmul_id_q8_0_f32_aligned_data[11888];
const uint64_t matmul_id_q8_0_f32_aligned_len = 11888;
extern unsigned char matmul_id_q3_k_f32_data[13616];
const uint64_t matmul_id_q3_k_f32_len = 13616;
extern unsigned char matmul_id_f16_f32_aligned_data[12048];
const uint64_t matmul_id_f16_f32_aligned_len = 12048;
extern unsigned char matmul_id_q4_0_f32_data[11036];
const uint64_t matmul_id_q4_0_f32_len = 11036;
extern unsigned char matmul_id_q4_k_f32_aligned_data[13336];
const uint64_t matmul_id_q4_k_f32_aligned_len = 13336;
extern unsigned char matmul_id_q5_k_f32_data[13048];
const uint64_t matmul_id_q5_k_f32_len = 13048;
extern unsigned char matmul_id_q8_0_f32_data[11060];
const uint64_t matmul_id_q8_0_f32_len = 11060;
extern unsigned char matmul_id_q6_k_f32_data[12216];
const uint64_t matmul_id_q6_k_f32_len = 12216;
extern unsigned char matmul_id_q4_k_f32_data[12524];
const uint64_t matmul_id_q4_k_f32_len = 12524;
extern unsigned char matmul_id_q6_k_f32_aligned_data[13012];
const uint64_t matmul_id_q6_k_f32_aligned_len = 13012;
extern unsigned char mul_mat_vec_f32_f32_f32_data[3616];
const uint64_t mul_mat_vec_f32_f32_f32_len = 3616;
extern unsigned char mul_mat_vec_f32_f16_f32_data[3684];
const uint64_t mul_mat_vec_f32_f16_f32_len = 3684;
extern unsigned char get_rows_f32_data[3028];
const uint64_t get_rows_f32_len = 3028;
extern unsigned char get_rows_f32_f32_data[2976];
const uint64_t get_rows_f32_f32_len = 2976;
extern unsigned char mul_mat_vec_id_f32_f32_data[3540];
const uint64_t mul_mat_vec_id_f32_f32_len = 3540;
extern unsigned char mul_mat_vec_f16_f16_f32_data[3716];
const uint64_t mul_mat_vec_f16_f16_f32_len = 3716;
extern unsigned char dequant_f32_data[3200];
const uint64_t dequant_f32_len = 3200;
extern unsigned char mul_mat_vec_f16_f32_f32_data[3684];
const uint64_t mul_mat_vec_f16_f32_f32_len = 3684;
extern unsigned char mul_mat_vec_id_f16_f32_data[3608];
const uint64_t mul_mat_vec_id_f16_f32_len = 3608;
extern unsigned char matmul_id_q6_k_f32_data[12236];
const uint64_t matmul_id_q6_k_f32_len = 12236;
extern unsigned char get_rows_f16_data[2996];
const uint64_t get_rows_f16_len = 2996;
extern unsigned char mul_mat_vec_f32_f32_f32_data[3616];
const uint64_t mul_mat_vec_f32_f32_f32_len = 3616;
extern unsigned char get_rows_f16_f32_data[3028];
const uint64_t get_rows_f16_f32_len = 3028;
extern unsigned char matmul_id_q6_k_f32_aligned_data[13012];
const uint64_t matmul_id_q6_k_f32_aligned_len = 13012;
extern unsigned char mul_mat_vec_q4_0_f32_f32_data[4076];
const uint64_t mul_mat_vec_q4_0_f32_f32_len = 4076;
extern unsigned char mul_mat_vec_id_f32_f32_data[3540];
const uint64_t mul_mat_vec_id_f32_f32_len = 3540;
extern unsigned char mul_mat_vec_q4_0_f16_f32_data[4108];
const uint64_t mul_mat_vec_q4_0_f16_f32_len = 4108;
extern unsigned char matmul_id_q2_k_f32_aligned_data[12488];
const uint64_t matmul_id_q2_k_f32_aligned_len = 12488;
