ggml-org/llama.cpp

Pull Requests Commits
gguf-py: Support 01.AI Yi models (#3943)
KerfuffleV2 committed 2 years ago
Verified f28af0d8
metal : round up to 16 to fix MTLDebugComputeCommandEncoder assertion (#3938)
psugihara committed 2 years ago
Verified d9b33fe9
ggml-metal: fix yarn rope (#3937)
jxy committed 2 years ago
Verified 5ba37461
ggml-cuda : move row numbers to x grid dim in mmv kernels (#3921)
slaren committed 2 years ago
Verified abb77e73
speculative : change default p_accept to 0.5 + CLI args (#3919)
ggerganov committed 2 years ago
Verified 8f961abd
common : YAYF (yet another YARN fix) (#3925)
ggerganov committed 2 years ago
Verified 05816027
llama : change yarn_ext_factor placeholder to -1 (#3922)
cebtenzzre committed 2 years ago
Verified 3fdbe6b6
cuda : add ROCM aliases for CUDA pool stuff (#3918)
KerfuffleV2 committed 2 years ago
Verified 629f917c
cmake : fix relative path to git submodule index (#3915)
abetlen committed 2 years ago
Verified 51b2fc11
readme : add notice about #3912
ggerganov committed 2 years ago
Verified 224e7d5b
cuda : fix const ptrs warning causing ROCm build issues (#3913)
ggerganov committed 2 years ago
Verified c7743fe1
cuda : use CUDA memory pool with async memory allocation/deallocation when available (#3903)
young-developer committed 2 years ago
Verified d6069051
gguf : print error for GGUFv1 files (#3908)
ggerganov committed 2 years ago
Verified 4ff1046d
cmake : disable LLAMA_NATIVE by default (#3906)
slaren committed 2 years ago
Verified 21958bb3
gguf : remove special-case code for GGUFv1 (#3901)
ggerganov committed 2 years ago
Verified 2756c4fb
llm : prevent from 1-D tensors being GPU split (#3697)
ggerganov committed 2 years ago
Verified 1efae9b7
build : link against build info instead of compiling against it (#3879)
cebtenzzre committed 2 years ago
Verified b12fa0d1
cuda : check if this fixes Pascal card regression (#3882)
ggerganov committed 2 years ago
Verified 4d719a6d
metal : fix build errors and kernel sig after #2268 (#3898)
ggerganov committed 2 years ago
Verified 183b3fac
cuda : fix RoPE after #2268 (#3897)
cebtenzzre committed 2 years ago
Verified 2fffa0d6
llama : fix llama_context_default_params after #2268 (#3893)
cebtenzzre committed 2 years ago
Verified 0eb332a1
ggml-cuda : compute ptrs for cublasGemmBatchedEx in a kernel (#3891)
slaren committed 2 years ago
Verified d02e98cd
llama : implement YaRN RoPE scaling (#2268)
cebtenzzre committed 2 years ago
Verified 898aeca9
llm : fix llm_build_kqv taking unused tensor (benign, #3837)
ggerganov committed 2 years ago
Verified c43c2da8
llm : fix falcon norm after refactoring (#3837)
ggerganov committed 2 years ago
Verified 523e49b1
metal : multi-simd softmax (#3710)
ggerganov committed 2 years ago
Verified e16b9fa4
common : minor (#3715)
ggerganov committed 2 years ago
Verified ff8f9a88
llm : add llm_build_context (#3881)
ggerganov committed 2 years ago
Verified 50337961
common : allow caller to handle help/argument exceptions (#3715)
bandoti committed 2 years ago
Verified 0e40806c
log : make generating separate log files optional (#3787)
staviq committed 2 years ago
Verified a2758d08
Terms Of Use Privacy Policy Refund Policy Impressum