upstream-archive/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggml-org/llama.cpp.git synced 2026-08-01 23:50:47 -05:00
Code Issues Packages Projects Releases Wiki Activity
10,223 Commits 645 Branches 6,962 Tags

Default Branch

master

11924d4c17 · test: fix some CI errors (#26415) · Updated 2026-08-01 17:16:29 -05:00
 ZIP  TAR.GZ

Branches

compilade/requirements-cpu-torch

a44f22e7d3 · py : use cpu-only torch in requirements.txt · Updated 2024-07-06 10:18:03 -05:00    upstream-archive

6898
1
 ZIP  TAR.GZ
gg/indent

f55b647300 · llama : minor indentation during tensor loading · Updated 2024-07-04 11:34:04 -05:00    upstream-archive

6920
16
 ZIP  TAR.GZ
sl/fix-ppl-seq-max

dcab343f2f · use 1 seq for kl_divergence · Updated 2024-07-03 09:22:58 -05:00    upstream-archive

6935
2
 ZIP  TAR.GZ
fairydreaming/t5-clean-3-gg

703764a382 · convert : use non-fast T5 tokenizer · Updated 2024-07-02 11:29:26 -05:00    upstream-archive

6977
10
 ZIP  TAR.GZ
gg/nix-remove-opencl

d4a1923d4e · minor : remove parentheses · Updated 2024-07-01 06:45:55 -05:00    upstream-archive

6955
2
 ZIP  TAR.GZ
add-gemma2-soft-capping

51f0bd50a1 · Remove custom pre attention scaling and use computed value instead. · Updated 2024-06-29 22:02:50 -05:00    upstream-archive

6958
10
 ZIP  TAR.GZ
codeplay/tg-warmup

712e4d9450 · Generate full token count during warm up · Updated 2024-06-28 07:29:00 -05:00    upstream-archive

6961
1
 ZIP  TAR.GZ
gg/fix-devops

65f9293d14 · devops : remove clblast + LLAMA_CUDA -> GGML_CUDA · Updated 2024-06-26 11:17:26 -05:00    upstream-archive

6987
1
 ZIP  TAR.GZ
sl/zero-max-size

1e6e363d7f · test zero max buffer size · Updated 2024-06-26 10:11:09 -05:00    upstream-archive

6988
1
 ZIP  TAR.GZ
sycl-mul-mat-id

ff0aa3abd1 · fix part of mul_mat_id · Updated 2024-06-20 22:38:00 -05:00    upstream-archive

7031
1
 ZIP  TAR.GZ
sl/test-mul-mat-backend

f3974cabac · all matrix multiplication backend · Updated 2024-06-18 06:18:26 -05:00    upstream-archive

7072
1
 ZIP  TAR.GZ
codeplay/fix-matmul-arith

ce6e28cc23 · Update ggml-sycl.cpp · Updated 2024-06-18 03:57:14 -05:00    upstream-archive

7083
6
 ZIP  TAR.GZ
gg/fa-req-kq-hs

ef79941ac9 · llama : disable FA if KV head size do not match · Updated 2024-06-17 11:20:24 -05:00    upstream-archive

7053
1
 ZIP  TAR.GZ
codeplay/dequant_q4_K_improvements

a235b7c532 · Vectorize q load · Updated 2024-06-17 04:30:40 -05:00    upstream-archive

7083
11
 ZIP  TAR.GZ
gg/no-char32_t

98f948b9d0 · unicode : avoid char32_t · Updated 2024-06-16 05:18:46 -05:00    upstream-archive

7068
1
 ZIP  TAR.GZ
gg/ggml-fix-zero-blocks

28f7a4d028 · ggml : fix handling of zero blocks in IQ quants · Updated 2024-06-16 02:41:53 -05:00    upstream-archive

7069
1
 ZIP  TAR.GZ
gg/bitnet

e9f2abfc8c · bitnet : pad tensors to 256 · Updated 2024-06-15 11:01:03 -05:00    upstream-archive

7087
25
 ZIP  TAR.GZ
gg/rpc-fix-misaligned

34bdbed481 · rpc : fix load/store misaligned addresses · Updated 2024-06-15 06:39:20 -05:00    upstream-archive

7071
1
 ZIP  TAR.GZ
gg/metal-mmid-max-rows

eaf34ba0cd · metal : utilize max shared memory for mul_mat_id · Updated 2024-06-14 05:02:25 -05:00    upstream-archive

7078
1
 ZIP  TAR.GZ
codeplay/revert-host-alloc

18133cab40 · Revert "use the correct SYCL context for host USM allocations" · Updated 2024-06-13 06:08:27 -05:00    upstream-archive

7083
4
 ZIP  TAR.GZ