extern unsigned char dequant_f32_data[3200];
const uint64_t dequant_f32_len = 3200;
extern unsigned char get_rows_f32_data[3028];
const uint64_t get_rows_f32_len = 3028;
extern unsigned char matmul_id_q5_k_f32_data[13068];
const uint64_t matmul_id_q5_k_f32_len = 13068;
extern unsigned char mul_mat_vec_id_q4_0_f32_data[4000];
const uint64_t mul_mat_vec_id_q4_0_f32_len = 4000;
@ -348,38 +342,47 @@ const uint64_t mul_mat_vec_id_q4_0_f32_len = 4000;
extern unsigned char dequant_q4_0_data[5184];
const uint64_t dequant_q4_0_len = 5184;
extern unsigned char get_rows_q4_0_data[3704];
const uint64_t get_rows_q4_0_len = 3704;
extern unsigned char mul_mat_vec_id_f16_f32_data[3608];
const uint64_t mul_mat_vec_id_f16_f32_len = 3608;
extern unsigned char get_rows_f16_data[2996];
const uint64_t get_rows_f16_len = 2996;
extern unsigned char mul_mat_vec_q4_1_f32_f32_data[4160];
const uint64_t mul_mat_vec_q4_1_f32_f32_len = 4160;
extern unsigned char get_rows_q4_0_f32_data[3688];
const uint64_t get_rows_q4_0_f32_len = 3688;
extern unsigned char mul_mat_vec_q4_0_f32_f32_data[4076];
const uint64_t mul_mat_vec_q4_0_f32_f32_len = 4076;
extern unsigned char get_rows_q4_1_data[3788];
const uint64_t get_rows_q4_1_len = 3788;
extern unsigned char mul_mat_vec_q4_1_f16_f32_data[4192];
const uint64_t mul_mat_vec_q4_1_f16_f32_len = 4192;
extern unsigned char dequant_q4_1_data[5248];
const uint64_t dequant_q4_1_len = 5248;
extern unsigned char get_rows_q4_1_f32_data[3772];
const uint64_t get_rows_q4_1_f32_len = 3772;
extern unsigned char matmul_id_q5_0_f32_aligned_data[12340];
const uint64_t matmul_id_q5_0_f32_aligned_len = 12340;
extern unsigned char mul_mat_vec_q4_0_f16_f32_data[4108];
const uint64_t mul_mat_vec_q4_0_f16_f32_len = 4108;
extern unsigned char mul_mat_vec_id_q4_1_f32_data[4068];
const uint64_t mul_mat_vec_id_q4_1_f32_len = 4068;
extern unsigned char matmul_id_q5_1_f32_aligned_data[12236];
const uint64_t matmul_id_q5_1_f32_aligned_len = 12236;
extern unsigned char get_rows_q4_0_data[3704];
const uint64_t get_rows_q4_0_len = 3704;
extern unsigned char dequant_q5_1_data[6412];
const uint64_t dequant_q5_1_len = 6412;
extern unsigned char mul_mat_vec_f16_f16_f32_data[3716];
const uint64_t mul_mat_vec_f16_f16_f32_len = 3716;
extern unsigned char get_rows_q4_0_f32_data[3688];
const uint64_t get_rows_q4_0_f32_len = 3688;
extern unsigned char get_rows_f16_f32_data[3028];
const uint64_t get_rows_f16_f32_len = 3028;
extern unsigned char mul_mat_vec_q4_1_f16_f32_data[4192];
const uint64_t mul_mat_vec_q4_1_f16_f32_len = 4192;
extern unsigned char matmul_id_q5_k_f32_aligned_data[13860];
const uint64_t matmul_id_q5_k_f32_aligned_len = 13860;
extern unsigned char mul_mat_vec_q5_0_f16_f32_data[4636];
const uint64_t mul_mat_vec_q5_0_f16_f32_len = 4636;
extern unsigned char get_rows_q4_1_f32_data[3772];
const uint64_t get_rows_q4_1_f32_len = 3772;
extern unsigned char mul_mat_vec_q5_0_f32_f32_data[4604];
