ggerganov/llama.cpp

Pull Requests Commits
feat: cuda implementation for `ggml_conv_transpose_1d` (ggml/854)
balisujohn committed 2 years ago
fde13b3b
common : preallocate sampling token data vector (#8363)
kevmo314 committed 2 years ago
Verified 470939d4
infill : assert prefix/suffix tokens + remove old space logic (#8351)
ggerganov committed 2 years ago
Verified 6f0dbf6a
common : avoid unnecessary logits fetch (#8358)
kevmo314 committed 2 years ago
Verified ffd00797
readme : add supported glm models (#8360)
youth123 committed 2 years ago
Verified 04ce3a8b
py : type-check all Python scripts with Pyright (#8341)
compilade committed 2 years ago
Verified 3fd62a6b
Update llama-cli documentation (#8315)
dspasyuk committed 2 years ago
Verified a8db2a9c
ci : add checks for cmake,make and ctest in ci/run.sh (#8200)
AlexsCode committed 2 years ago
Verified 4090ea55
readme : update bindings list (#8222)
andy-tai committed 2 years ago
Verified f1948f1e
gguf-hash: model wide and per tensor hashing using xxhash and sha1 (#8048)
mofosyne committed 2 years ago
Verified f7cab35e
llama : support glm3 and glm4 (#8031)
youth123 committed 2 years ago
Verified 905942ab
llama : fix n_rot default (#8348)
ggerganov committed 2 years ago
Verified b5040086
py : use cpu-only torch in requirements.txt (#8335)
compilade committed 2 years ago
Verified d39130a3
finetune: Rename command name in README.md (#8343)
standby24x7 committed 2 years ago
Verified b81ba1f9
finetune: Rename an old command name in finetune.sh (#8344)
standby24x7 committed 2 years ago
Verified 210eb9ed
server: Retrieve prompt template in /props (#8337)
bviksoe committed 2 years ago
Verified cb4d86c4
added support for Authorization Bearer tokens when downloading model (#8307)
dwoolworth committed 2 years ago
Verified 86e7299e
update main readme (#8333)
ngxson committed 2 years ago
Verified 60d83a01
llama : add early return for empty range (#8327)
danbev committed 2 years ago
Verified 87e25a1d
Detokenizer fixes (#8039)
jaime-m-p committed 2 years ago
Verified 213701b5
Reorganize documentation pages (#8325)
ngxson committed 2 years ago
Verified be20e7f4
llama : fix compile warning (#8304)
ggerganov committed 2 years ago
Verified 7ed03b89
cmake : add GGML_BUILD and GGML_SHARED macro definitions (#8281)
akemimadoka committed 2 years ago
Verified 1d894a79
Enabled more data types for oneMKL gemm_batch (#8236)
OuadiElfarouki committed 2 years ago
Verified 1f3e1b66
convert : remove AWQ remnants (#8320)
ggerganov committed 2 years ago
Verified 148ec970
llama : minor indentation during tensor loading (#8304)
ggerganov committed 2 years ago
Verified 2cccbaa0
CUDA: MMQ support for iq4_nl, iq4_xs (#8278)
JohannesGaessler committed 2 years ago
Verified 8e558309
CUDA: revert part of the RDNA1 optimizations (#8309)
daniandtheweb committed 2 years ago
Verified 0a423800
llama : streamline embeddings from "non-embedding" models (#8087)
iamlemec committed 2 years ago
Verified d12f7810
CUDA: fix MMQ stream-k rounding if ne00 % 128 != 0 (#8311)
JohannesGaessler committed 2 years ago
Verified bcefa03b
Terms Of Use Privacy Policy Refund Policy Impressum