ggerganov/llama.cpp

Pull Requests Commits
server : parallel decoding and multimodal (#3677)
ggerganov committed 2 years ago
Verified 438c2ca8
Add test for MPT tokenization (#3728)
goerch committed 2 years ago
Verified 9e70cc03
readme : remove unsupported node.js library (#3703)
ianscrivener committed 2 years ago
Verified 5a42a5f8
llama : validate special token ids are in range when loading GGUF model (#3635)
KerfuffleV2 committed 2 years ago
Verified a5e7dbd6
main : escape prompt for cfg_negative_prompt and consecutive inputs in main with interactive (#3623)
vvhg1 committed 2 years ago
Verified d3956aea
batched : add len CLI argument
ggerganov committed 2 years ago
Verified 22c69a27
CLBlast: Add outer loops over src0 for broadcasting in mulmat
shibe2 committed 2 years ago
465219b9
sampling : refactor init to use llama_sampling_params (#3696)
ggerganov committed 2 years ago
Verified d1031cf4
gguf : support big endian platform (#3552)
chenqiny committed 2 years ago
Verified 8cf19d60
server : fix uninitialized sampling context (close #3685)
ggerganov committed 2 years ago
Verified a0edf73b
ggml : fix rope + llama minor optimizations (#3560)
GermanAizek committed 2 years ago
Verified f439e506
convert : restore compat with old Falcon models (#3680)
cebtenzzre committed 2 years ago
Verified e78f3ef2
multimodal : add BakLLaVA conversion support (#3682)
monatis committed 2 years ago
Verified f3b25e40
llava : avoid segfault in case of non-existent mmproj file (#3674)
monatis committed 2 years ago
Verified 60abea97
readme : update hot topics
ggerganov committed 2 years ago
Verified 004797f6
speculative : bug fixes
ggerganov committed 2 years ago
Verified 4e82b2ea
speculative : add tree-based sampling example (#3624)
ggerganov committed 2 years ago
Verified 0e89203b
metal : implement q5_0 and q5_1 kernels (#3648)
jhen0409 committed 2 years ago
Verified c67fe68e
opencl : fix element-wise multiplication (#3656)
shibe2 committed 2 years ago
Verified 1117d066
fix embeddings when using CUDA (#3657)
slaren committed 2 years ago
Verified cb33f43a
llama : avoid fprintf in favor of LLAMA_LOG (#3538)
ggerganov committed 2 years ago
Verified e1675d13
readme : update hot-topics & models, detail windows release in usage (#3615)
BarfingLemurs committed 2 years ago
Verified 8402566a
CLBlast: Fix temporary buffer size for f16 conversion (wsize)
shibe2 committed 2 years ago
40e5ce05
train-text-from-scratch : fix assert failure in ggml-alloc (#3618)
slaren committed 2 years ago
Verified a5e8c1d8
editorconfig : remove trailing spaces
ggerganov committed 2 years ago
Verified e74c705e
server : documentation of JSON return value of /completion endpoint (#3632)
coezbek committed 2 years ago
Verified 3ad1e3f1
save-load-state : fix example + add ci test (#3655)
ggerganov committed 2 years ago
Verified 1142013d
readme : add Aquila2 links (#3610)
ftgreat committed 2 years ago
Verified 5fe268a4
tokenizer : special token handling (#3538)
staviq committed 2 years ago
Verified 1a159553
k-quants : fix quantization ranges (#3646)
ggerganov committed 2 years ago
Verified 281ef73c
Terms Of Use Privacy Policy Refund Policy Impressum