const uint64_t mul_mat_vec_q5_0_f32_f32_len = 4604;
@ -387,29 +390,53 @@ const uint64_t mul_mat_vec_q5_0_f32_f32_len = 4604;
extern unsigned char get_rows_q5_0_data[4232];
const uint64_t get_rows_q5_0_len = 4232;
extern unsigned char get_rows_q5_1_f32_data[4112];
const uint64_t get_rows_q5_1_f32_len = 4112;
extern unsigned char mul_mat_vec_q5_1_f32_f32_data[4500];
const uint64_t mul_mat_vec_q5_1_f32_f32_len = 4500;
extern unsigned char dequant_q5_0_data[6644];
const uint64_t dequant_q5_0_len = 6644;
extern unsigned char mul_mat_vec_id_q5_1_f32_data[4392];
const uint64_t mul_mat_vec_id_q5_1_f32_len = 4392;
extern unsigned char get_rows_q5_0_f32_data[4216];
const uint64_t get_rows_q5_0_f32_len = 4216;
extern unsigned char get_rows_q4_1_data[3788];
const uint64_t get_rows_q4_1_len = 3788;
extern unsigned char mul_mat_vec_q5_1_f16_f32_data[4532];
const uint64_t mul_mat_vec_q5_1_f16_f32_len = 4532;
extern unsigned char get_rows_q5_1_data[4128];
const uint64_t get_rows_q5_1_len = 4128;
extern unsigned char dequant_q5_1_data[6412];
const uint64_t dequant_q5_1_len = 6412;
extern unsigned char dequant_q4_1_data[5248];
const uint64_t dequant_q4_1_len = 5248;
extern unsigned char get_rows_q5_1_f32_data[4112];
const uint64_t get_rows_q5_1_f32_len = 4112;
extern unsigned char mul_mat_vec_q8_0_f16_f32_data[4064];
const uint64_t mul_mat_vec_q8_0_f16_f32_len = 4064;
extern unsigned char mul_mat_vec_q8_0_f32_f32_data[4032];
const uint64_t mul_mat_vec_q8_0_f32_f32_len = 4032;
extern unsigned char get_rows_q8_0_data[3644];
const uint64_t get_rows_q8_0_len = 3644;
extern unsigned char mul_mat_vec_id_q8_0_f32_data[3956];
const uint64_t mul_mat_vec_id_q8_0_f32_len = 3956;
extern unsigned char dequant_q8_0_data[4780];
const uint64_t dequant_q8_0_len = 4780;
extern unsigned char get_rows_q8_0_f32_data[3628];
const uint64_t get_rows_q8_0_f32_len = 3628;
extern unsigned char get_rows_q8_0_data[3644];
const uint64_t get_rows_q8_0_len = 3644;
extern unsigned char mul_mat_vec_q8_0_f16_f32_data[4064];
const uint64_t mul_mat_vec_q8_0_f16_f32_len = 4064;
extern unsigned char mul_mat_vec_q4_k_f16_f32_data[10548];
const uint64_t mul_mat_vec_q4_k_f16_f32_len = 10548;
extern unsigned char mul_mat_vec_id_q2_k_f32_data[8412];
const uint64_t mul_mat_vec_id_q2_k_f32_len = 8412;
@ -417,23 +444,14 @@ const uint64_t mul_mat_vec_id_q2_k_f32_len = 8412;
extern unsigned char mul_mat_vec_q2_k_f16_f32_data[8760];
const uint64_t mul_mat_vec_q2_k_f16_f32_len = 8760;
extern unsigned char mul_mat_vec_q2_k_f32_f32_data[8504];
const uint64_t mul_mat_vec_q2_k_f32_f32_len = 8504;
extern unsigned char mul_mat_vec_q3_k_f32_f32_data[10048];
const uint64_t mul_mat_vec_q3_k_f32_f32_len = 10048;
extern unsigned char dequant_q2_k_data[4032];
const uint64_t dequant_q2_k_len = 4032;
extern unsigned char mul_mat_vec_q3_k_f32_f32_data[10048];
const uint64_t mul_mat_vec_q3_k_f32_f32_len = 10048;
