upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-08-01 15:40:46 -05:00
Code Issues Packages Projects Releases Wiki Activity
10,222 Commits 643 Branches 6,961 Tags

Default Branch

master

a7a6d0d269 · vulkan: extend topk_moe fusion to support sqrt(softplus) (#26124) · Updated 2026-08-01 14:18:07 -05:00
 ZIP  TAR.GZ

Branches

try-fix-metal

5cc672a9a5 · metal : try to utilize more of the shared memory using smaller views · Updated 2023-06-26 14:23:04 -05:00    upstream-archive

9474
1
 ZIP  TAR.GZ
avoid-gnu-source

78fafcaf10 · ggml : do not use _GNU_SOURCE gratuitously · Updated 2023-06-25 09:21:02 -05:00    upstream-archive

9482
1
 ZIP  TAR.GZ
fix_clblast

20054a38c1 · Fix directory name · Updated 2023-05-26 18:00:08 -05:00    upstream-archive

9632
1
 ZIP  TAR.GZ
chunks

a1cdd29cd2 · ggml : rms_norm in chunks · Updated 2023-05-20 02:15:54 -05:00    upstream-archive

9653
2
 ZIP  TAR.GZ
steering

95dc4d7270 · Merge 'origin/master' into steering · Updated 2023-05-19 15:19:57 -05:00    upstream-archive

9655
9
 ZIP  TAR.GZ
f16c

40ec4882c4 · ggml : use F16C conversion when available · Updated 2023-05-17 12:05:51 -05:00    upstream-archive

9664
1
 ZIP  TAR.GZ
dequantize-matmul-3-gg

a3e6d62283 · cuda : alternative q4_q8 kernel · Updated 2023-05-12 09:02:39 -05:00    upstream-archive

9698
8
 ZIP  TAR.GZ
remove-vzip

e116eb638c · ggml : speed-up Q5_0 + Q5_1 at 4 threads · Updated 2023-05-11 10:51:56 -05:00    upstream-archive

9700
20
 ZIP  TAR.GZ
jed/spm-clblast

4baa85633a · Fix build · Updated 2023-05-06 20:44:07 -05:00    upstream-archive

9708
5
 ZIP  TAR.GZ
ci_cublas

31ff9e2e83 · ci : add cublas to windows release · Updated 2023-05-03 16:21:20 -05:00    upstream-archive

9723
1
 ZIP  TAR.GZ
q4_3-range-fix

102cd98074 · ggml : Q4_3c using 2x "Full range" approach · Updated 2023-04-23 06:56:44 -05:00    upstream-archive

9804
8
 ZIP  TAR.GZ
q4_0-q4_2-range-fix

71e6ae3779 · ggml : continue from #729 (wip) · Updated 2023-04-22 10:49:07 -05:00    upstream-archive

9804
7
 ZIP  TAR.GZ
gg/rmse_quantization

a0242a833c · Minor, plus rebase on master · Updated 2023-04-22 09:07:10 -05:00    upstream-archive

9804
2
 ZIP  TAR.GZ
quant-attn

4b8d5e3890 · llama : quantize attention results · Updated 2023-04-22 03:35:13 -05:00    upstream-archive

9809
1
 ZIP  TAR.GZ
mmap-pages-stats

1506737499 · Add mmap pages stats (disabled by default) · Updated 2023-04-16 11:22:30 -05:00    upstream-archive

9859
1
 ZIP  TAR.GZ
flash-attn

36ddd12924 · llama : add flash attention (demo) · Updated 2023-04-05 14:12:04 -05:00    upstream-archive

9925
1
 ZIP  TAR.GZ
mmap

c9c820ff36 · Added support for _POSIX_MAPPED_FILES if defined in source (#564) · Updated 2023-03-28 16:26:25 -05:00    upstream-archive

10159
8
 ZIP  TAR.GZ
q4_1_more_accel

4aeee216fd · Regroup q4_1 dot addition for better numerics. · Updated 2023-03-24 15:20:57 -05:00    upstream-archive

10040
2
 ZIP  TAR.GZ
q4_1_more_accel_kahan

66ea164e1d · Kahan summation on Q4_1 · Updated 2023-03-22 22:28:51 -05:00    upstream-archive

10067
2
 ZIP  TAR.GZ
q4_1_more_accel_loopsplit

711224708d · Break up loop for numeric stability · Updated 2023-03-22 21:14:44 -05:00    upstream-archive

10067
2
 ZIP  TAR.GZ