Releases · liuwei-git/llama.cpp · GitHub

21 May 18:38

b2977 Latest

Latest

convert : read/write n_head_kv

Assets 21

cudart-llama-bin-win-cu11.7.1-x64.zip

293 MB 2024-05-21T18:38:04Z
cudart-llama-bin-win-cu12.2.0-x64.zip

413 MB 2024-05-21T18:38:11Z
llama-b2977-bin-macos-arm64.zip

41.3 MB 2024-05-21T18:38:19Z
llama-b2977-bin-macos-x64.zip

37.9 MB 2024-05-21T18:38:20Z
llama-b2977-bin-ubuntu-x64.zip

45.8 MB 2024-05-21T18:38:21Z
llama-b2977-bin-win-avx-x64.zip

6.66 MB 2024-05-21T18:38:23Z
llama-b2977-bin-win-avx2-x64.zip

6.64 MB 2024-05-21T18:38:23Z
llama-b2977-bin-win-avx512-x64.zip

6.66 MB 2024-05-21T18:38:24Z
llama-b2977-bin-win-clblast-x64.zip

7.83 MB 2024-05-21T18:38:25Z
llama-b2977-bin-win-cuda-cu11.7.1-x64.zip

65 MB 2024-05-21T18:38:25Z
Source code (zip)

2024-05-21T16:40:01Z
Source code (tar.gz)

2024-05-21T16:40:01Z

21 May 18:20

b2976

cuda : fix compile warning

Assets 21

21 May 18:20

b2975

llama : remove tmp assert

Assets 21

21 May 18:19

b2974

llama : move rope factors from KV header to tensors

Assets 21