extern unsigned char dequant_q4_k_data[5916];
const uint64_t dequant_q4_k_len = 5916;
extern unsigned char mul_mat_vec_id_q3_k_f32_data[10020];
const uint64_t mul_mat_vec_id_q3_k_f32_len = 10020;
extern unsigned char mul_mat_vec_q6_k_f16_f32_data[12108];
const uint64_t mul_mat_vec_q6_k_f16_f32_len = 12108;
extern unsigned char dequant_q6_k_data[4272];
const uint64_t dequant_q6_k_len = 4272;
extern unsigned char mul_mat_vec_q3_k_f16_f32_data[10176];
const uint64_t mul_mat_vec_q3_k_f16_f32_len = 10176;
@ -444,62 +462,59 @@ const uint64_t dequant_q3_k_len = 4804;
extern unsigned char mul_mat_vec_q4_k_f32_f32_data[10036];
const uint64_t mul_mat_vec_q4_k_f32_f32_len = 10036;
extern unsigned char mul_mat_vec_q4_k_f16_f32_data[10548];
const uint64_t mul_mat_vec_q4_k_f16_f32_len = 10548;
extern unsigned char get_rows_q8_0_f32_data[3628];
const uint64_t get_rows_q8_0_f32_len = 3628;
extern unsigned char mul_mat_vec_id_q4_k_f32_data[9944];
const uint64_t mul_mat_vec_id_q4_k_f32_len = 9944;
extern unsigned char mul_mat_vec_id_q5_k_f32_data[12816];
const uint64_t mul_mat_vec_id_q5_k_f32_len = 12816;
extern unsigned char mul_mat_vec_q5_k_f16_f32_data[13436];
const uint64_t mul_mat_vec_q5_k_f16_f32_len = 13436;
extern unsigned char mul_mat_vec_q6_k_f32_f32_data[11852];
const uint64_t mul_mat_vec_q6_k_f32_f32_len = 11852;
extern unsigned char dequant_q4_k_data[5916];
const uint64_t dequant_q4_k_len = 5916;
extern unsigned char mul_mat_vec_q5_k_f32_f32_data[12924];
const uint64_t mul_mat_vec_q5_k_f32_f32_len = 12924;
extern unsigned char dequant_q5_k_data[5980];
const uint64_t dequant_q5_k_len = 5980;
extern unsigned char sqr_f32_data[3252];
const uint64_t sqr_f32_len = 3252;
extern unsigned char mul_mat_vec_q6_k_f32_f32_data[11852];
const uint64_t mul_mat_vec_q6_k_f32_f32_len = 11852;
extern unsigned char mul_mat_vec_q6_k_f16_f32_data[12108];
const uint64_t mul_mat_vec_q6_k_f16_f32_len = 12108;
extern unsigned char mul_mat_vec_id_q6_k_f32_data[11760];
const uint64_t mul_mat_vec_id_q6_k_f32_len = 11760;
extern unsigned char matmul_id_q5_0_f32_data[11532];
const uint64_t matmul_id_q5_0_f32_len = 11532;
extern unsigned char mul_mat_vec_id_q3_k_f32_data[10020];
const uint64_t mul_mat_vec_id_q3_k_f32_len = 10020;
extern unsigned char cpy_f32_f32_data[3232];
const uint64_t cpy_f32_f32_len = 3232;
extern unsigned char relu_f32_data[1212];
const uint64_t relu_f32_len = 1212;
extern unsigned char mul_mat_vec_p021_f16_f32_data[2768];
const uint64_t mul_mat_vec_p021_f16_f32_len = 2768;
extern unsigned char norm_f32_data[2624];
const uint64_t norm_f32_len = 2624;
extern unsigned char mul_mat_vec_nc_f16_f32_data[2824];
const uint64_t mul_mat_vec_nc_f16_f32_len = 2824;
extern unsigned char rope_neox_f32_data[3852];
const uint64_t rope_neox_f32_len = 3852;
extern unsigned char rms_norm_f32_data[2344];
const uint64_t rms_norm_f32_len = 2344;
extern unsigned char cpy_f32_f16_data[3284];
