root/llama.cpp
1
0
Fork 0
mirror of https://github.com/ggerganov/llama.cpp.git synced 2024-12-24 10:24:35 +00:00
Code Issues Actions 5 Packages Projects Releases Wiki Activity
4,388 Commits 359 Branches 2,865 Tags 382 MiB

Default Branch

master

30caac3a68 · llama : the WPM vocabs use the CLS token as BOS (#10930) · Updated 2024-12-24 07:44:20 +00:00
 ZIP  TAR.GZ

Branches

compilade/pyright-fix-ignores

a34fc0dd86 · ci : reduce severity of unused Pyright ignore comments · Updated 2024-09-30 17:59:40 +00:00
537
1
 ZIP  TAR.GZ
gg/tfs-ob1

114ab6347e · sampling : fix off-by-one in tail-free sampling · Updated 2024-09-23 08:44:55 +00:00
581
1
 ZIP  TAR.GZ
gg/tokenizer-cleanup

6e873e561a · llama : make llm_tokenizer more private · Updated 2024-09-20 08:41:51 +00:00
600
2
 ZIP  TAR.GZ
sl/fix-omp-one-thread

6b0248c29a · Update ggml/src/ggml.c · Updated 2024-09-18 16:00:26 +00:00
605
2
 ZIP  TAR.GZ
fix_ctx_default
Some checks failed
flake8 Lint / Lint (push) Has been cancelled
Details

a6a8f8d09c · Update docs/backend/SYCL.md · Updated 2024-09-17 08:25:43 +00:00
637
2
 ZIP  TAR.GZ
gg/cb-naming
Some checks failed
Python check requirements.txt / check-requirements (push) Has been cancelled
Details
flake8 Lint / Lint (push) Has been cancelled
Details
Python Type-Check / pyright type-check (push) Has been cancelled
Details

cc1c017191 · naming : normalize the name of callback-related identifiers · Updated 2024-09-16 06:11:42 +00:00
625
1
 ZIP  TAR.GZ
sycl-cmake-append

73ef3f769c · Update llama-server-intel.Dockerfile · Updated 2024-09-15 15:21:46 +00:00
630
3
 ZIP  TAR.GZ
gg/cmake-dedup-link
Some checks failed
flake8 Lint / Lint (push) Has been cancelled
Details

fb8f142554 · one more CMAKE_CXX_FLAGS fix (#9471) · Updated 2024-09-13 13:13:07 +00:00
639
5
 ZIP  TAR.GZ
gg/ggml-atomic-int

d7c042d1ae · ggml : make n_threads_cur atomic_int · Updated 2024-09-11 18:12:11 +00:00
655
1
 ZIP  TAR.GZ
gg/ggml-rework-cgraph

f9968f661d · ggml : update comments [no ci] · Updated 2024-09-11 10:16:39 +00:00
668
5
 ZIP  TAR.GZ
compilade/imatrix-batched-chunks

2d79a7077c · quantize : use unused imatrix chunk_size with LLAMA_TRACE · Updated 2024-09-10 16:09:17 +00:00    root

685
13
 ZIP  TAR.GZ
SVE-vector-length-agnostic-VLA-gg

cfbf33a705 · ggml : style changes + fix 512-bit nb loop check · Updated 2024-09-09 09:50:35 +00:00    root

729
4
 ZIP  TAR.GZ
sl/fix-rpc-nkvo

c3e2bb6dcf · rpc : fix nkvo · Updated 2024-09-07 01:24:47 +00:00    root

710
1
 ZIP  TAR.GZ
fix-ninja-metallib-build

b979fc97ba · cmake : use ggml-metal.metal from source dir to build default.metallib · Updated 2024-09-05 16:17:56 +00:00    root

719
1
 ZIP  TAR.GZ
compilade/bitnet-ternary

75b3a09602 · test-backend-ops : add TQ1_0 and TQ2_0 comments for later · Updated 2024-09-04 19:00:21 +00:00    root

721
33
 ZIP  TAR.GZ
gg/llama-refactor-sampling

f648ca2cee · llama : add llama_sampling API + move grammar in libllama · Updated 2024-09-03 07:31:54 +00:00    root

728
1
 ZIP  TAR.GZ
gg/llama-disambiguate

40fa68cb46 · readme : add API change notice · Updated 2024-09-02 15:32:24 +00:00    root

737
3
 ZIP  TAR.GZ
compilade/refactor-kv-cache

375de5b1f8 · llama : use unused n_embd_k_gqa in k_shift · Updated 2024-09-02 01:59:24 +00:00    root

737
41
 ZIP  TAR.GZ
gg/metal-fix-fa-2

a95225cdfd · metal : another fix for the fa kernel · Updated 2024-08-26 12:08:38 +00:00    root

761
1
 ZIP  TAR.GZ
gg/metal-fix-fa

aa931d0375 · metal : fix fa kernel · Updated 2024-08-26 10:09:50 +00:00    root

761
1
 ZIP  TAR.GZ