Tags

Tags give the ability to mark specific points in history as being important

b1748

83e633c2 · llama : differentiate the KV dims in the attention (#4657) · Jan 02, 2024
b1747

32866c5e · editorconfig : fix whitespace and indentation #4710 · Jan 02, 2024
b1746

5d7002d4 · server : add --override-kv parameter (#4710) · Jan 02, 2024
b1743

58ba655a · metal : enable shader debugging (cmake option) (#4705) · Jan 02, 2024
b1742

edd1ab7b · flake.lock: update · Dec 31, 2023
b1732

e39106c0 · ggml : add ggml_vdotq_s32 alias (#4715) · Dec 31, 2023
b1731

9fbda719 · clip : refactor + bug fixes (#4696) · Dec 30, 2023
b1730

39d8bc71 · CUDA: fixed tensor cores not being used on RDNA3 (#4697) · Dec 30, 2023
b1729

24a447e2 · ggml : add ggml_cpu_has_avx_vnni() (#4589) · Dec 30, 2023
b1728

a20f3c74 · CUDA: fix tensor core logic for Pascal and HIP (#4682) · Dec 29, 2023
b1727

0235b9b5 · clip : use ggml_backend_buffer_is_host (#4205) · Dec 29, 2023
b1726

ce18d727 · clip : enable gpu backend (#4205) · Dec 29, 2023
b1725

91bb39ce · cuda: fix vmm oom issue on NVIDIA AGX Orin (#4687) · Dec 29, 2023
b1724

04ac0607 · python : add check-requirements.sh and GitHub workflow (#4585) · Dec 29, 2023
b1723

68eccbdc · flake.nix : rewrite (#4605) · Dec 29, 2023
b1722

97bbca6e · cmake : fix ld warning duplicate libraries libllama.a (#4671) · Dec 29, 2023
b1721

4af48015 · llava-cli : refactor to use sampling library (#4669) · Dec 29, 2023
b1720

db49ff8e · server : replace sleep with condition variables (#4673) · Dec 29, 2023
b1719

60f55e88 · server : fix OpenAI server sampling w.r.t. penalty. (#4675) · Dec 29, 2023
b1718

b93edd22 · server : allow to generate multimodal embeddings (#4681) · Dec 29, 2023

Prev
1
…
37
38
39
40
41
42
43
44
45
…
99
Next