const uint64_t cpy_f32_f16_len = 3284;
extern unsigned char rope_neox_f16_data[3952];
const uint64_t rope_neox_f16_len = 3952;
extern unsigned char cpy_f16_f16_data[3252];
const uint64_t cpy_f16_f16_len = 3252;
extern unsigned char dequant_q6_k_data[4272];
const uint64_t dequant_q6_k_len = 4272;
extern unsigned char split_k_reduce_data[1416];
const uint64_t split_k_reduce_len = 1416;
extern unsigned char add_f32_data[4276];
const uint64_t add_f32_len = 4276;
extern unsigned char split_k_reduce_data[1416];
const uint64_t split_k_reduce_len = 1416;
extern unsigned char norm_f32_data[2624];
const uint64_t norm_f32_len = 2624;
extern unsigned char mul_f32_data[4276];
const uint64_t mul_f32_len = 4276;
extern unsigned char div_f32_data[4276];
const uint64_t div_f32_len = 4276;
@ -507,93 +522,78 @@ const uint64_t div_f32_len = 4276;
extern unsigned char scale_f32_data[3320];
const uint64_t scale_f32_len = 3320;
extern unsigned char gelu_f32_data[1484];
const uint64_t gelu_f32_len = 1484;
extern unsigned char mul_f32_data[4276];
const uint64_t mul_f32_len = 4276;
extern unsigned char clamp_f32_data[3512];
const uint64_t clamp_f32_len = 3512;
extern unsigned char diag_mask_inf_f32_data[1504];
const uint64_t diag_mask_inf_f32_len = 1504;
extern unsigned char gelu_f32_data[1484];
const uint64_t gelu_f32_len = 1484;
extern unsigned char mul_mat_vec_id_q5_k_f32_data[12816];
const uint64_t mul_mat_vec_id_q5_k_f32_len = 12816;
extern unsigned char silu_f32_data[1264];
const uint64_t silu_f32_len = 1264;
extern unsigned char relu_f32_data[1212];
const uint64_t relu_f32_len = 1212;
extern unsigned char diag_mask_inf_f32_data[1504];
const uint64_t diag_mask_inf_f32_len = 1504;
extern unsigned char soft_max_f32_data[4836];
const uint64_t soft_max_f32_len = 4836;
extern unsigned char sqr_f32_data[3252];
const uint64_t sqr_f32_len = 3252;
extern unsigned char rope_norm_f32_data[3852];
const uint64_t rope_norm_f32_len = 3852;
extern unsigned char rope_norm_f16_data[3952];
const uint64_t rope_norm_f16_len = 3952;
extern unsigned char mul_mat_vec_p021_f16_f32_data[2768];
const uint64_t mul_mat_vec_p021_f16_f32_len = 2768;
extern unsigned char soft_max_f32_f16_data[4904];
const uint64_t soft_max_f32_f16_len = 4904;
extern unsigned char rope_neox_f32_data[3852];
const uint64_t rope_neox_f32_len = 3852;
extern unsigned char rope_norm_f16_data[3952];
const uint64_t rope_norm_f16_len = 3952;
extern unsigned char rope_neox_f16_data[3952];
const uint64_t rope_neox_f16_len = 3952;
extern unsigned char cpy_f32_f32_data[3232];
const uint64_t cpy_f32_f32_len = 3232;
extern unsigned char dequant_q5_0_data[6644];
const uint64_t dequant_q5_0_len = 6644;
extern unsigned char sum_rows_f32_data[2112];
const uint64_t sum_rows_f32_len = 2112;
extern unsigned char get_rows_q5_0_f32_data[4216];
const uint64_t get_rows_q5_0_f32_len = 4216;
extern unsigned char get_rows_q5_1_data[4128];
