ggerganov/llama.cpp

Pull Requests Commits
llama : switch KQ multiplication to F32 precision by default (#10015)
ggerganov committed 1 year ago
Verified 8841ce3f
sync : ggml
ggerganov committed 1 year ago
Verified cc2983d3
increase cuda_cpy block size (ggml/996)
bssrdf committed 1 year ago
Verified 8c60a8a4
scripts : fix amx sync [no ci]
ggerganov committed 1 year ago
Verified 9e4a2563
metal : support permuted matrix multiplicaions (#10033)
ggerganov committed 1 year ago
Verified 66875035
llama : add DRY sampler (#9702)
wwoodsTM committed 1 year ago
Verified ff252ea4
llama: string_split fix (#10022)
Xarbirus committed 1 year ago
Verified d80fb71f
llamafile : extend sgemm.cpp support for Q5_0 models (#10010)
Srihari-mcw committed 1 year ago
Verified 2f8bd2b9
server : check that the prompt fits in the slot's context (#10030)
ggerganov committed 1 year ago
Verified bc5ba007
server : refactor slot input data, move tokenizer to HTTP thread (#10023)
ngxson committed 1 year ago
Verified 958367bf
ci : fix cmake flags for SYCL
ggerganov committed 1 year ago
Verified 40f25557
CUDA: fix insufficient buffer clearing for MMQ (#10032)
JohannesGaessler committed 1 year ago
Verified 167a5156
CUDA: fix MMQ for non-contiguous src0, add tests (#10021)
JohannesGaessler committed 1 year ago
Verified c39665f5
server : samplers accept the prompt correctly (#10019)
wwoodsTM committed 1 year ago
Verified 0a1c750c
sync : ggml
ggerganov committed 1 year ago
Verified 190a37d7
llama.vim : bump generation time limit to 3s [no ci]
ggerganov committed 1 year ago
Verified 2d3aba9e
CUDA: fix 1D im2col, add tests (ggml/993)
JohannesGaessler committed 1 year ago
80273a30
ggml : remove redundant set of contexts used field (ggml/978)
danbev committed 1 year ago
c19af0ac
llama.vim : add classic vim support (#9995)
m18coppola committed 1 year ago
Verified ac113a0f
metal : add POOL2D and fix IM2COL (#9943)
junhee-yoo committed 1 year ago
Verified 4c9388fb
flake.lock: Update
github-actions[bot] committed 1 year ago
873279b1
llama : fix empty batch causing llama_batch_allocr to crash (#9966)
ngxson committed 1 year ago
Verified c8c07d65
llama : rename batch to ubatch (#9950)
danbev committed 1 year ago
Verified 19d900a7
Rwkv chat template fix (#10001)
MollySophia committed 1 year ago
Verified 11d47057
lora : warn user if new token is added in the adapter (#9948)
ngxson committed 1 year ago
Verified c421ac07
llama : add chat template for RWKV-World + fix EOT (#9968)
MollySophia committed 1 year ago
Verified 4ff7fe1f
[CANN] Adapt to dynamically loadable backends mechanism (#9970)
leo-pony committed 1 year ago
Verified 6b844735
arg : fix typo in embeddings argument help [no ci] (#9994)
danbev committed 1 year ago
Verified 674804a9
llama.vim : fix info text display [no ci] (#9787)
ggerganov committed 1 year ago
Verified e94a138d
llama.vim : move info to the right of screen [no ci] (#9787)
ggerganov committed 1 year ago
Verified e01c67af
Terms Of Use Privacy Policy Refund Policy Impressum