ggml-org/llama.cpp

Pull Requests Commits
ggml : workaround for missing _mm256_setr_m128i in GCC < 8 (#1638)
xingchensong committed 2 years ago
Verified ef3171d1
make : add SSSE3 compilation use case (#1659)
rankaiyx committed 2 years ago
Verified 555275a6
OpenCL: Add release memory (#1741)
edp1096 committed 2 years ago
Verified 98ed1655
Windows nvcc workaround (#1753)
JohannesGaessler committed 2 years ago
Verified ae9663f1
metal : fix build "tanhf" -> "tanh"
ggerganov committed 2 years ago
Verified b33dee28
metal : add GELU implementation (#1770)
manyoso committed 2 years ago
Verified 92f44ff7
metal : faster q4_0 (#1775)
ikawrakow committed 2 years ago
Verified 245fc3c3
metal : add Q2_K implementation (#1762)
ikawrakow committed 2 years ago
Verified 72ff5282
Revert "ggml : load data into int8x16x4_t using vld4q_s8 on arm64 (#1738)"
ggerganov committed 2 years ago
Verified 0bf7cf1b
ggml : load data into int8x16x4_t using vld4q_s8 on arm64 (#1738)
lindeer committed 2 years ago
Verified 8432d4d9
metal : Q6_K implementation (#1752)
ikawrakow committed 2 years ago
Verified 0f291e1f
Add llama.cpp docker support for non-latin languages (#1673)
qingfengfenga committed 2 years ago
Verified 8fc81799
ggml : fix fprintf warnings (#1720)
sroussey committed 2 years ago
Verified b50b570e
clang-tidy : restore dot file from accidental deletion
ggerganov committed 2 years ago
Verified 53aba3f3
metal : add Q4_K implementation (#1733)
ikawrakow committed 2 years ago
Verified 4161bdc0
k-quants : add missing compile definition to CMakeLists (#1748)
johnson442 committed 2 years ago
Verified 00358582
k-quants : allow to optionally disable at compile time (#1734)
ggerganov committed 2 years ago
Verified 5c64a095
flake : update to support metal on m1/m2 (#1724)
jpetrucciani committed 2 years ago
Verified 5b57a5b7
readme : add June roadmap
ggerganov committed 2 years ago
Verified 4dc62c54
main: add the possibility to open the prompt cache read-only (#1640)
wtarreau committed 2 years ago
Verified 35a84916
llama : fix vram_scratch var
ggerganov committed 2 years ago
Verified 2d7bf110
llama : fix compile warnings
ggerganov committed 2 years ago
Verified 2a4e41a0
Multi GPU support, CUDA refactor, CUDA scratch buffer (#1703)
JohannesGaessler committed 2 years ago
Verified 17366df8
metal : add f16 support
ggerganov committed 2 years ago
Verified 44f906e8
Clblast fixes + enhancements to save VRAM and offload more layers (#1675)
LostRuins committed 2 years ago
Verified d5b111f5
ggml : fix builds, add ggml-quants-k.o (close #1712, close #1710)
ggerganov committed 2 years ago
Verified 2d43387d
gitignore : add .clang-tidy
ggerganov committed 2 years ago
Verified 7ad7750c
llama : temporary disable Q6_K output quantization (#1711)
ggerganov committed 2 years ago
Verified 7a74dee6
metal : add checks for buffer size (#1706)
spencersutton committed 2 years ago
Verified 590250f7
docs : add performance troubleshoot + example benchmark documentation (#1674)
Yuval-Peled committed 2 years ago
Verified f4c55d3b
Terms Of Use Privacy Policy Refund Policy Impressum