const uint64_t get_rows_q5_1_len = 4128;
extern unsigned char matmul_id_f16_f32_aligned_data[12048];
const uint64_t matmul_id_f16_f32_aligned_len = 12048;
extern unsigned char matmul_id_q2_k_f32_aligned_data[12488];
const uint64_t matmul_id_q2_k_f32_aligned_len = 12488;
extern unsigned char matmul_q4_k_f32_fp32_data[11932];
const uint64_t matmul_q4_k_f32_fp32_len = 11932;
extern unsigned char matmul_f32_f32_aligned_fp32_data[10124];
const uint64_t matmul_f32_f32_aligned_fp32_len = 10124;
extern unsigned char matmul_q4_k_f32_aligned_fp32_data[11900];
const uint64_t matmul_q4_k_f32_aligned_fp32_len = 11900;
extern unsigned char matmul_id_q4_1_f32_data[11104];
const uint64_t matmul_id_q4_1_f32_len = 11104;
extern unsigned char matmul_q5_1_f32_aligned_fp32_data[10800];
const uint64_t matmul_q5_1_f32_aligned_fp32_len = 10800;
extern unsigned char matmul_id_q4_1_f32_aligned_data[11912];
const uint64_t matmul_id_q4_1_f32_aligned_len = 11912;
extern unsigned char mul_mat_vec_q5_0_f16_f32_data[4636];
const uint64_t mul_mat_vec_q5_0_f16_f32_len = 4636;
extern unsigned char mul_mat_vec_id_q5_0_f32_data[4512];
const uint64_t mul_mat_vec_id_q5_0_f32_len = 4512;
extern unsigned char mul_mat_vec_q5_1_f16_f32_data[4532];
const uint64_t mul_mat_vec_q5_1_f16_f32_len = 4532;
extern unsigned char mul_mat_vec_id_q5_1_f32_data[4392];
const uint64_t mul_mat_vec_id_q5_1_f32_len = 4392;
extern unsigned char rope_norm_f32_data[3852];
const uint64_t rope_norm_f32_len = 3852;
extern unsigned char argsort_f32_data[4096];
const uint64_t argsort_f32_len = 4096;
extern unsigned char cpy_f32_f16_data[3284];
const uint64_t cpy_f32_f16_len = 3284;
extern unsigned char sum_rows_f32_data[2112];
const uint64_t sum_rows_f32_len = 2112;
extern unsigned char mul_mat_vec_q5_k_f16_f32_data[13436];
const uint64_t mul_mat_vec_q5_k_f16_f32_len = 13436;
extern unsigned char matmul_id_f32_f32_data[11440];
const uint64_t matmul_id_f32_f32_len = 11440;
extern unsigned char matmul_q3_k_f32_fp32_data[13024];
const uint64_t matmul_q3_k_f32_fp32_len = 13024;
extern unsigned char matmul_id_q3_k_f32_aligned_data[14392];
const uint64_t matmul_id_q3_k_f32_aligned_len = 14392;
extern unsigned char mul_mat_vec_id_q5_0_f32_data[4512];
const uint64_t mul_mat_vec_id_q5_0_f32_len = 4512;
extern unsigned char matmul_q4_0_f32_aligned_data[11156];
const uint64_t matmul_q4_0_f32_aligned_len = 11156;
extern unsigned char mul_mat_vec_f32_f16_f32_data[3684];
const uint64_t mul_mat_vec_f32_f16_f32_len = 3684;
extern unsigned char matmul_id_f32_f32_aligned_data[12120];
const uint64_t matmul_id_f32_f32_aligned_len = 12120;
extern unsigned char dequant_q5_k_data[5980];
const uint64_t dequant_q5_k_len = 5980;
extern unsigned char mul_mat_vec_nc_f16_f32_data[2824];
const uint64_t mul_mat_vec_nc_f16_f32_len = 2824;
extern unsigned char mul_mat_vec_q2_k_f32_f32_data[8504];
const uint64_t mul_mat_vec_q2_k_f32_f32_len = 8504;